Am J Physiol Endocrinol Metab 282: E672–E678, 2002;
10.1152/ajpendo.00280.2001.
Retinoic acid is detected at relatively high levels
in the CNS of adult rats
ELIZABETH A. WERNER AND HECTOR F. DELUCA
Department of Biochemistry, University of Wisconsin at Madison, Madison, Wisconsin 53706
Received 26 June 2001; accepted in final form 8 November 2001
all-trans-retinoic acid; vitamin A deficiency; central nervous
system; all-trans-retinol
VITAMIN A IS AN essential nutrient required for vision,
reproduction, cellular growth and differentiation, the
immune response, and embryological development (9,
18, 26). The majority of vitamin A actions are mediated
through the binding of its active metabolite, all-transretinoic acid (atRA), to nuclear receptors (4). Two families of retinoid receptors have been identified, retinoic
acid (RA) receptors (RAR-␣, -␤, and -␥) and retinoid X
receptors (RXR-␣, -␤, and -␥). RXR binds 9-cis-retinoic
acid (9cRA), whereas RAR binds both atRA and 9cRA
(2). Retinoid receptors have been identified in numerous tissues, including liver, kidney, spleen, testis, lung,
spinal cord, and brain (4).
The importance of vitamin A, particularly RA, for the
proper development of the central nervous system
(CNS) has been well established (4, 14, 28). An overabundance and lack of vitamin A cause a number of
defects in brain patterning (5, 11). Aldehyde dehydro-
Address for correspondence: H. F. DeLuca, Dept. of Biochemistry,
Univ. of Wisconsin-Madison, 433 Babcock Dr., Madison, WI 537061544 (E-mail: deluca@biochem.wisc.edu).
E672
genases that catalyze the formation of RA, retinaldehyde dehydrogenase-1 and -2, have been identified in
areas of embryos where the presence of RA is important for proper CNS development (19). A cytochrome
P-450 enzyme, P-450RAI-1, which is capable of catabolizing RA into polar derivatives, has been identified in
areas of embryos that are highly sensitive to the teratogenic actions of RA (1).
The conversion of retinol (ROL) to RA and the function of RA have not been well established in the adult
CNS. RA is synthesized in crude adult rabbit brain
homogenates (8) and was detected in specific areas in
the brain of songbirds (6), suggesting that RA is not
only required for the proper development of the CNS
but continues to play a role in the adult CNS as well.
Biochemical apparati, such as retinoid receptors and
cellular retinoid-binding proteins, that are required for
retinoid signaling are present in adult brain (4, 17, 31).
Impaired dopamine signaling and impaired locomotion
were observed in mutant mice in which a combination
of RAR-␤ and RXR-␤ or -␥ was ablated, suggesting a
possible function for RA in the adult mouse brain (12).
A novel cytochrome P-450 enzyme, P-450RAI-2, was
detected in adult brain, primarily in the cerebellum
(29), which may play an essential role in regulating RA
levels in the adult brain.
In this study, RA was detected throughout the brain
and in the spinal cord or vitamin A-deficient (VAD)
rats treated with 1 ␮g or 3.52 nmol all-trans-retinol
(atROL), an amount of atROL required for a biological
response. RA comprises a greater proportion of the
retinoid pool in the CNS compared with other target
tissues (27). The relatively high levels of RA present in
the brains of these rats is not the result of preferential
transport of RA from the blood to the brain. Although
we show that RA enters the brain, the levels of RA
transported to the liver, testis, and spleen were higher
than those observed for the brain 10 h after an injection of atRA. Because RA is not transported preferentially to the brain, it seems more likely that RA is
synthesized more efficiently in the tissues of the CNS
compared with other target tissues.
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
0193-1849/02 $5.00 Copyright © 2002 the American Physiological Society
http://www.ajpendo.org
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on September 29, 2016
Werner, Elizabeth A., and Hector F. Deluca. Retinoic
acid is detected at relatively high levels in the CNS of adult
rats. Am J Physiol Endocrinol Metab 282: E672–E678, 2002;
10.1152/ajpendo.00280.2001.—Retinoic acid (RA) is essential
for cellular growth and differentiation in developing and
adult animals. The central nervous system (CNS) suffers
developmental defects if embryonic levels of RA are too high
or too low. The production and function of RA in adult brain
are unclear. We report that RA is present throughout the
brain and spinal cord of adult, vitamin A-deficient (VAD) rats
treated with a physiological amount of all-trans-retinol. The
hippocampus/cortex contained the highest proportion of RA
in the brain (27.2 ⫾ 2.9% of the organic phase radioactivity,
and 23.5 ⫾ 0.8% of the organic phase radioactivity extracted
from spinal cord was RA). RA comprises a higher proportion
of the retinoid pool in the CNS compared with amounts
reported in other target tissues (E Werner and HF DeLuca.
Arch Biochem Biophys 393: 262–270, 2001). However, RA is
not preferentially transported from the blood to the brain.
There were 2.90 ⫾ 0.20 fmol RA/g tissue transported to the
brain of VAD rats treated with 2.00 nmol [20-3H]all-transretinoic acid, but higher amounts of RA were delivered to the
liver, testis, and spleen. Because RA is not transported preferentially to brain, this tissue likely synthesizes RA more
efficiently than other target tissues.
