- Email: [email protected]
Behavioral and neurochemical changes in folate-deficient mice
Physiology & Behavior, Vol. 58, No. 5, pp. 935-941, 1995 Copyright © 1995 Elsevier Science Inc. Printed in the USA. All rights reserved 0031-9384/95 $9.50 + .00
Pergamon 0031-9384(95)00156-5
Behavioral and Neurochemical Changes in Folate-Deficient Mice S I D N E Y M. GOSPE, JR.,*t 1 D O R O T H Y W. GIETZEN,:[: PHILIP J. S U M M E R S , § JENNINE M. L U N E T T A , ¶ J O S H U A W. IV[ILLER, I [ J A C O B SELHUB, I t W I L L I A M G. ELLIS,** A N D A N D R E W J. C L I F F O R D ¶
*Department of Neurology, University of California, Davis, Davis, CA 95616, tDepartment of Pediatrics, University of California, Davis, Davis, CA 95616, ~Department of Anatomy, Physiology & Cell Biology, University of California, Davis, Davis, CA 9561,6, §Department of Animal Science, University of California, Davis, Davis, CA 95616, ¶Department of Nutrition, University of California, Davis, Davis, CA 95616, ][ USDA Human Nutrition Research Center on Aging, Tufts University, Boston, MA 02111, and **Department of Pathology, University of California, Davis, Davis, CA 95616 Received 3 February 1995 GOSPE, JR., S. M., D. W GIETZEN, P. J. SUMMERS, J. M. LUNETI'A, J. W. MILLER, J. SELHUB, W. G. ELLIS AND A. J. CLIFFORD. Behavioral and neurochemical changes in folate-deficient mice. PHYSIOL BEHAV 58(5) 935-941, 1995.--Weanling mice were fed an amino acid-based diet supplemented with 0 or 11.3 /.~mol folic acid/kg diet for ~ 38 days to study behavior and neurochemistry in folate deficiency. After ~ 5 wk, mice fed the unsupplemented diet weighed ~ 70% as much those fed the supplemented diet. After 2 wk, mice fed the unsupplemented diet consistently discarded (spilled) more food, and after ~ 5 wk, they had spilled 3 times more than mice fed the supplemented diet. Serum folate, brain folate and brain S-adenosylmethionine of mice fed the unsupplemented diet were 4, 53, and 60% as high, respectively, as those of mice fed the supplemented diet. Pathologic changes were not evident in brain, spinal cord, or skeletal muscle of folate-deficient mice. The hypothalamic 5-hydroxyindole acetic acid/serotonin ratio and caudate dopamine, homovanillic acid, and 3,4-dihydroxyphenylacetic acid concentrations were lower in deficient than control mice. Folate-deficient mice develop a behavioral activity, food spilling, which may have a neurochemical basis in the serotonin and dopamine systems. Folate deficiency
Food spilling behavior
Serotonin
FOLIC acid is a water-soluble micronutrient that is essential for growth, reproduction, and maintenance of normal body function. It is involved in one-carbon metabolism for synthesis of nucleic acids and certain amino acids. Folate deficiency is thought to be associated with several neurologic and psychiatric disorders including dementia (23,25,31), subacute combined degeneration of the spinal cord (SACD) (9,23,25), peripheral neuropathy (9,25), restless leg syndrome (9), and depression (1,9,34). The biochemical mechanisms underlying the neurobehavioral dysfunction associated with folate deficien~cy are not well understood. Depression of serotonin levels, has been noted in folate-deficient rats (8), and low levels of serotonin metabolites were found in the cerebrospinal fluid (CSF) of patients with folate responsive neuropsychiatric disorders (12). The concentration of S-adenosylmethionine (SAM) (24) is also reduced in brain from folate-deficient rats. The concentratJion of SAM is also reduced in the CSF of children with 5,10-melhylenetetrahydrofolate reductase (5,10CH2-THFR) deficiency (22), an inherited disorder characterized
Dopamine
Mice
by demyelination of the brain and SACD. Therefore, it has been suggested that depletion of these chemicals in brain and CSF may contribute to the depression (35), demyelination and SACD (22) that seems to occur in folate deficiency. Recent studies in our laboratory have demonstrated that brain folate concentrations can be markedly reduced depending on the folic acid content of the diet (6,7,21). During the course of these studies, it was noted that the folate concentration in brain from deficient mice was about half that in brain from control mice. It was also noticed that the deficient mice consistently discarded (spilled) much more food than did nondeficient control mice (unpublished observations). Food spilling activity in the folatedeficient mice was so intense that large quantities of food had to be added to their food cups each day to assure that food would be available at all times. This food spilling behavior suggested a neurochemical disturbance as an accompaniment of severe folate deficiency in mice. This report describes the food spilling behavior and neu-
1 Requests for reprints should be addressed to Sidney M. Gospe, Jr., M.D, Ph.D., Child Neurology, University of California, Davis Medical Center, 2315 Stockton Blvd., Sacran:tento,CA 95817. E-Mail: [email protected].
935
936
GOSPE, JR. ET AL.
ropathologic and neurochemical changes associated with folate deficiency in mice. The results from three experiments are presented. The first describes the effect of folate deficiency on the histopathology of brain and skeletal muscle. The second compares the growth, food spilling behavior and the concentrations of total folate, cysteine and homocysteine in serum, and of neurotransmitters in the hypothalamus and caudate nucleus from folate-deficient and control mice. The third experiment compares the growth, food spilling behavior and the concentrations of folate, SAM and S-adenosylhomocysteine (SAH) in whole brain and hematologic characteristics of folate-deficient and control mice. MATERIALS AND METHODS
Weanling Swiss Webster female mice weighing 12-14 g were obtained from Simonsen Laboratories, Inc., Gilroy, CA. Mice were fed a commercially available cereal-based diet overnight (Purina Mouse Chow, Ralston Purina, St. Louis, MO). Mice were individually housed in stainless steel wire-bottomed cages in a room with a 12-h light/dark cycle (lights on 07:00-19:00 h), 20-23°C, and 50% humidity. The experimental protocol was approved by the Animal Use and Care Administrative Advisory Committee of the University of California, Davis. On the following day, mice were systematically divided into two groups of equal mean body weights and randomly assigned to an amino acid-based diet (5) supplemented with either 0/zmol or 11.3 /zmol folic acid per kg diet for 38 days. During this time, all mice had free access to food and water, and were weighed daily with a few exceptions. The amount of food that was given, spilled by the mice, and left in the feed cup was measured almost every day during the 37 and 39 day periods in experiments 2 and 3, respectively. Measuring of body weight, food given, food spilled and food left in the feed cup was always performed between 14:00 and 16:00 h. The measurement of spilled food was accomplished by the following method: An 8.5 by 11 inch sheet of paper was placed beneath each cage (10 X 6 inches) where spilled food, feces, hair and dander accumulated. Beneath the paper was a layer of wood shavings. At the time the animals were weighed, the paper was replaced, the feces were removed from the paper and weighed, and the hair and dander were removed and discarded. Urine was absorbed into the paper and the majority of it evaporated. Spilled food was then removed as carefully as possible, weighed and placed into a collection jar. Food eaten during experiments 2 and 3 was calculated by subtracting the amount of food spilled plus the food remaining in the cup from that given. The feeding patterns during the dark cycle were not monitored. At the end of the test periods, folate-deficient mice, on average, weighed ~ 70% as much as the nondeficient controis. At the end of the feeding periods, the mice were killed by overdosing with diethyl ether, and bled by cardiac puncture between 1000 and 1200 h. The mice were killed in pairs, a folic acid-deficient mouse followed by a control, to avoid diurnal change as a confounding variable. All stomachs contained food and the amounts did not appear to be different (subjective assessment) between the 2 groups of mice. Collected blood was transferred to tubes without anticoagulant and allowed to clot for 15 min at 23°C. The serum was isolated by centrifugation, transferred to plastic vials and stored at -20°C. Serum from experiment 2 was analyzed for total-folate by the 96 wells microtiter plate assay using Lactobacillus casei (33) and cysteine and homocysteine by HPLC (32). In experiment 1, brains were removed, weighed and fixed in formalin for pathologic evaluation. The vertebral columns were
