Folate and other methyl-donor pathway components are widely supplemented due to their ability to prevent prenatal neural tube defects. Several lines of evidence suggest that these supplements act through epigenetic mechanisms (e.g. altering DNA methylation). Primary among these are the experiments on the mouse viable yellow allele of the agouti locus (Avy). In the Avy allele, an Intracisternal A-particle retroelement has inserted into the genome adjacent to the agouti gene and is preferentially methylated. To further test these effects, we tested the same diet used in the Avy studies on wild-derived Peromyscus maniculatus, a native North American rodent. We collected tissues from neonatal offspring whose parents were fed the high-methyl donor diet as well as controls. In addition, we assayed coat-color of a natural variant (wide-band agouti = ANb) that overexpresses agouti as a phenotypic biomarker. Our data indicate that these dietary components affected agouti protein production, despite the lack of a retroelement at this locus. Surprisingly, the methyl-donor diet was associated with defects (e.g. ovarian cysts, cataracts) and increased mortality. We also assessed the effects of the diet on behavior: We scored animals in open field and social interaction tests. We observed significant increases in female repetitive behaviors. Thus these data add to a growing number of studies that suggest that these ubiquitously added nutrients may be a human health concern.
Citation: Shorter KR, Anderson V, Cakora P, Owen A, Lo K, Crossland J, et al. (2014) Pleiotropic Effects of a Methyl Donor Diet in a Novel Animal Model. PLoS ONE9(8): e104942. https://doi.org/10.1371/journal.pone.0104942
Editor: Claudia Kappen, Pennington Biomedical Research Center/LSU, United States of America
Received: March 9, 2014; Accepted: July 1, 2014; Published: August 14, 2014
Copyright: © 2014 Shorter et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All Data found in paper, or supplementary files, except for video files of behavioral analyses which are available upon request from the corresponding author.
Funding: This study was made possible by NIH P40 OD010961 and the University of South Carolina VP for research office. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Folic acid and related B vitamins are widely supplemented in the US and western countries due to their ability to prevent neural tube defects such as spina bifida . This consumption has increased over the last decade, due not only to direct supplementation (i.e. vitamin tablets/capsules) but also to enrichment of grains , , and addition to other products such as energy drinks e.g. (http://www.5hourenergy.com/QandA.asp).
While it is clear these compounds have beneficial effects, the underlying mechanisms are unknown. These molecules contribute to the 1-carbon/methyl donor pathway. This pathway contributes to many biological processes. Notably, these components are involved in production of SAM (S-Adenosyl Methionine), which is the ultimate donor responsible for adding methyl groups to proteins and nucleic acids. This and other data suggests that these nutrients act through epigenetic mechanisms, as methylation of DNA and histone amino acid residues are known to mediate epigenetic effects , .
These data include experiments on the viable yellow allele of the agouti locus (Avy) in the lab mouse. In the Avy allele, an Intracisternal A-particle (IAP) class retroelement has inserted into the genome adjacent to the Agouti (a) gene . The strength of the IAP promoter results in constitutive expression of the agouti locus. Thus, the Avy allele results in hair that is all yellow (as opposed to hairs having regions of both black and yellow) as well as obesity and tumor predisposition. Maternal consumption of a diet high in components of the 1-carbon/methyl donor pathway restores Avy animals to a wild-type appearance, presumably due to the observed increased DNA methylation of IAP promoter , , . A similar IAP insertion at the axin locus (AxinFu allele) is similarly affected by the diet .
Few such studies have been done on natural variants or examination of other potential effects of such a diet. Peromyscus are wild-derived North American rodents and thus represent natural populations/genomes in ways that more widely used models do not . Peromyscus have proven useful for evaluating the impacts of environmental factors. We therefore tested the (1X; Table 1) diet originally used in the Avy studies on P. maniculatus. We employed a naturally occurring variant termed wide-band agouti (ANb) as a biomarker for the effects of the diet , , . The ANb allele is otherwise on a BW (http://stkctr.biol.sc.edu/wild-stock/p_manicu_bw.html) genetic background, a P. maniculatus stock whose genome has recently been sequenced (http://www.ncbi.nlm.nih.gov/assembly/84591/) and mapped . Effects of the diet on the ANb animals would suggest general effects of the diet, as there is no evidence for a retroelement in this allele .