also removed, fixed in formalin, decalcified and the spinal cord was cut into small pieces prior to further processing. The fixed brain and spinal cord pieces were dehydrated and embedded in paraffin and histologic sections were cut and stained with either hematoxylin/eosin or cresyl violet/luxol fast blue. Also in experiment i, gastrocnemius and soleus muscles from 2 depleted and 1 control mouse were removed and frozen in a dry ice-acetone bath. Cross sections, 16 /xm thick, were cut, fixed and stained with hematoxylin/eosin. Muscle fiber cross sections were traced onto acetate sheets using a Reichert microscope (Model Visopan, Leica, Deerfield, IL). Serial sections were stained for succinic dehydrogenase (28), and the percentage of fibers high in oxidative metabolism, staining dark with this technique, was determined. Muscle fiber cross-sectional areas were measured from the traces using Sigma-Scan software (Jandel Scientific, San Rafael, CA). In experiment 2, brains were also removed, but these were frozen in liquid nitrogen and weighed. The caudate nuclei and hypothalamus were dissected from frozen coronal sections. The tissues were deproteinized with perchloric acid, centrifuged, and the supernatants were filtered (0.45 /zm pore size). The samples were then analyzed for neurotransmitters and their respective metabolites using HPLC with electrochemical detection in a modification of a previously described method (19). The HPLC column was reversed-phase 3 /xm ODS, 250 mm x 4.6 mm (Bioanalytical Systems, West Lafayette, IN). The mobile phase contained 109 mmol/L citric acid, 8.8 mmol/L NaOH, 1.3 mmol/L octane sulfonic acid, 0.94 mmol/L EDTA and 1.3 mmol/L acetonitrile at pH 3.15. Electrochemical detection was at 850 inV. Since in our experience, filtrates that are injected show an internal standard recovery ratio of 100 + 2%, the internal standard is no longer included in these assays. External standards included the transmitters norepinephrine (NE), dopamine (DA), and serotonin (5HT), along with their respective metabolites as follows: for NE: 3-methoxy-4-hydroxy-mandelic acid (VMA) and 3-methoxy-4-hydroxy-phenylglycol (MHPG); for DA: homovanillic acid (HVA) and 3,4-dihydroxyphenylacetic acid (DOPAC); for 5-HT: 5-hydroxyindole acetic acid (5HIAA). Also included in the external standards was 3-hydroxykynurenin, because this tryptophan metabolite can co-elute with NE, and is elevated in vitamin B-6-deficient animals (20). In experiment 3, brains were also removed, frozen in liquid nitrogen and weighed. These whole brains were analyzed for total-folate (33) and for SAH and SAM using HPLC with UV detection (16) with minor modifications. For folate analysis, aliquots ( ~ 100 rag) of frozen brain were homogenized in 9 volumes of a 50 mM phosphate buffer (pH 6.1) containing 0.2% ascorbate, autoclaved for 10 min at 121°C, and cooled in an ice water bath. Cooled homogenates were centrifuged for 15 min at 2,000 g, and 150 /~L aliquots of the clear supematant extracts were transferred to tubes containing 2.79 ml of the 50 mM phosphate buffer (pH 6.1) containing 0.2% ascorbate plus 60 /xL chicken pancreas conjugase and mixed. The mixture was incubated at 37°C for 6 h to allow folacin polyglutamates to be converted to diglutamates. Aliquots of the conjugase treated extracts containing 0.1 to 2.0 ng folate were then analyzed for total folate (33). For SAH and SAM determinations, the frozen brain samples were homogenized in 5 volumes cold 0.4 M perchloric acid using a Polytron homogenizer (20 s., setting 7) (Brinkman Instruments, Westbury, NY) and centrifuged at 1500 × g for 10 min. A 1.0 ml aliquot of the superuatant was filtered through a 0.45 p,m Gelman syringe filter (VWR Scientific, San Francisco, CA), and 100 /xL of the filtered supernatant from each sample was analyzed by HPLC. The column was 3 /xm ODS Hypersil, 150 mm x 2 mm
NEUROCHEMISTRY, BEHAVIOR AND FOLIC ACID
(Keystone Scientific, Inc, Bellafonte, PA). The mobile phase was a linear gradient over 20 min for 100% solvent A (0.01 m o l / L ammonium formate, 4 m m o l / L heptanesulfonic acid, pH 4.0) to 75% solvent A and 25% solvent B (50% mobile phase A + 50% acetonitrile, pH 4.0). The 75% A / 2 5 % B ratio was maintained from 20 to 25 min, and then the system was returned to 100% solvent A for a 15 min equilibration. The flow rate was
30
2S,/
i
20,
m
15.
11:3HmolFolicAcid/kDiet g 0~la'nolFolioAci ~ d/kDi g et
,o
,o
,o
3o
,o
ZOO-"mOBtl-.~molFolicA~gOiet -"~ 1501'~1"1~0I~m°lF°hcAcid/kD' gt TS
937
0.3 m l / m i n . SAM and SAIl peaks were detected by UV absorption at 254 rim. SAM and SAH concentrations were determined using SAM and SAH external standards. Statistical analysis of data was performed with StatView software (Abacus Concepts, Berkeley, CA) for the Macintosh Computer. Since there were just two groups, Student's unpaired 2-tailed t-tests were used to determine significance of the effects of the folate-deficient diet on weight, food spill, food consumption, and levels of serum and brain chemicals. In experiment 2, one mouse from each dietary group failed to demonstrate spilling behavior consistent with their cohort. These animals were removed from the data set, leaving 4 mice per group in experiment 2. These two mice were statistical outliers because the average amount of food they spilled differed by - 3 . 2 6 (folate-deficient mouse) and - 2 . 2 2 (control mouse) standard deviations from the mean values of the 4 remaining mice in their respective groups. There were no mice in experiment 3 (n = 7/group), with such inconsistent spilling behavior. Some of the analyte data are expressed as concentration ratios because simultaneous observation of analyte concentration ratios, in addition to the concentrations, often provide insight into the biological implications of the data as has been illustrated for dopamine (3). Statistical significance was defined by a p value < 0.05. The values presented throughout this manuscript are means ___SEM. RESULTS
Systemic Effects of Folate Depletion
100
-
T /
50o zso T ~
,o
,;
~
3o
~
j ----o,---11.3lamolFol~Ac~kgDiet I ---o--- 0p.rnolFoiicAcid/kgDiet
1001
O
TT
,TI]J~'
~,i~-
I'3
2; Days Fed
3;
40
TestDiets
FIG. 1. Growth, cumulative food spilled, and cumulative food eaten by mice fed amino acid-based diets containing 0 or 11.3/zmol folic acid/kg. Values are mean + SEM of 4 mice fed the amino acid-based diet with and without folic acid, respectively. Body weight of mice fed the diet without folic acid were smaller after 3.5 wk. Cumulative diet spilled by mice fed the diet without folic acid was greater at 2.5 wk. Cumulative diet eaten was greater for mice fed the .diet without added folic acid throughout this experiment. The patterns of growth, and cumulative food spilled and eaten, as shown in this figure, were consistent for all three experiments.
After approximately 3.5 wk on the folic acid-deficient diet, growth rate declined and eventually these mice lost weight (Fig. 1). In experiment 2, at the end of the 37 day feeding period, folate-deficient animals weighed ~ 70% of the weight of the mice who were fed the control diet ( p = 0.01). This was a consistent finding among the three studies.Animals fed the folic acid-deficient diet consistently spilled their food (Table 1), and differences between the two groups in amount of food spilled first appeared after ~ 2.5 wk (Fig. 1). In experiment 2, the cumulative food spilled over the 37 day experiment in the folate-deficient group was 166 + 21 g vs. 5 2 + 11 g in the control group ( p = 0.0015). Even though the folate-deficient animals spilled more food, and gained less weight, they actually consumed more food during the course of the experiment. In experiment 2, the folate-deficient group consumed 219 + 21 g while the control group consumed 174 ___4 g ( p = 0.039). Similar observations were recorded in the animals from experiment 3 (Table 1). Folate deficiency was documented in experiment 2 by measuring the concentrations of folate, cysteine and homocysteine in serum (Table 1). Serum folate concentrations in the animals fed the folic acid-deficient diet were reduced to < 5% of those fed the control diet ( p = 0.0001). Folate depletion was also associated with a significant reduction of serum cysteine, elevation of serum homocysteine and deterioration of hematologic status (Table 1).
Neuropathological Studies Histologic examination of brain and spinal cord sections revealed no specific changes in the tissue from the folate-deficient animals. In particular, the cerebral cortex, caudate nucleus, corpus callosum, cerebellum, and white matter tracts showed no evidence of necrosis, atrophy or demyelination. Spinal cord sections from cervical, thoracic, lumbar and sacral levels were
938
GOSPE, JR. ETAL.
B. Soleus muscle fiber area distribution (~.m2)
A. Gastrocnemius muscle fiber area distribution (pm 2)
i "--
-coo,+
[] Folatedeficient
20
o ' I .n . . . . . . . 400
800
1200
111] 1600
2000
•
2400
•
1
I
2800
3200
Muscle Fiber Area (~m2) C. Control
i
~
•
.++
[] Folatedeficient
.~ 20 ~.
0
.
.
.
400
.
.
800
.
In .
.
1200
.
.
ii,,,
1600
.
.
2000
.
.
2400
.
.
2800
3200
Muscle Fiber Area (l~mz) D, Folate deficient
FIG. 2. Frequency distribution histogram of fiber cross sectional area in gastrocnemius (panel A) and soleus (panel B) muscles of mice fed an amino acid based diet supplemented with 0 or 11.3/zmole relic acid per kg. Distributions are based on data from ~ 150 fibers• Folate depletion induced a reduction in the number of larger fibers in both muscles• Representative cross sections of succinic dehydrogenase-stained gastrocnemius muscle from control and folate-deficient mice are shown in panels C and D, respectively• Magnification = 700.
examined and were without neuropathologic changes in the anterior horn, posterior columns or spinal roots. Brain weight of folate-deficient m i c e w a s less ( 4 2 7 - I - 6 vs. 480--I-6 m g , p = 0.006) c o m p a r e d to control m i c e . H o w e v e r , the b r a i n w e i g h t / b o d y w e i g h t ratio did not differ b e t w e e n the two groups• Since n e u r o m u s c u l a r d i s e a s e induced by f e e d i n g a folic aciddeficient diet for 3 7 - 3 9 d a y s could c a u s e the increased food spilling, the histology o f skeletal m u s c l e w a s e x a m i n e d . Soleus and g a s t r o c n e m i u s m u s c l e s h a d a n o r m a l distribution o f fibers h i g h in oxidative m e t a b o l i s m and s h o w e d no e v i d e n c e o f m y o p a thy or n e u r o g e n i c atrophy (data not shown). H o w e v e r , g a s t r o c n e m i u s and soleus m u s c l e s f r o m folate-deficient m i c e s h o w e d a reduction in the n u m b e r o f larger fibers c o m p a r e d to the s a m e m u s c l e s f r o m control mice (Fig. 2). T h e s e c h a n g e s in m u s c l e fiber size are not likely to be a m a j o r c a u s e o f the behavioral c h a n g e s associated w i t h relate depletion.
Biogenic Amines Dietary folic acid deficiency had significant effects on the concentrations o f b i o g e n i c a m i n e s a n d their m e t a b o l i t e s in brain (Table 2). T h e alterations in the h y p o t h a l a m i c region o f the brain were different f r o m t h o s e o b s e r v e d in the caudate nucleus. In the h y p o t h a l a m u s , folate-deficient a n i m a l s tended to h a v e m o r e 5 - H T ( p = 0.063) a n d less 5 - H I A A ( p = 0.086) e v e n t h o u g h neither trend w a s statistically significant (Table 2). T h e 5 - H I A A / 5 H T ratio w a s m u c h lower in relate-deficient m i c e c o m p a r e d to
TABLE 1 FINAL WEIGHT, FOOD SPILLED AND EATEN, KEY METABOLITES AND HEMATOLOGY IN MICE RECEIVING CONTROL AND FOLATE-DEF1CIENT DIETS* Dietary Folic Acid ( / z m o l / k g )
Experiment 2 Final bodyweight(g) Food spilled (g/37d) Food eaten (g/37d) Cysteine (/xmol/L serum) Homocysteine (/xmol/L serum) Folate(nmol/Lserum) Experiment 3 Final body weight (g) Food spilled (g/39d) Food eaten (g/39d) Brain weight (rag) Erythrocytes (10 12/ L ) Reticulocytes (109/L) Packed cell volume (%) Mean corpuscular volume (fL) Folate (nmol/kg brain) SAM (/xmol/kg brain) SAH (/xmol/kg brain) SAM/SAH ratio in brain
0.0
19.3 166 219 83 125 5.1
± ± ± ± ± ±
20.7 + 191 + 238 + 427 ± 2.09 ± 2.22 ± 22.6 ± 116 ± 61 ± 4.85 ± 4.00 ± 1.40 +
I 1.3
2.7t 21# 21t 12t 36t 0.9t
27.8 52 174 212 9 124
-t-_ 0.8 ± 11 ± 4 ± 14 ± 1 ± 6
0.5t 43t 16t 6t 0.29t 0.62 2.3t 16t 10t 0.65t 0.70 0.39t
28.2 30 202 480 5.74 3.04 50.0 87 116 8.13 3.43 2.47
± 1.0 ± 18 + 11 ± 6 ± 0.11 + 0.49 -I- 1.2 ± 3 ± 12 ± 1.19 ± 0.64 ± 0.34
* Values are means ± SEM. Experiment 2 lasted 37 days and results a r e means of 4 mice/group; data for one deficient and one control mouse were omitted because they spilled very little food (see text). Experiment 3 lasted 39 days and involved 7 deficient and 7 control mice. t Values that are significantly affected by relate deficiency at p < 0.05 based on unpaired, 2-tailed Student's t-tests.
NEUROCHEMISTRY, BEHAVIOR AND FOLIC ACID
TABLE 2 CONCENTRATIONS O!F BIOGENIC AMINES AND METABOLITES IN HYPOTHALAMUS AND CAUDATE OF CONTROL AND FOLATE-DEFICIENT MICE IN THE SECOND EXPERIMENT*,t p,Mol Analyte/kg wet weight of tissue Dietary Folic Acid (p, mol/kg) Hypothalamus L-DOPA DOPAC DA 5HIAA HVA 5HT DOPAC/DA HVA/DA DOPAC + H V A / D A 5HIAA/5HT Caudate L-DOPA DOPAC DA 5HIAA HVA 5HT DOPAC/DA HVA/DA DOPAC + H V A / D A 5HIAA/5HT
0.0
11.3
0.37 1.03 5.71 4.70 1.14 3.72 0.18 0.20 0.38 1.33
+ 0.02 + 0.15 + 0.32 + 0.77 + 0.33 + 0.67 _+ 0.01 + 0.06 5:0.06 + 0.22:~
0.31 1.04 6.63 7.20 1.79 1.71 0.17 0.27 0.45 4.38
0.13 1.38 26.5 2.41 2.15 2.28 0.05 0.09 0.15 1.14
5:0.01 + 0.39:~ + 5.6:~ + 0.34 5: 0.11:~ + 0.41 + 0.02 -t- 0.02 + 0.03 _+ 0.22
0.15 2.64 41.3 2.73 3.74 1.78 0.06 0.09 0.15 2.25
+ 0.02 + 0.38 + 0.69 + 0.82 + 0.28 + 0.36 5:0.08 + 0.04 + 0.11 + 0.43 + + + + + + + + + +
0.02 0.21 1.3 0.65 0.13 0.58 0.01 0.01 0.01 1.17
* Values are means + SEM of 4 folate-deficient and 3 control mice for the hypothalamus data and of 4 folate-deficient and 4 control mice for the caudate data. ~" Norepinephrine (NE) and metabolites, VMA and MHPG, were also measured but differences in concentrations of these metabolites between the folate-deficient and control groups were not significant, therefore only DA- and 5-HT-related compounds are shown. :~ Values that are significantly smaller in the folate-deficient group at p < 0.05 based on unpaired, 2-tailed Student's t-tests.
control mice ( p = 0.015). Differences in hypothalamic concentrations of norepinephrine, dopamine, or their metabolites between deficient and control mice were not significant. Caudate dopamine levels were reduced by 36% ( p = 0.04), DOPAC levels by 48% ( p = 0.03), and HVA levels by 43% ( p = 0.0002) in folate-deficient mice compared to control mice (Table 2). Brain Folate, SAM and SAH Animals fed the folic acid-deficient diet also had significantly lower whole brain folate and SAM (Table 1). Whole brain folate was 47% lower ( p = 0.01) and SAM was 40% lower ( p = 0.035). Whole brain SAH was not affected by folate depletion, however the SAM/SAH ratio was 43% lower ( p = 0.047) in folate-deficient mice. DISCUSSION
These studies have shown that feeding mice a folic acid-deficient diet for 37 - 39 days results in a reproducible decrease in weight gain followed by weight loss. In addition, even though the folate depleted mice spilled large amounts of their diet, they also consumed more. This combination of weight loss and augmented food consumption sugges~ts that folate depletion may reduce the efficiency by which animals utilize dietary nutrients to meet physiological functions, suggesting that an alteration in hypothalamic and/or neuroendocfine function may underlie this effect of folate depletion. In addition, these changes may be due to a folate deficiency-induced reductiLon in nutrient absorption. These studies were designed to characterize the food spilling behavior of folate-deficient mice and to determine if it might have a neurochemical basis. Several precautions were taken to
939
minimize the effects of diurnal variation on the results of this study. These included housing the mice in a room with a 12 h light dark cycle, feeding and weighing the mice at the same time each day and alternately killing control and deficient mice within a 2 h window at the end of each experiment. Alterations of serotonergic and dopaminergic systems by folate-depletion are suggested by these experiments. A significant reduction in the hypothalamic 5-HIAA/5HT ratio was obtained, while the folate-deficiency induced 35% reduction of hypothalamic 5-HIAA was not significant ( p = 0.086). It is possible that these changes were due to an inhibition of 5-HT metabolism. A reduction in whole brain 5-HT in folate-deficient rats has been shown in one previous study (11), whereas a more recent study did not show a change in either whole brain 5-HT or 5-HIAA in response to 10 wk of folate depletion (14). In that study, there was a small increase in 5-HIAA in the CSF of folate depleted rats. However, the concentrations of folate in serum and brain of those depleted rats were much higher than those observed in the folate-deficient mice in the present study. Therefore, the rats might not have been adequately depleted to alter biogenic amine metabolism. Clinical studies have also shown that folate deficiency (10-12) and 5,10-CHz-THFR deficiency (22,29) are associated with a reduction in 5-HIAA in the CSF. These results suggest that folate depletion does affect the hypothalamic serotonergic system. Examination of the effects of folate depletion on the serotonergic neurochemistry of the raphe nuclei, which contain the cell bodies that supply serotonin to the forebrain, may provide more specific information about these alterations. This study has also demonstrated an alteration in dopaminergic neurochemistry due to folate depletion. In the caudate, significant reductions in DA, DOPAC and HVA were noted, without a change in the level of L-DOPA. Although this might suggest that folate deficiency may reduce DA synthesis at the level of DOPA-decarboxylase, this is unlikely to be a rate limiting step in vivo because of the large quantities of DOPA-decarboxylase present. Alternatively, if low metabolite concentrations can be used to suggest a reduction in dopaminergic activity, folate depletion may have reduced nigrostriatal activity as measured neurochemically in the caudate nucleus. One previous animal study failed to show a change in whole brain levels of DA and HVA of rats in response to folate depletion (14). In that study, the use of whole rat brain may have obscured the changes in dopaminergic neurochemistry that were noted in caudate but not hypothalamus of the mouse. Because the deficient rat serum and brain folate concentrations were much higher than those observed in our deficient mice, those folate-deficient rats may not have been depleted of folate as severely as were the folate-deficient mice in the present study. Clinical studies have also shown reductions of HVA in the CSF of folate-deficient patients (10) and of HVA, as well as 5-HIAA and MHPG with 5,10-CHz-THFR deficiency (22,29). The clinical observations together with the data presented in this report suggest that folate deficiency alters monoaminergic function. This study also demonstrated a reduction in whole brain folate and SAM. Reduction of brain folate in folate-depleted mice is consistent with several recent studies (6,7,14,21) which have demonstrated that the concentration of folate in brain can vary markedly depending on the folic acid content of the diets. Given the close relationship between the metabolism of folates and SAM (30), the 40% reduction of whole brain SAM in the folate depleted mice was not unexpected. This study did not reveal any significant pathologic changes in brain, spinal cord, spinal roots or skeletal muscle in response to folate depletion. It was hypothesized that the food spilling behavior represented a symptom of folate-depletion-induced
940
GOSPE, JR. ET AL.