BW animals have a tendency towards repetitive behaviors (e.g. jumps, backflips), and thus have been used as a model for Autism Spectrum (ASD) and Obsessive-Compulsive disorders (OCD) , , , , , . We therefore wished to assess whether the diet overtly affected behavior in addition to potential effects on the ANb allele. These studies provide novel evidence of deleterious effects of large doses of these compounds typically considered therapeutic or preventive to disease.
All procedures were approved by the University of South Carolina Institutional Animal Care and Use Committee (IACUC; protocol #1809-100340-061011).
Animal Husbandry & Mating Schemes
Animals were taken from the stocks maintained at the Peromyscus Genetic Stock Center (http://stkctr.biol.sc.edu/). Animals were kept on a 16∶8 hour light-dark cycle and were given food and water ad libitum. Matings of BW female×ANb male were established and maintained on either the methyl donor diet (Table 1) or normal rodent chow (i.e. controls). Offspring were weaned at approximately 25 days of age and maintained on the methyl donor diet or normal rodent chow until reaching six months of age (to obviate any concerns about maturity of coat-color; note that these animals live >4 yrs). Other animals were sacrificed at birth for future nucleic acid analyses; additional tissues from both ages are available to interested investigators.
Offspring of the BW female×ANb male matings were evaluated in Open Field and Social Interaction Tests at 4–6 months of age, as previously described . We tested 62 experimental animals (39 ♀ & 23 ♂) and 30 controls (12 ♀ & 18 ♂). Briefly, these tests consisted of first placing a single animal into a standard rat (10.25″W×19″L×8″H) cage with aspen shavings and ventilated transparent cover. After five minutes of observation, we introduced a novel animal of the same sex and species. The subsequent five minute period constituted the social interaction test. The novel animal’s tail was marked with a non-toxic marker to distinguish it from the animal being tested. The cage was cleaned between each animal tested (including replacement of bedding).
All behaviors were recorded with a digital camcorder. We used the Noldus Observer XT software (http://www.noldus.com/) to score behaviors from the video data. For the open field test, we scored the following behaviors: burrowing, freezing, jumping, back-flipping, running in circles, and grooming. Based on these videos, we considered straight vertical jumping, back-flipping, and running circles as repetitive behaviors.
For the social interaction test videos, we scored the same behaviors as in the open field test with the addition of social and aggressive behaviors. General social behaviors included sniffing, following, and allogrooming. Aggressive behaviors included biting, chasing, boxing, and mounting.
All behaviors were scored by incidence; we assessed behavior type at five second intervals throughout the video. Three people scored each video; overall inter-rater reliability was at least 80 percent. At least two scorers were blind to the diet of the animals being scored. When specific behavioral assessments disagreed, we alternated accepting the assessment of the three scorers. The data collected by scoring videos were graphed with Microsoft Excel. Behaviors are reported as percentage of incidence of behavior. Statistics were calculated using the Minitab and SPSS software packages. Note that we used Kruskal–Wallis one-way analysis of variance in cases where there was clearly a non-normal distribution.
After behavioral testing, animals were euthanized via CO2 chamber. Whole pelts were taken in order to analyze coat color differences. Tissues (skin sample, brain, and liver) were obtained and flash frozen in liquid nitrogen.
Measurement of Agouti (Yellow) Band Lengths
Hair tufts were pulled from the dorsal midline behind the ears from each pelt. Tufts of hair were placed on a microscope beside a micrometer and pictures were taken using a light microscope/digital camera combination. Agouti (yellow) band lengths in the hair were measured in millimeters (mm). We assessed 67 experimental animals (40 ♀ & 27 ♂) and 30 controls (12 ♀ & 18 ♂).