ataxia, weakness, or extrapyramidal dysfunction. While no microscopic changes were detected, the neurochemical changes do suggest some possible mechanisms which may underlie this food spilling behavior. For example food spilling due to folate-depletion may be a reflection of altered basal ganglia dopaminergic motor function. Active avoidance learning, but not motor function, was impaired in folate-deficient rats (2). In the present study, the mice appeared (subjectively) to have adequate motor function, but if spilling is associated with conditioned taste aversion (26), then this form of learning was apparently intact in these animals. Alternatively, this excessive spilling behavior may in some way represent a form of obsessive compulsive behavior associated with altered serotonergic function, as has been seen in other animal models (18). Reductions in CSF SAM levels have been noted in depressed patients (13), and SAM has been shown to have anti-depressant properties (17). Increased food spilling has been noted in other nutrient deficiencies (19,26,27) but the neurochemical basis for the behavior has not been studied. In a behavioral study of rats becoming thiamin deficient, the animals showed two responses that are not often seen in normal rats. Some rats vigorously spilled their food out of the food cup; they could empty 50 g of food in 15 min. Some deficient rats also demonstrated "redirected feeding" by chewing on the cage or other nonfood objects. These responses were interpreted as aversive behaviors directed toward the deficient diet as might be seen with poisoning (26). The poison-
avoiding responses of the rat have been reviewed (4), and may also relate to the conditioned taste aversion or " b a i t shyness" (15). Whatever the psychological mechanisms, some neurochemical alterations may be associated with the dramatic increases in food spillage that was observed. The present findings of alterations in both the serotonin and dopamine systems with folate deficiency suggest that the food spilling may have a neurochemical component. In addition, the lower feed efficiency of folatedeficient mice, as evidenced by greater feed intake but smaller weight gain, may imply neuroendocrine dysfunction, which could also have dopaminergic or serotonergic components. Clearly, more research on neurochemical mechanisms and neuroendocrine changes is needed. While these data suggest that food spilling behavior is associated with neurochemical changes, it is not clear which may come first. A separate longitudinal experiment with many animals and time points is needed to adequately address this important point. ACKNOWLEDGEMENTS This work was supported by grants DK-38637 from U.S. Public Health Service, regional research W-143 from U.S.D.A., and Hatch 2850 from the California Experiment Station. The authors thank Miriam Watson for assistance in measuring growth and the intake and spillage of food in the second experiment, Shan Shan Zhou for assistance with the histology, and Keith Hyland, Baylor Research Institute, Dallas TX, for thoughtfully reviewing this manuscript.
REFERENCES
1. Abou-Saleh, M. T.; Coppen, A. The biology of folate in depression: Implications for nutritional hypotheses of the psychoses. J. Psychiatr. Res. 20:91-101;1986. 2. Bachevalier, J.; Botez, M. I. Avoidance behavior in folate-deficient rats. Tohoku J. Exp. Med. 126:111-116;1978. 3. Bannon M. J.; Roth, R. H. Pharmacology of mesocortical dopamine neurons. Pharmacol. Rev. 35:53-68;1983. 4. Barnett, S. A. The rat. A study in behavior. London: Methuen; 1963. 5. Bills, N. D.; Koury, M. J.; Clifford, A. J.; Dessypris, E. N. Ineffective hematopoiesis in folate-deficient mice. Blood 79:22732280;1992. 6. Bills, N. D.; Hinrichs, S. H.; Morgan, R.; Clifford, A. J. Tissue folate levels in transgenic mice with tumors and in nontransgenic controls. J. Nutr. Biochem. 3:113-117;1992. 7. Bills, N. D.; Hinrichs, S. H.; Morgan, R; Clifford, A. J. Delayed tumor onset in transgenic mice fed a low folate diet. J. Natl. Cancer Inst. 84:332-337;1992. 8. Botez, M. I.; Botez, T.; Leveille, J.; Bielmann, P; Cadotte, M. Neuropsychological correlates of folic acid deficiency: Facts and hypotheses. In: Botez, M. I.; Reynolds, E. H., eds. Folic acid in neurology. Psychiatry, and internal medicine. New York: Raven Press; 1979:435-461. 9. Botez, M. I.; Fontaine, F.; Botez, T.; Bachevalier, J. Folate-responsive neurological and mental disorders:report of 16 cases. Eur. Neurol. 16:230-246;1977. 10. Botez, M. I.; Young, S. N. Effects of anticonvulsant treatment and low levels of folate and thiamine on amine metabolites in cerebrospinal fluid. Brain 114:333-348;1991. 11. Botez, M. I.; Young, S. N.; Bachevalier, J.; Ganthier, S. Folate deficiency and decreased brain 5-hydroxytryptamine synthesis in man and rat. Nature 278:182-183;1979. 12. Botez, M. I.; Young, S. N.; Bachevalier, J.; Gauthier, S. Effect of folic acid and vitamin B12 deficiencies on 5-hydroxyindoleacetic acid in human cerebrospinal fluid. Ann. Neurol. 12:479-484;1982. 13. Bottliglieri, T.; Laundy, M.; Carney, M. W. P.; Toone, B. K.; Reynolds, E. H. Cerebrospinal fluid S-adenosylmethionine in depression and dementia:effects of treatment with parenteral and oral
14. 15. 16. 17.
18. 19. 20.
21. 22.
23. 24.
S-adenosylmethionine. J. Neurol. Neurosurg. Psychiatry 53:10961098;1990. Butler, I. J.; Rothenberg, S. P. Dietary folate and biogenic amines in the CNS. J. Neurochem. 49:268-271;1987. Chambers, K. C. A neural model for conditioned taste aversions. Ann. Rev. Neurosci. 13:373-385;1990. Fell D.; Benjamin L. E.; Steele R. D. Determination of adenosine and S-adenosyl derivatives of sulfur amino acid in rat liver by high-performance liquid chromatography. J. Chromatogr. 345:150-156;1985. Friedel, H. A.; Goa, K. L.; Benfeld, P. S-adenosyl-L-methionine: A review of its therapeutic potential in liver dysfunction and affective disorders in relation to its physiological role in cell metabolism. Drugs 38:389-416;1989. Gietzen, D. W. Animal models elucidate the role of serotonin in OCD. Psychiat. Times Med. Behav. 9:17-20;1990. Gietzen, D. W.; Leung, P. M. B.; Rogers, Q. R. Norepinephrine and amino acids in prepyriform cortex of rats fed imbalanced amino acid diets. Physiol. Behav. 36:1071-1080;1986. Guilarte, T. R.; Block, L. D.; Wagner H. D. The putative endogenous convulsant 3-hydroxykyneurine decreases benzodiazepine receptor binding affinity:implications to seizures associated with neonatal vitamin B-6 deficiency. Pharmacot. Biochem. Behav. 30:665668;1988. Heid, M. K.; Bills, N.D.; Hinrichs, S. H.; Clifford, A. J. Folate deficiency alone appears insufficient to produce neural tube defects in mice. J. Nutr. 122:888-894;1992. Hyland, K.; Smith, I.; Bottiglieri, T.; Perry, J.; Wendel, U.; Clayton, P. T.; Leonard, J. V. Demyelination and decreased S-adenosylmethionine in 5,10-methylenetetrahydrofolate reductase deficiency. Neurology 38:459-462;1986. Melamed, R.; Reches, A.; Hershko, C. Reversible central nervous system dysfunction in folate deficiency. J. Neurol. Sci. 25:9398;1975. Ordonez, L. A.; Wurtman, R. J. Folic acid deficiency and methyl group metabolism in rat brain:effects of L-dopa. Reversal effect of S-adenosyl-L-methionine. Arch. Biochem. Biophys. 160:372376;1974.
NEUROCHEMISTRY, BEHAVIOR AND FOLIC ACID
25. Pincus, J. H.; Reynolds, E. H.; Giaser, G. H. Subacute combined degeneration with folate deficiency. JAMA J. Am. Med. Assoc. 221:496-497;1972. 26. Rozin, P. Specific aversions as a component of specific hungers. J. Comp. Physiol. Psych. 64:237-242;1967. 27. Rozin, P.; Kalat, J. W. Specific hungers and poison avoidance as adaptive specializations of learning. Psychol. Rev. 78:459-486;1971. 28. Sarnat, H. B. Muscle pathology and histochemistry. Chicago: American Society of Clinical Pathology Press; 1983. 29. Singer, H. S.; Butler, I; Rothenberg, S.; Valle, D.; Freeman, J. Interrelationships among serum folate, CSF folate, neurotransmitters, and neuropsychiatric symptoms. Neurology 30:419;1990. 30. Stokstad, E. L. R. Historical perspective on key advances in the biochemistry and physiology of folates. In: Picciano, M. F.; Stokstad, E. L. R.; Gregory, III, J. F., eds. Folic acid metabolism in health and disease, vol. 13, New York: Wiley and Sons, Inc.; 1990:1-21.
941
31. Strachan, R. W.; Henderson J. G. Dementia and folate deficiency. Quart. J. Med. 36:189-204;1967. 32. Vester B.; Rasmussen K. High performance liquid chromatography method for rapid and accurate determination of homocysteine in plasma and serum. Eur. J. Clin. Chem. Clin. Biochem. 29:549554;1991. 33. Wilson, S. D.; Home D. W. Use of glycerol-protected Lactobacillus casei for microbiological assay of folic acid. Clin. Chem. 28:11981200;1982. 34. Young, S. N.; Ghadirian, A. M. Folic acid and psychopathology. Prog. Neuro-psychopharmacol. Biol. Psychiatr. 13:841-863;1989. 35. Young, S. N. Some effects of dietary components (amino acids, carbohydrate, folic acid) on brain serotonin synthesis, mood, and behavior. Can. J. Physiol. Pharmacol. 69:893-903;1991.