Methyl Diet Affects Coat Color & Body Weight
Matings were established to obtain offspring heterozygous for the dominant ANb allele. As this allele results in higher expression of agouti, heterozygotes exhibit a longer yellow band of hair and thus overall lighter appearance. A number of animals raised on the methyl-donor diet exhibited visibly darker coats than the controls (Fig. 1A).
(A) Whole pelts and (B) corresponding hair tufts from representative six-month old female ANb methyl diet (#1) and control diet (#2) animals. Note the visible differences in yellow band length in hair tufts and size. (C) Distribution of yellow band lengths (in mm) in tufts of hair. T test was used to determine significance between methyl diet animals and control animals: t(107) = 15.9, p<0.005, d = 2.2. The calculated Cohen’s D value of 2.2 indicates a large treatment effect.
To quantify these changes, we prepared pelts and measured the yellow (agouti) band length on the dorsal midline from 67 methyl diet animals (40♀, 27♂) and 41 controls (18♀, 23♂; Fig. 1B). These data revealed that while the control ANb animals had yellow band lengths tightly clustered around 3.1 mm, the treatment group had a broader distribution with an average yellow band length of 2.21 mm (Fig. 1C). These differences were deemed significant by T-test (p<0.005).
A number of the methyl diet ANb animals appeared visibly larger than the controls. We therefore weighed the animals at the time of sacrifice (Fig. 2). Female methyl diet animals averaged 20.2 g compared to 18.7 g for control females; this shift was significant (p<0.05; t-test). Despite the presence of two much larger animals, the male methyl diet average (22.6 g) was essentially the same as the control average (22.0 g).
We weighed 68 experimental animals (40 ♀ & 28 ♂) and 40 controls (12 ♀ & 18 ♂) at six months of age. The difference between female experimental & female control (ctrl) was significant (p<0.05; t-test), male averages were not significant. However, there were two methyl-diet males that were much larger than the control population.
Abnormalities & Mortality
Unexpectedly, we noted that a number of methyl-donor animals died between weaning and adult assessments of coat-color and behavior (4–6 months). While mortality was especially pronounced in males (p<0.001; Table 2), it was also significant in females (p = 0.005). Note that there was no mortality in control animals over this time period (P. maniculatus live 4–5 years in captivity).
When we took tissues from sacrificed animals for nucleic acid analyses, we noted a number of abnormalities in methyl diet animals not present in controls (Table 2). Again, the number was higher in methyl diet males (9 of 28 methyl diet males had at least one abnormality; p<0.005), but also significant in females (5 of 40 methyl diet females had at least one abnormality; p<0.01). These apparent defects (Table 2) were varied, and showed no effect of litter (i.e. were randomly distributed between the litters). They included ovarian cysts (Fig. 3A), size/consistency differences in ribcage, heart, and lungs (Fig. 3B), cataracts (Fig. 3C) and asymmetrical testes (Fig. 3D, E). In addition, we noted consistency differences in other organs (e.g. brain).
(A) Hemorrhagic ovarian cyst in a methyl diet female. (B) Normal diet animal’s ribcage, heart, and lungs (left) compared to one methyl diet animal’s ribcage, heart and lungs; note abnormalities in size and shape of lungs and heart. (C) Cataracts were visible in the left eye of some animals. (D) Left and right testes from a control diet male (top) and a methyl diet male (bottom). Chi squared tests suggest significant size differences between right and left testes in these three methyl diet males.
Methyl Diet Affects Behavior
Animals still alive at six months were subjected to a simple open-field test and social interaction test, as described . Major categories scored included repetitive behaviors (jumping, backflips, circle running) and general social behaviors (sniffing, following, allogrooming). We also assessed aggressive behaviors, including biting, boxing, mounting, and chasing.
Female methyl diet animals performed significantly higher numbers of repetitive behaviors than control diet females (Fig. 4A; p<0.01, Kruskal-Wallis test). Examples are shown in Video S1. Female methyl diet animals were, on average, more social, but this was not deemed significant (Fig. 4B; p = 0.064, Kruskal-Wallis). Similarly, male methyl diet animals trended towards more aggression than control diet males, but this was not statistically significant (p = 0.069, Kruskal-Wallis test). ANb animals are more aggressive and exhibit less repetitive behavior than standard BW animals . Thus, it is possible that some of these behavioral effects are due to suppression of the agouti (or a tightly linked) locus itself.
(A) Repetitive behaviors in each group tested. Repetitive behaviors included jumping, back-flipping, and running in circles. Female methyl diet animals performed significantly higher numbers of repetitive behaviors than control diet females (p<0.01, Kruskal-Wallis test). (B) Social behaviors and aggressive behaviors for each group tested. Social behaviors included sniffing, following, and allogrooming. Female methyl diet animals were, on average, less social, but this was statistically insignificant (p = 0.064, Kruskal-Wallis). Aggressive behaviors included biting, boxing, mounting, and chasing. Male methyl diet animals were, on average, more aggressive than control diet males, but this was statistically insignificant (p = 0.069). In both cases, bars represent standard error.
We set out to assess whether the methyl-donor diet would affect the Peromyscus natural agouti variant ANb in a similar manner to the Mus Avy and whether the behavior of these wild-derived animals was obviously altered by the diet. The data presented here further indicate that these dietary components do indeed affect the ANb agouti allele, although whether this is via DNA methylation, or even a cis-effect, is unknown (our preliminary data does not suggest significant DNA methylation changes at the agouti promoter). The apparent lack of a retroelement at this allele suggests more broad effects than the mouse Avy and axinFu studies. Further, female repetitive behavior and weights were significantly increased. Unexpectedly, the diet resulted in significant increases in mortality and abnormalities, with a greater effect in males.
The data presented here indicate that dietary intake of methyl-donors may have multiple adverse outcomes in a true wild-type mammalian model. To our knowledge, this is the first study to associate these particular defects, mortality or altered behavior in wild-type animals with these dietary factors.
We note that increasing evidence points to gene-environment interactions underlying the etiology of many diseases. Folic acid and other methyl-donor pathway components are typically thought of as preventing, rather than being causal to human health issues. Addition of these nutrients to flour appears to have dramatically reduced neural tube defects , and deficiencies are also thought to contribute to neuro-cognitive disorders . However, this study adds to a growing number of recent studies suggesting deleterious effects of developmental exposure to high doses of these compounds , , , , , , , . For example, mutations in some loci involved in neural tube development are exacerbated (rather than rescued) by excess folic acid , and neurons developmentally exposed to high folic acid may be more susceptible to seizure . Further, studies using these same components have shown increased colitis susceptibility and allergic airway disease (e.g. allergic asthma) in standard laboratory mice (C57BL/6J) , .
Through counting of food pellets consumed, we estimated that these animals took in approximately one food pellet per day. This amount is roughly equivalent to a human consuming around 1750–2000 micrograms of folic acid in a day (based on weight of the animals and 0.0043 grams folate/kg food). We note that such consumption is quite feasible, as many commercial supplements contain 800 micrograms folate (e.g. http://www.vitaminshoppe.com/p/folic-acid-800-mcg-100-capsules/vs-1148#.UwetE8pWQ7w), which are taken in addition to the amounts found in enriched flour and sports drinks. Other ingredients in this diet are also consumed in copious amounts. For example, the decaffeinated version of the popular 5- hour energy drink contains additional Vitamin B12 and choline in addition to folic acid (http://www.5hourenergy.com/healthfacts.asp?Product=decaf). While rodent and human metabolism differ substantially, it is worth considering whether these dietary components may contribute to human behavioral variation .