Pergamon 0031-9384(95)00156-5
Behavioral and Neurochemical Changes in Folate-Deficient Mice S I D N E Y M. GOSPE, JR.,*t 1 D O R O T H Y W. GIETZEN,:[: PHILIP J. S U M M E R S , § JENNINE M. L U N E T T A , ¶ J O S H U A W. IV[ILLER, I [ J A C O B SELHUB, I t W I L L I A M G. ELLIS,** A N D A N D R E W J. C L I F F O R D ¶
*Department of Neurology, University of California, Davis, Davis, CA 95616, tDepartment of Pediatrics, University of California, Davis, Davis, CA 95616, ~Department of Anatomy, Physiology & Cell Biology, University of California, Davis, Davis, CA 9561,6, §Department of Animal Science, University of California, Davis, Davis, CA 95616, ¶Department of Nutrition, University of California, Davis, Davis, CA 95616, ][ USDA Human Nutrition Research Center on Aging, Tufts University, Boston, MA 02111, and **Department of Pathology, University of California, Davis, Davis, CA 95616 Received 3 February 1995 GOSPE, JR., S. M., D. W GIETZEN, P. J. SUMMERS, J. M. LUNETI'A, J. W. MILLER, J. SELHUB, W. G. ELLIS AND A. J. CLIFFORD. Behavioral and neurochemical changes in folate-deficient mice. PHYSIOL BEHAV 58(5) 935-941, 1995.--Weanling mice were fed an amino acid-based diet supplemented with 0 or 11.3 /.~mol folic acid/kg diet for ~ 38 days to study behavior and neurochemistry in folate deficiency. After ~ 5 wk, mice fed the unsupplemented diet weighed ~ 70% as much those fed the supplemented diet. After 2 wk, mice fed the unsupplemented diet consistently discarded (spilled) more food, and after ~ 5 wk, they had spilled 3 times more than mice fed the supplemented diet. Serum folate, brain folate and brain S-adenosylmethionine of mice fed the unsupplemented diet were 4, 53, and 60% as high, respectively, as those of mice fed the supplemented diet. Pathologic changes were not evident in brain, spinal cord, or skeletal muscle of folate-deficient mice. The hypothalamic 5-hydroxyindole acetic acid/serotonin ratio and caudate dopamine, homovanillic acid, and 3,4-dihydroxyphenylacetic acid concentrations were lower in deficient than control mice. Folate-deficient mice develop a behavioral activity, food spilling, which may have a neurochemical basis in the serotonin and dopamine systems. Folate deficiency
Food spilling behavior
Serotonin
FOLIC acid is a water-soluble micronutrient that is essential for growth, reproduction, and maintenance of normal body function. It is involved in one-carbon metabolism for synthesis of nucleic acids and certain amino acids. Folate deficiency is thought to be associated with several neurologic and psychiatric disorders including dementia (23,25,31), subacute combined degeneration of the spinal cord (SACD) (9,23,25), peripheral neuropathy (9,25), restless leg syndrome (9), and depression (1,9,34). The biochemical mechanisms underlying the neurobehavioral dysfunction associated with folate deficien~cy are not well understood. Depression of serotonin levels, has been noted in folate-deficient rats (8), and low levels of serotonin metabolites were found in the cerebrospinal fluid (CSF) of patients with folate responsive neuropsychiatric disorders (12). The concentration of S-adenosylmethionine (SAM) (24) is also reduced in brain from folate-deficient rats. The concentratJion of SAM is also reduced in the CSF of children with 5,10-melhylenetetrahydrofolate reductase (5,10CH2-THFR) deficiency (22), an inherited disorder characterized
Dopamine
Mice
by demyelination of the brain and SACD. Therefore, it has been suggested that depletion of these chemicals in brain and CSF may contribute to the depression (35), demyelination and SACD (22) that seems to occur in folate deficiency. Recent studies in our laboratory have demonstrated that brain folate concentrations can be markedly reduced depending on the folic acid content of the diet (6,7,21). During the course of these studies, it was noted that the folate concentration in brain from deficient mice was about half that in brain from control mice. It was also noticed that the deficient mice consistently discarded (spilled) much more food than did nondeficient control mice (unpublished observations). Food spilling activity in the folatedeficient mice was so intense that large quantities of food had to be added to their food cups each day to assure that food would be available at all times. This food spilling behavior suggested a neurochemical disturbance as an accompaniment of severe folate deficiency in mice. This report describes the food spilling behavior and neu-
1 Requests for reprints should be addressed to Sidney M. Gospe, Jr., M.D, Ph.D., Child Neurology, University of California, Davis Medical Center, 2315 Stockton Blvd., Sacran:tento,CA 95817. E-Mail: [email protected].
935
936
GOSPE, JR. ET AL.
ropathologic and neurochemical changes associated with folate deficiency in mice. The results from three experiments are presented. The first describes the effect of folate deficiency on the histopathology of brain and skeletal muscle. The second compares the growth, food spilling behavior and the concentrations of total folate, cysteine and homocysteine in serum, and of neurotransmitters in the hypothalamus and caudate nucleus from folate-deficient and control mice. The third experiment compares the growth, food spilling behavior and the concentrations of folate, SAM and S-adenosylhomocysteine (SAH) in whole brain and hematologic characteristics of folate-deficient and control mice. MATERIALS AND METHODS
Weanling Swiss Webster female mice weighing 12-14 g were obtained from Simonsen Laboratories, Inc., Gilroy, CA. Mice were fed a commercially available cereal-based diet overnight (Purina Mouse Chow, Ralston Purina, St. Louis, MO). Mice were individually housed in stainless steel wire-bottomed cages in a room with a 12-h light/dark cycle (lights on 07:00-19:00 h), 20-23°C, and 50% humidity. The experimental protocol was approved by the Animal Use and Care Administrative Advisory Committee of the University of California, Davis. On the following day, mice were systematically divided into two groups of equal mean body weights and randomly assigned to an amino acid-based diet (5) supplemented with either 0/zmol or 11.3 /zmol folic acid per kg diet for 38 days. During this time, all mice had free access to food and water, and were weighed daily with a few exceptions. The amount of food that was given, spilled by the mice, and left in the feed cup was measured almost every day during the 37 and 39 day periods in experiments 2 and 3, respectively. Measuring of body weight, food given, food spilled and food left in the feed cup was always performed between 14:00 and 16:00 h. The measurement of spilled food was accomplished by the following method: An 8.5 by 11 inch sheet of paper was placed beneath each cage (10 X 6 inches) where spilled food, feces, hair and dander accumulated. Beneath the paper was a layer of wood shavings. At the time the animals were weighed, the paper was replaced, the feces were removed from the paper and weighed, and the hair and dander were removed and discarded. Urine was absorbed into the paper and the majority of it evaporated. Spilled food was then removed as carefully as possible, weighed and placed into a collection jar. Food eaten during experiments 2 and 3 was calculated by subtracting the amount of food spilled plus the food remaining in the cup from that given. The feeding patterns during the dark cycle were not monitored. At the end of the test periods, folate-deficient mice, on average, weighed ~ 70% as much as the nondeficient controis. At the end of the feeding periods, the mice were killed by overdosing with diethyl ether, and bled by cardiac puncture between 1000 and 1200 h. The mice were killed in pairs, a folic acid-deficient mouse followed by a control, to avoid diurnal change as a confounding variable. All stomachs contained food and the amounts did not appear to be different (subjective assessment) between the 2 groups of mice. Collected blood was transferred to tubes without anticoagulant and allowed to clot for 15 min at 23°C. The serum was isolated by centrifugation, transferred to plastic vials and stored at -20°C. Serum from experiment 2 was analyzed for total-folate by the 96 wells microtiter plate assay using Lactobacillus casei (33) and cysteine and homocysteine by HPLC (32). In experiment 1, brains were removed, weighed and fixed in formalin for pathologic evaluation. The vertebral columns were
also removed, fixed in formalin, decalcified and the spinal cord was cut into small pieces prior to further processing. The fixed brain and spinal cord pieces were dehydrated and embedded in paraffin and histologic sections were cut and stained with either hematoxylin/eosin or cresyl violet/luxol fast blue. Also in experiment i, gastrocnemius and soleus muscles from 2 depleted and 1 control mouse were removed and frozen in a dry ice-acetone bath. Cross sections, 16 /xm thick, were cut, fixed and stained with hematoxylin/eosin. Muscle fiber cross sections were traced onto acetate sheets using a Reichert microscope (Model Visopan, Leica, Deerfield, IL). Serial sections were stained for succinic dehydrogenase (28), and the percentage of fibers high in oxidative metabolism, staining dark with this technique, was determined. Muscle fiber cross-sectional areas were measured from the traces using Sigma-Scan software (Jandel Scientific, San Rafael, CA). In experiment 2, brains were also removed, but these were frozen in liquid nitrogen and weighed. The caudate nuclei and hypothalamus were dissected from frozen coronal sections. The tissues were deproteinized with perchloric acid, centrifuged, and the supernatants were filtered (0.45 /zm pore size). The samples were then analyzed for neurotransmitters and their respective metabolites using HPLC with electrochemical detection in a modification of a previously described method (19). The HPLC column was reversed-phase 3 /xm ODS, 250 mm x 4.6 mm (Bioanalytical Systems, West Lafayette, IN). The mobile phase contained 109 mmol/L citric acid, 8.8 mmol/L NaOH, 1.3 mmol/L octane sulfonic acid, 0.94 mmol/L EDTA and 1.3 mmol/L acetonitrile at pH 3.15. Electrochemical detection was at 850 inV. Since in our experience, filtrates that are injected show an internal standard recovery ratio of 100 + 2%, the internal standard is no longer included in these assays. External standards included the transmitters norepinephrine (NE), dopamine (DA), and serotonin (5HT), along with their respective metabolites as follows: for NE: 3-methoxy-4-hydroxy-mandelic acid (VMA) and 3-methoxy-4-hydroxy-phenylglycol (MHPG); for DA: homovanillic acid (HVA) and 3,4-dihydroxyphenylacetic acid (DOPAC); for 5-HT: 5-hydroxyindole acetic acid (5HIAA). Also included in the external standards was 3-hydroxykynurenin, because this tryptophan metabolite can co-elute with NE, and is elevated in vitamin B-6-deficient animals (20). In experiment 3, brains were also removed, frozen in liquid nitrogen and weighed. These whole brains were analyzed for total-folate (33) and for SAH and SAM using HPLC with UV detection (16) with minor modifications. For folate analysis, aliquots ( ~ 100 rag) of frozen brain were homogenized in 9 volumes of a 50 mM phosphate buffer (pH 6.1) containing 0.2% ascorbate, autoclaved for 10 min at 121°C, and cooled in an ice water bath. Cooled homogenates were centrifuged for 15 min at 2,000 g, and 150 /~L aliquots of the clear supematant extracts were transferred to tubes containing 2.79 ml of the 50 mM phosphate buffer (pH 6.1) containing 0.2% ascorbate plus 60 /xL chicken pancreas conjugase and mixed. The mixture was incubated at 37°C for 6 h to allow folacin polyglutamates to be converted to diglutamates. Aliquots of the conjugase treated extracts containing 0.1 to 2.0 ng folate were then analyzed for total folate (33). For SAH and SAM determinations, the frozen brain samples were homogenized in 5 volumes cold 0.4 M perchloric acid using a Polytron homogenizer (20 s., setting 7) (Brinkman Instruments, Westbury, NY) and centrifuged at 1500 × g for 10 min. A 1.0 ml aliquot of the superuatant was filtered through a 0.45 p,m Gelman syringe filter (VWR Scientific, San Francisco, CA), and 100 /xL of the filtered supernatant from each sample was analyzed by HPLC. The column was 3 /xm ODS Hypersil, 150 mm x 2 mm
NEUROCHEMISTRY, BEHAVIOR AND FOLIC ACID
(Keystone Scientific, Inc, Bellafonte, PA). The mobile phase was a linear gradient over 20 min for 100% solvent A (0.01 m o l / L ammonium formate, 4 m m o l / L heptanesulfonic acid, pH 4.0) to 75% solvent A and 25% solvent B (50% mobile phase A + 50% acetonitrile, pH 4.0). The 75% A / 2 5 % B ratio was maintained from 20 to 25 min, and then the system was returned to 100% solvent A for a 15 min equilibration. The flow rate was
30
2S,/
i
20,
m
15.
11:3HmolFolicAcid/kDiet g 0~la'nolFolioAci ~ d/kDi g et
,o
,o
,o
3o
,o
ZOO-"mOBtl-.~molFolicA~gOiet -"~ 1501'~1"1~0I~m°lF°hcAcid/kD' gt TS
937
0.3 m l / m i n . SAM and SAIl peaks were detected by UV absorption at 254 rim. SAM and SAH concentrations were determined using SAM and SAH external standards. Statistical analysis of data was performed with StatView software (Abacus Concepts, Berkeley, CA) for the Macintosh Computer. Since there were just two groups, Student's unpaired 2-tailed t-tests were used to determine significance of the effects of the folate-deficient diet on weight, food spill, food consumption, and levels of serum and brain chemicals. In experiment 2, one mouse from each dietary group failed to demonstrate spilling behavior consistent with their cohort. These animals were removed from the data set, leaving 4 mice per group in experiment 2. These two mice were statistical outliers because the average amount of food they spilled differed by - 3 . 2 6 (folate-deficient mouse) and - 2 . 2 2 (control mouse) standard deviations from the mean values of the 4 remaining mice in their respective groups. There were no mice in experiment 3 (n = 7/group), with such inconsistent spilling behavior. Some of the analyte data are expressed as concentration ratios because simultaneous observation of analyte concentration ratios, in addition to the concentrations, often provide insight into the biological implications of the data as has been illustrated for dopamine (3). Statistical significance was defined by a p value < 0.05. The values presented throughout this manuscript are means ___SEM. RESULTS
Systemic Effects of Folate Depletion
100
-
T /
50o zso T ~
,o
,;
~
3o
~
j ----o,---11.3lamolFol~Ac~kgDiet I ---o--- 0p.rnolFoiicAcid/kgDiet
1001
O
TT
,TI]J~'
~,i~-
I'3
2; Days Fed
3;
40
TestDiets
FIG. 1. Growth, cumulative food spilled, and cumulative food eaten by mice fed amino acid-based diets containing 0 or 11.3/zmol folic acid/kg. Values are mean + SEM of 4 mice fed the amino acid-based diet with and without folic acid, respectively. Body weight of mice fed the diet without folic acid were smaller after 3.5 wk. Cumulative diet spilled by mice fed the diet without folic acid was greater at 2.5 wk. Cumulative diet eaten was greater for mice fed the .diet without added folic acid throughout this experiment. The patterns of growth, and cumulative food spilled and eaten, as shown in this figure, were consistent for all three experiments.
After approximately 3.5 wk on the folic acid-deficient diet, growth rate declined and eventually these mice lost weight (Fig. 1). In experiment 2, at the end of the 37 day feeding period, folate-deficient animals weighed ~ 70% of the weight of the mice who were fed the control diet ( p = 0.01). This was a consistent finding among the three studies.Animals fed the folic acid-deficient diet consistently spilled their food (Table 1), and differences between the two groups in amount of food spilled first appeared after ~ 2.5 wk (Fig. 1). In experiment 2, the cumulative food spilled over the 37 day experiment in the folate-deficient group was 166 + 21 g vs. 5 2 + 11 g in the control group ( p = 0.0015). Even though the folate-deficient animals spilled more food, and gained less weight, they actually consumed more food during the course of the experiment. In experiment 2, the folate-deficient group consumed 219 + 21 g while the control group consumed 174 ___4 g ( p = 0.039). Similar observations were recorded in the animals from experiment 3 (Table 1). Folate deficiency was documented in experiment 2 by measuring the concentrations of folate, cysteine and homocysteine in serum (Table 1). Serum folate concentrations in the animals fed the folic acid-deficient diet were reduced to < 5% of those fed the control diet ( p = 0.0001). Folate depletion was also associated with a significant reduction of serum cysteine, elevation of serum homocysteine and deterioration of hematologic status (Table 1).
Neuropathological Studies Histologic examination of brain and spinal cord sections revealed no specific changes in the tissue from the folate-deficient animals. In particular, the cerebral cortex, caudate nucleus, corpus callosum, cerebellum, and white matter tracts showed no evidence of necrosis, atrophy or demyelination. Spinal cord sections from cervical, thoracic, lumbar and sacral levels were
938
GOSPE, JR. ETAL.
B. Soleus muscle fiber area distribution (~.m2)
A. Gastrocnemius muscle fiber area distribution (pm 2)
i "--
-coo,+
[] Folatedeficient
20
o ' I .n . . . . . . . 400
800
1200
111] 1600
2000
•
2400
•
1
I
2800
3200
Muscle Fiber Area (~m2) C. Control
i
~
•
.++
[] Folatedeficient
.~ 20 ~.
0
.
.
.
400
.
.
800
.
In .
.
1200
.
.
ii,,,
1600
.
.
2000
.
.
2400
.
.
2800
3200
Muscle Fiber Area (l~mz) D, Folate deficient
FIG. 2. Frequency distribution histogram of fiber cross sectional area in gastrocnemius (panel A) and soleus (panel B) muscles of mice fed an amino acid based diet supplemented with 0 or 11.3/zmole relic acid per kg. Distributions are based on data from ~ 150 fibers• Folate depletion induced a reduction in the number of larger fibers in both muscles• Representative cross sections of succinic dehydrogenase-stained gastrocnemius muscle from control and folate-deficient mice are shown in panels C and D, respectively• Magnification = 700.
examined and were without neuropathologic changes in the anterior horn, posterior columns or spinal roots. Brain weight of folate-deficient m i c e w a s less ( 4 2 7 - I - 6 vs. 480--I-6 m g , p = 0.006) c o m p a r e d to control m i c e . H o w e v e r , the b r a i n w e i g h t / b o d y w e i g h t ratio did not differ b e t w e e n the two groups• Since n e u r o m u s c u l a r d i s e a s e induced by f e e d i n g a folic aciddeficient diet for 3 7 - 3 9 d a y s could c a u s e the increased food spilling, the histology o f skeletal m u s c l e w a s e x a m i n e d . Soleus and g a s t r o c n e m i u s m u s c l e s h a d a n o r m a l distribution o f fibers h i g h in oxidative m e t a b o l i s m and s h o w e d no e v i d e n c e o f m y o p a thy or n e u r o g e n i c atrophy (data not shown). H o w e v e r , g a s t r o c n e m i u s and soleus m u s c l e s f r o m folate-deficient m i c e s h o w e d a reduction in the n u m b e r o f larger fibers c o m p a r e d to the s a m e m u s c l e s f r o m control mice (Fig. 2). T h e s e c h a n g e s in m u s c l e fiber size are not likely to be a m a j o r c a u s e o f the behavioral c h a n g e s associated w i t h relate depletion.