Clearly, much additional work is required to assess the scope and mechanisms of these adverse effects. For example, we are currently undertaking additional behavioral assays (e.g. Barnes Maze). Besides molecular characterization of these changes, we plan to test the dietary effects on an interfertile species (P. polionotus), which is more social and less prone to repetitive behaviors . We hypothesize that certain genotypes will be more susceptible to specific epimutations that result in neurological disorders or have other deleterious effects.
That is, we hypothesize that certain genotypes in combination with threshold amounts of these nutrients at specific developmental time points may result in negative effects. As observed in our studies, we predict that such effects will also be highly sexually dimorphic.
We thank Frances Lee for photography assistance. We thank Drs. Jay Gargus, Pauline Filipek, Rachel O’Neill and Chris Wiley for discussions; we thank the labs of Dr. Kim Creek and Dr. Sean Place for equipment use.
Conceived and designed the experiments: KRS PBV MRF. Performed the experiments: KRS VA PC AO KL JC MRF PBV. Analyzed the data: KRS PBV ACHS. Wrote the paper: PBV KRS.
- 1. Godwin KA, Sibbald B, Bedard T, Kuzeljevic B, Lowry RB, et al. (2008) Changes in frequencies of select congenital anomalies since the onset of folic acid fortification in a Canadian birth defect registry. Can J Public Health 99: 271–275.
- 2. Kim YI (2007) Folic acid fortification and supplementation–good for some but not so good for others. Nutrition Reviews 65: 504–511.
- 3. Lamers Y (2011) Folate recommendations for pregnancy, lactation, and infancy. Ann Nutr Metab 59: 32–37.
- 4. Waterland RA, Jirtle RL (2003) Transposable elements: targets for early nutritional effects on epigenetic gene regulation. Mol Cell Biol 23: 5293–5300.
- 5. Barua S, Kuizon S, Chadman KK, Flory MJ, Brown WT, et al. (2014) Single-base resolution of mouse offspring brain methylome reveals epigenome modifications caused by gestational folic acid. Epigenetics Chromatin 7: 3.
- 6. Michaud EJ, van Vugt MJ, Bultman SJ, Sweet HO, Davisson MT, et al. (1994) Differential expression of a new dominant agouti allele (Aiapy) is correlated with methylation state and is influenced by parental lineage. Genes Dev 8: 1463–1472.
- 7. Wolff GL, Kodell RL, Moore SR, Cooney CA (1998) Maternal epigenetics and methyl supplements affect agouti gene expression in Avy/a mice. The FASAB Journal 12: 949–957.
- 8. Cooney CA, Dave AA, Wolff GL (2002) Maternal methyl supplements in mice affect epigenetic variation and DNA methylation of offspring. The Journal of Nutrition 132: 2393S–2400S.
- 9. Waterland RA, Dolinoy DC, Lin JR, Smith CA, Shi X, et al. (2006) Maternal methyl supplements increase offspring DNA methylation at Axin Fused. Genesis 44: 401–406.
- 10. Vrana PB, Shorter KR, Szalai G, Felder MR, Crossland JP, et al. (2014) Peromyscus (Deer Mice) as Developmental Models. WIREs Developmental Biology 3: 211–230.
- 11. Robinson R (1981) The agouti alleles of Peromyscus. J Hered 72: 132.
- 12. Linnen CR, Kingsley EP, Jensen JD, Hoekstra HE (2009) On the origin and spread of an adaptive allele in deer mice. Science 325: 1095–1098.
- 13. Kenney-Hunt J, Lewandowski A, Glenn TC, Glenn JL, Tsyusko OV, et al. (2014) A genetic map of Peromyscus with chromosomal assignment of linkage groups (a Peromyscus genetic map). Mamm Genome 25: 160–179.
- 14. Linnen CR, Poh YP, Peterson BK, Barrett RD, Larson JG, et al. (2013) Adaptive evolution of multiple traits through multiple mutations at a single gene. Science 339: 1312–1316.