Biogenic Amines Dietary folic acid deficiency had significant effects on the concentrations o f b i o g e n i c a m i n e s a n d their m e t a b o l i t e s in brain (Table 2). T h e alterations in the h y p o t h a l a m i c region o f the brain were different f r o m t h o s e o b s e r v e d in the caudate nucleus. In the h y p o t h a l a m u s , folate-deficient a n i m a l s tended to h a v e m o r e 5 - H T ( p = 0.063) a n d less 5 - H I A A ( p = 0.086) e v e n t h o u g h neither trend w a s statistically significant (Table 2). T h e 5 - H I A A / 5 H T ratio w a s m u c h lower in relate-deficient m i c e c o m p a r e d to
TABLE 1 FINAL WEIGHT, FOOD SPILLED AND EATEN, KEY METABOLITES AND HEMATOLOGY IN MICE RECEIVING CONTROL AND FOLATE-DEF1CIENT DIETS* Dietary Folic Acid ( / z m o l / k g )
Experiment 2 Final bodyweight(g) Food spilled (g/37d) Food eaten (g/37d) Cysteine (/xmol/L serum) Homocysteine (/xmol/L serum) Folate(nmol/Lserum) Experiment 3 Final body weight (g) Food spilled (g/39d) Food eaten (g/39d) Brain weight (rag) Erythrocytes (10 12/ L ) Reticulocytes (109/L) Packed cell volume (%) Mean corpuscular volume (fL) Folate (nmol/kg brain) SAM (/xmol/kg brain) SAH (/xmol/kg brain) SAM/SAH ratio in brain
0.0
19.3 166 219 83 125 5.1
± ± ± ± ± ±
20.7 + 191 + 238 + 427 ± 2.09 ± 2.22 ± 22.6 ± 116 ± 61 ± 4.85 ± 4.00 ± 1.40 +
I 1.3
2.7t 21# 21t 12t 36t 0.9t
27.8 52 174 212 9 124
-t-_ 0.8 ± 11 ± 4 ± 14 ± 1 ± 6
0.5t 43t 16t 6t 0.29t 0.62 2.3t 16t 10t 0.65t 0.70 0.39t
28.2 30 202 480 5.74 3.04 50.0 87 116 8.13 3.43 2.47
± 1.0 ± 18 + 11 ± 6 ± 0.11 + 0.49 -I- 1.2 ± 3 ± 12 ± 1.19 ± 0.64 ± 0.34
* Values are means ± SEM. Experiment 2 lasted 37 days and results a r e means of 4 mice/group; data for one deficient and one control mouse were omitted because they spilled very little food (see text). Experiment 3 lasted 39 days and involved 7 deficient and 7 control mice. t Values that are significantly affected by relate deficiency at p < 0.05 based on unpaired, 2-tailed Student's t-tests.
NEUROCHEMISTRY, BEHAVIOR AND FOLIC ACID
TABLE 2 CONCENTRATIONS O!F BIOGENIC AMINES AND METABOLITES IN HYPOTHALAMUS AND CAUDATE OF CONTROL AND FOLATE-DEFICIENT MICE IN THE SECOND EXPERIMENT*,t p,Mol Analyte/kg wet weight of tissue Dietary Folic Acid (p, mol/kg) Hypothalamus L-DOPA DOPAC DA 5HIAA HVA 5HT DOPAC/DA HVA/DA DOPAC + H V A / D A 5HIAA/5HT Caudate L-DOPA DOPAC DA 5HIAA HVA 5HT DOPAC/DA HVA/DA DOPAC + H V A / D A 5HIAA/5HT
0.0
11.3
0.37 1.03 5.71 4.70 1.14 3.72 0.18 0.20 0.38 1.33
+ 0.02 + 0.15 + 0.32 + 0.77 + 0.33 + 0.67 _+ 0.01 + 0.06 5:0.06 + 0.22:~
0.31 1.04 6.63 7.20 1.79 1.71 0.17 0.27 0.45 4.38
0.13 1.38 26.5 2.41 2.15 2.28 0.05 0.09 0.15 1.14
5:0.01 + 0.39:~ + 5.6:~ + 0.34 5: 0.11:~ + 0.41 + 0.02 -t- 0.02 + 0.03 _+ 0.22
0.15 2.64 41.3 2.73 3.74 1.78 0.06 0.09 0.15 2.25
+ 0.02 + 0.38 + 0.69 + 0.82 + 0.28 + 0.36 5:0.08 + 0.04 + 0.11 + 0.43 + + + + + + + + + +
0.02 0.21 1.3 0.65 0.13 0.58 0.01 0.01 0.01 1.17
* Values are means + SEM of 4 folate-deficient and 3 control mice for the hypothalamus data and of 4 folate-deficient and 4 control mice for the caudate data. ~" Norepinephrine (NE) and metabolites, VMA and MHPG, were also measured but differences in concentrations of these metabolites between the folate-deficient and control groups were not significant, therefore only DA- and 5-HT-related compounds are shown. :~ Values that are significantly smaller in the folate-deficient group at p < 0.05 based on unpaired, 2-tailed Student's t-tests.
control mice ( p = 0.015). Differences in hypothalamic concentrations of norepinephrine, dopamine, or their metabolites between deficient and control mice were not significant. Caudate dopamine levels were reduced by 36% ( p = 0.04), DOPAC levels by 48% ( p = 0.03), and HVA levels by 43% ( p = 0.0002) in folate-deficient mice compared to control mice (Table 2). Brain Folate, SAM and SAH Animals fed the folic acid-deficient diet also had significantly lower whole brain folate and SAM (Table 1). Whole brain folate was 47% lower ( p = 0.01) and SAM was 40% lower ( p = 0.035). Whole brain SAH was not affected by folate depletion, however the SAM/SAH ratio was 43% lower ( p = 0.047) in folate-deficient mice. DISCUSSION
These studies have shown that feeding mice a folic acid-deficient diet for 37 - 39 days results in a reproducible decrease in weight gain followed by weight loss. In addition, even though the folate depleted mice spilled large amounts of their diet, they also consumed more. This combination of weight loss and augmented food consumption sugges~ts that folate depletion may reduce the efficiency by which animals utilize dietary nutrients to meet physiological functions, suggesting that an alteration in hypothalamic and/or neuroendocfine function may underlie this effect of folate depletion. In addition, these changes may be due to a folate deficiency-induced reductiLon in nutrient absorption. These studies were designed to characterize the food spilling behavior of folate-deficient mice and to determine if it might have a neurochemical basis. Several precautions were taken to
939
minimize the effects of diurnal variation on the results of this study. These included housing the mice in a room with a 12 h light dark cycle, feeding and weighing the mice at the same time each day and alternately killing control and deficient mice within a 2 h window at the end of each experiment. Alterations of serotonergic and dopaminergic systems by folate-depletion are suggested by these experiments. A significant reduction in the hypothalamic 5-HIAA/5HT ratio was obtained, while the folate-deficiency induced 35% reduction of hypothalamic 5-HIAA was not significant ( p = 0.086). It is possible that these changes were due to an inhibition of 5-HT metabolism. A reduction in whole brain 5-HT in folate-deficient rats has been shown in one previous study (11), whereas a more recent study did not show a change in either whole brain 5-HT or 5-HIAA in response to 10 wk of folate depletion (14). In that study, there was a small increase in 5-HIAA in the CSF of folate depleted rats. However, the concentrations of folate in serum and brain of those depleted rats were much higher than those observed in the folate-deficient mice in the present study. Therefore, the rats might not have been adequately depleted to alter biogenic amine metabolism. Clinical studies have also shown that folate deficiency (10-12) and 5,10-CHz-THFR deficiency (22,29) are associated with a reduction in 5-HIAA in the CSF. These results suggest that folate depletion does affect the hypothalamic serotonergic system. Examination of the effects of folate depletion on the serotonergic neurochemistry of the raphe nuclei, which contain the cell bodies that supply serotonin to the forebrain, may provide more specific information about these alterations. This study has also demonstrated an alteration in dopaminergic neurochemistry due to folate depletion. In the caudate, significant reductions in DA, DOPAC and HVA were noted, without a change in the level of L-DOPA. Although this might suggest that folate deficiency may reduce DA synthesis at the level of DOPA-decarboxylase, this is unlikely to be a rate limiting step in vivo because of the large quantities of DOPA-decarboxylase present. Alternatively, if low metabolite concentrations can be used to suggest a reduction in dopaminergic activity, folate depletion may have reduced nigrostriatal activity as measured neurochemically in the caudate nucleus. One previous animal study failed to show a change in whole brain levels of DA and HVA of rats in response to folate depletion (14). In that study, the use of whole rat brain may have obscured the changes in dopaminergic neurochemistry that were noted in caudate but not hypothalamus of the mouse. Because the deficient rat serum and brain folate concentrations were much higher than those observed in our deficient mice, those folate-deficient rats may not have been depleted of folate as severely as were the folate-deficient mice in the present study. Clinical studies have also shown reductions of HVA in the CSF of folate-deficient patients (10) and of HVA, as well as 5-HIAA and MHPG with 5,10-CHz-THFR deficiency (22,29). The clinical observations together with the data presented in this report suggest that folate deficiency alters monoaminergic function. This study also demonstrated a reduction in whole brain folate and SAM. Reduction of brain folate in folate-depleted mice is consistent with several recent studies (6,7,14,21) which have demonstrated that the concentration of folate in brain can vary markedly depending on the folic acid content of the diets. Given the close relationship between the metabolism of folates and SAM (30), the 40% reduction of whole brain SAM in the folate depleted mice was not unexpected. This study did not reveal any significant pathologic changes in brain, spinal cord, spinal roots or skeletal muscle in response to folate depletion. It was hypothesized that the food spilling behavior represented a symptom of folate-depletion-induced
940
GOSPE, JR. ET AL.
ataxia, weakness, or extrapyramidal dysfunction. While no microscopic changes were detected, the neurochemical changes do suggest some possible mechanisms which may underlie this food spilling behavior. For example food spilling due to folate-depletion may be a reflection of altered basal ganglia dopaminergic motor function. Active avoidance learning, but not motor function, was impaired in folate-deficient rats (2). In the present study, the mice appeared (subjectively) to have adequate motor function, but if spilling is associated with conditioned taste aversion (26), then this form of learning was apparently intact in these animals. Alternatively, this excessive spilling behavior may in some way represent a form of obsessive compulsive behavior associated with altered serotonergic function, as has been seen in other animal models (18). Reductions in CSF SAM levels have been noted in depressed patients (13), and SAM has been shown to have anti-depressant properties (17). Increased food spilling has been noted in other nutrient deficiencies (19,26,27) but the neurochemical basis for the behavior has not been studied. In a behavioral study of rats becoming thiamin deficient, the animals showed two responses that are not often seen in normal rats. Some rats vigorously spilled their food out of the food cup; they could empty 50 g of food in 15 min. Some deficient rats also demonstrated "redirected feeding" by chewing on the cage or other nonfood objects. These responses were interpreted as aversive behaviors directed toward the deficient diet as might be seen with poisoning (26). The poison-
avoiding responses of the rat have been reviewed (4), and may also relate to the conditioned taste aversion or " b a i t shyness" (15). Whatever the psychological mechanisms, some neurochemical alterations may be associated with the dramatic increases in food spillage that was observed. The present findings of alterations in both the serotonin and dopamine systems with folate deficiency suggest that the food spilling may have a neurochemical component. In addition, the lower feed efficiency of folatedeficient mice, as evidenced by greater feed intake but smaller weight gain, may imply neuroendocrine dysfunction, which could also have dopaminergic or serotonergic components. Clearly, more research on neurochemical mechanisms and neuroendocrine changes is needed. While these data suggest that food spilling behavior is associated with neurochemical changes, it is not clear which may come first. A separate longitudinal experiment with many animals and time points is needed to adequately address this important point. ACKNOWLEDGEMENTS This work was supported by grants DK-38637 from U.S. Public Health Service, regional research W-143 from U.S.D.A., and Hatch 2850 from the California Experiment Station. The authors thank Miriam Watson for assistance in measuring growth and the intake and spillage of food in the second experiment, Shan Shan Zhou for assistance with the histology, and Keith Hyland, Baylor Research Institute, Dallas TX, for thoughtfully reviewing this manuscript.
REFERENCES
1. Abou-Saleh, M. T.; Coppen, A. The biology of folate in depression: Implications for nutritional hypotheses of the psychoses. J. Psychiatr. Res. 20:91-101;1986. 2. Bachevalier, J.; Botez, M. I. Avoidance behavior in folate-deficient rats. Tohoku J. Exp. Med. 126:111-116;1978. 3. Bannon M. J.; Roth, R. H. Pharmacology of mesocortical dopamine neurons. Pharmacol. Rev. 35:53-68;1983. 4. Barnett, S. A. The rat. A study in behavior. London: Methuen; 1963. 5. Bills, N. D.; Koury, M. J.; Clifford, A. J.; Dessypris, E. N. Ineffective hematopoiesis in folate-deficient mice. Blood 79:22732280;1992. 6. Bills, N. D.; Hinrichs, S. H.; Morgan, R.; Clifford, A. J. Tissue folate levels in transgenic mice with tumors and in nontransgenic controls. J. Nutr. Biochem. 3:113-117;1992. 7. Bills, N. D.; Hinrichs, S. H.; Morgan, R; Clifford, A. J. Delayed tumor onset in transgenic mice fed a low folate diet. J. Natl. Cancer Inst. 84:332-337;1992. 8. Botez, M. I.; Botez, T.; Leveille, J.; Bielmann, P; Cadotte, M. Neuropsychological correlates of folic acid deficiency: Facts and hypotheses. In: Botez, M. I.; Reynolds, E. H., eds. Folic acid in neurology. Psychiatry, and internal medicine. New York: Raven Press; 1979:435-461. 9. Botez, M. I.; Fontaine, F.; Botez, T.; Bachevalier, J. Folate-responsive neurological and mental disorders:report of 16 cases. Eur. Neurol. 16:230-246;1977. 10. Botez, M. I.; Young, S. N. Effects of anticonvulsant treatment and low levels of folate and thiamine on amine metabolites in cerebrospinal fluid. Brain 114:333-348;1991. 11. Botez, M. I.; Young, S. N.; Bachevalier, J.; Ganthier, S. Folate deficiency and decreased brain 5-hydroxytryptamine synthesis in man and rat. Nature 278:182-183;1979. 12. Botez, M. I.; Young, S. N.; Bachevalier, J.; Gauthier, S. Effect of folic acid and vitamin B12 deficiencies on 5-hydroxyindoleacetic acid in human cerebrospinal fluid. Ann. Neurol. 12:479-484;1982. 13. Bottliglieri, T.; Laundy, M.; Carney, M. W. P.; Toone, B. K.; Reynolds, E. H. Cerebrospinal fluid S-adenosylmethionine in depression and dementia:effects of treatment with parenteral and oral
14. 15. 16. 17.
18. 19. 20.
21. 22.
23. 24.
S-adenosylmethionine. J. Neurol. Neurosurg. Psychiatry 53:10961098;1990. Butler, I. J.; Rothenberg, S. P. Dietary folate and biogenic amines in the CNS. J. Neurochem. 49:268-271;1987. Chambers, K. C. A neural model for conditioned taste aversions. Ann. Rev. Neurosci. 13:373-385;1990. Fell D.; Benjamin L. E.; Steele R. D. Determination of adenosine and S-adenosyl derivatives of sulfur amino acid in rat liver by high-performance liquid chromatography. J. Chromatogr. 345:150-156;1985. Friedel, H. A.; Goa, K. L.; Benfeld, P. S-adenosyl-L-methionine: A review of its therapeutic potential in liver dysfunction and affective disorders in relation to its physiological role in cell metabolism. Drugs 38:389-416;1989. Gietzen, D. W. Animal models elucidate the role of serotonin in OCD. Psychiat. Times Med. Behav. 9:17-20;1990. Gietzen, D. W.; Leung, P. M. B.; Rogers, Q. R. Norepinephrine and amino acids in prepyriform cortex of rats fed imbalanced amino acid diets. Physiol. Behav. 36:1071-1080;1986. Guilarte, T. R.; Block, L. D.; Wagner H. D. The putative endogenous convulsant 3-hydroxykyneurine decreases benzodiazepine receptor binding affinity:implications to seizures associated with neonatal vitamin B-6 deficiency. Pharmacot. Biochem. Behav. 30:665668;1988. Heid, M. K.; Bills, N.D.; Hinrichs, S. H.; Clifford, A. J. Folate deficiency alone appears insufficient to produce neural tube defects in mice. J. Nutr. 122:888-894;1992. Hyland, K.; Smith, I.; Bottiglieri, T.; Perry, J.; Wendel, U.; Clayton, P. T.; Leonard, J. V. Demyelination and decreased S-adenosylmethionine in 5,10-methylenetetrahydrofolate reductase deficiency. Neurology 38:459-462;1986. Melamed, R.; Reches, A.; Hershko, C. Reversible central nervous system dysfunction in folate deficiency. J. Neurol. Sci. 25:9398;1975. Ordonez, L. A.; Wurtman, R. J. Folic acid deficiency and methyl group metabolism in rat brain:effects of L-dopa. Reversal effect of S-adenosyl-L-methionine. Arch. Biochem. Biophys. 160:372376;1974.
NEUROCHEMISTRY, BEHAVIOR AND FOLIC ACID
25. Pincus, J. H.; Reynolds, E. H.; Giaser, G. H. Subacute combined degeneration with folate deficiency. JAMA J. Am. Med. Assoc. 221:496-497;1972. 26. Rozin, P. Specific aversions as a component of specific hungers. J. Comp. Physiol. Psych. 64:237-242;1967. 27. Rozin, P.; Kalat, J. W. Specific hungers and poison avoidance as adaptive specializations of learning. Psychol. Rev. 78:459-486;1971. 28. Sarnat, H. B. Muscle pathology and histochemistry. Chicago: American Society of Clinical Pathology Press; 1983. 29. Singer, H. S.; Butler, I; Rothenberg, S.; Valle, D.; Freeman, J. Interrelationships among serum folate, CSF folate, neurotransmitters, and neuropsychiatric symptoms. Neurology 30:419;1990. 30. Stokstad, E. L. R. Historical perspective on key advances in the biochemistry and physiology of folates. In: Picciano, M. F.; Stokstad, E. L. R.; Gregory, III, J. F., eds. Folic acid metabolism in health and disease, vol. 13, New York: Wiley and Sons, Inc.; 1990:1-21.
941
31. Strachan, R. W.; Henderson J. G. Dementia and folate deficiency. Quart. J. Med. 36:189-204;1967. 32. Vester B.; Rasmussen K. High performance liquid chromatography method for rapid and accurate determination of homocysteine in plasma and serum. Eur. J. Clin. Chem. Clin. Biochem. 29:549554;1991. 33. Wilson, S. D.; Home D. W. Use of glycerol-protected Lactobacillus casei for microbiological assay of folic acid. Clin. Chem. 28:11981200;1982. 34. Young, S. N.; Ghadirian, A. M. Folic acid and psychopathology. Prog. Neuro-psychopharmacol. Biol. Psychiatr. 13:841-863;1989. 35. Young, S. N. Some effects of dietary components (amino acids, carbohydrate, folic acid) on brain serotonin synthesis, mood, and behavior. Can. J. Physiol. Pharmacol. 69:893-903;1991.