- 15. Powell SB, Newman HA, Pendergast JF, Lewis MH (1999) A rodent model of spontaneous stereotypy: initial characterization of developmental, environmental, and neurobiological factors. Physiology & Behavior 66: 355–363.
- 16. Powell SB, Newman HA, McDonald TA, Bugenhagen P, Lewis MH (2000) Development of spontaneous stereotyped behavior in deer mice: effects of early and late exposure to a more complex environment. Developmental Psychobiology 37: 100–108.
- 17. Tanimura Y, Yang MC, Lewis MH (2008) Procedural learning and cognitive flexibility in a mouse model of restricted, repetitive behaviour. Behavioural Brain Research 189: 250–256.
- 18. Tanimura Y, Yang MC, Ottens AK, Lewis MH (2010) Development and temporal organization of repetitive behavior in an animal model. Dev Psychobiol 52: 813–824.
- 19. Korff S, Stein DJ, Harvey BH (2008) Stereotypic behaviour in the deer mouse: pharmacological validation and relevance for obsessive compulsive disorder. Prog Neuropsychopharmacol Biol Psychiatry 32: 348–355.
- 20. Shorter KR, Owen A, Anderson V, Hall-South AC, Hayford S, et al. (2014) Natural Genetic Variation Underlying Differences in Peromyscus Repetitive and Social/Aggressive Behaviors. Behav Genet 44: 126–135.
- 21. Dice LR (1938) Variation in Nine Stocks of the Deer Mouse, Peromyscus maniculatus, from Arizona. Occasional Papers of the Museum of Zoology, University of Michigan 375: 1–19.
- 22. Smith AD, Kim YI, Refsum H (2008) Is folic acid good for everyone? The American journal of clinical nutrition 87: 517–533.
- 23. Ly A, Lee H, Chen J, Sie KK, Renlund R, et al. (2011) Effect of maternal and postweaning folic Acid supplementation on mammary tumor risk in the offspring. Cancer Res 71: 988–997.
- 24. Marean A, Graf A, Zhang Y, Niswander L (2011) Folic acid supplementation can adversely affect murine neural tube closure and embryonic survival. Hum Mol Genet 20: 3678–3683.
- 25. Hoyo C, Murtha AP, Schildkraut JM, Jirtle RL, Demark-Wahnefried W, et al. (2011) Methylation variation at IGF2 differentially methylated regions and maternal folic acid use before and during pregnancy. Epigenetics 6: 928–936.
- 26. Girotto F, Scott L, Avchalumov Y, Harris J, Iannattone S, et al. (2013) High dose folic acid supplementation of rats alters synaptic transmission and seizure susceptibility in offspring. Sci Rep 3: 1465.
- 27. Junaid MA, Kuizon S, Cardona J, Azher T, Murakami N, et al. (2011) Folic acid supplementation dysregulates gene expression in lymphoblastoid cells–implications in nutrition. Biochem Biophys Res Commun 412: 688–692.
- 28. Vasquez K, Kuizon S, Junaid M, Idrissi AE (2013) The effect of folic acid on GABA(A)-B 1 receptor subunit. Adv Exp Med Biol 775: 101–109.
- 29. Schaible TD, Harris RA, Dowd SE, Smith CW, Kellermayer R (2011) Maternal methyl-donor supplementation induces prolonged murine offspring colitis susceptibility in association with mucosal epigenetic and microbiomic changes. Hum Mol Genet 20: 1687–1696.
- 30. Hollingsworth JW, Maruoka S, Boon K, Garantziotis S, Li Z, et al. (2008) In utero supplementation with methyl donors enhances allergic airway disease in mice. J Clin Invest 118: 3462–3469.
- 31. McGowan PO, Meaney MJ, Szyf M (2008) Diet and the epigenetic (re)programming of phenotypic differences in behavior. Brain Res 1237: 12–24.