RETINOIC ACID IN THE CNS
METHODS
AJP-Endocrinol Metab • VOL
nomenex) was used as the stationary phase. A mixture of
30% acetonitrile, 25% methanol, 15% isopropyl alcohol, and
30% water [all containing 1.2% (vol/vol) acetic acid] was used
as the mobile phase. Putative ROL compounds from normalphase HPLC were quantitated using an isocratic reverse
phase system based on a method by MacCrehan and Schonberger (16). A 250 ⫻ 4.6-mm Develosil ODS-HG-5 column
(Phenomenex) was used as the stationary phase. A mixture
of 60% methanol, 10% n-butanol, and 30% water containing
10 mM ammonium acetate (Mallinkrodt, AR) was used as the
mobile phase.
RA transport experiment. Ten-week-old VAD male
Sprague-Dawley rats were maintained on a purified diet free
of vitamin A containing 12 ␮g RA/g diet while the experimental methodology was developed. Approximately 10 days before the experiment, the animals were fed a purified diet free
of RA. Weights of the animals were monitored until a weightgain plateau and subsequent loss of ⬃20 g body wt were
observed, indicating retinoid deficiency. Once all animals
displayed signs of vitamin A deficiency, loss of weight, and
xerophthalmia, the rats were injected intrajugularly with 0.6
␮g (116 ␮Ci) of [20-3H]atRA, 93% pure (NEN), in 100 ␮l
ethanol. Nine hours after RA injection, 1.5 mg (1.5 ␮Ci) of
[14C-carboxyl]inulin (Amersham Pharmacia) were injected in
50 ␮l of 0.15 M NaCl through the tail vein of each animal.
After 30 min, the animals were anesthetized. Blood was
collected, and the animals were perfused through the heart
with 80 ml PBS. Brain, liver, lung, spleen, and testis were
collected. Serum was isolated through centrifugation at 2,800
g for 15 min. Tissues were solubilized using Soluene (Packard). Levels of tritium and carbon-14 were measured using a
Tri-Carb 2300TR liquid scintillation counter (Packard).
Statistical analysis. Statistical analysis was performed
using the Student’s t-test. A P value ⬍0.05 was considered
significant.
RESULTS
RA is found throughout the brain and in the spinal
cord and comprises a proportionally higher level of the
retinoid pool compared with other target tissues. VAD
rats were treated with 250 ␮Ci (1 ␮g, 3.52 nmol) of
atROL. The brain and spinal cord were harvested 10 h
posttreatment. The brain was dissected into the following four sections: hippocampus-cortex, cerebellum,
brain stem, and remaining brain matter. Radiolabeled
retinoids were extracted from the tissue and were
analyzed using normal-phase HPLC. The chromatograms of the cerebellum and hippocampus-cortex extracts from one of the four rats and the organic extract
from the ROL degradation control experiment are
shown in Fig. 1. All areas of brain and spinal cord from
the four animals showed similar chromatograms.
Three peaks of radioactivity were detected in the organic extracts from the brain and spinal cord. A peak
coeluting with RA was observed between 7 and 10 min,
a peak coeluting with ROL was observed at 33 min,
and the final peak was eluted at 45 min. Because
retinoids are extremely labile compounds, control experiments were essential to determine whether the
radiolabeled compounds detected upon HPLC analysis
were metabolites or degradation products. The degradation control experiments were performed by homogenizing brains harvested from VAD rats not injected
with [20-3H]atROL. The homogenates were boiled and
282 • MARCH 2002 •
www.ajpendo.org
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on September 29, 2016
Metabolism of atROL. Weanling (21-day-old) male
Sprague-Dawley rats were maintained on a purified VAD
diet. After 4 wk, rats were weighed. On average, the rats
weighed ⬃250 g. Weights were monitored until a weight-gain
plateau and subsequent loss of ⬃20 g body wt were observed,
indicating vitamin A deficiency. Four rats were injected intrajugularly with 250 ␮Ci (1 ␮g, 3.53 nmol) of [20-3H]atROL
(NEN), 95% purity, in 100 ␮l ethanol. Ten hours after treatment, animals were killed. Brains were removed, and the
brain stem, cerebellum, and the hippocampus along with
cortex were isolated. The spinal cord was also harvested from
the four animals.
Extraction. The brain stem, cerebellum, hippocampus, cortex, remaining brain tissue, and spinal cord were weighed,
minced, and homogenized in degassed, deionized water (1
ml/g tissue) containing 0.15% n-propyl gallate (Sigma) with a
Teflon homogenizer (Wheaton).
Extraction of the homogenates was performed using a
modified version of the Bligh and Dyer method for lipid
extraction (13). Degassed methanol (Burdick and Jackson;
Muskegon, MI) containing 0.1% butylated hydroxytoluene
(BHT; Sigma) and degassed methylene chloride (Burdick and
Jackson) containing 0.1% BHT (Sigma) were added to the
homogenized tissue to give a 1:2:1 water-methanol-methylene chloride ratio. Tissue was extracted for 2 h at 4°C using
an extraction wheel. Samples were centrifuged at 3,200 g for
7 min using a Beckman Accuspin FR centrifuge. Supernatant
was decanted, and the tissue debris was resuspended in a
1:2:1 water-methanol-methylene chloride mixture. The tissue pellet was extracted overnight at 4°C using an extraction
wheel.
Samples were centrifuged as before, and the supernatants
from the washes were pooled with the initial extracts. Water
and methylene chloride were added to the supernatants to
give a 1:1:1 ratio of water-methanol-methylene chloride. The
organic and aqueous fractions were separated, and the organic extract was dried under N2. Samples were dissolved in
900 ␮l of column solvent for HPLC analysis.
Control experiments were performed by harvesting brains
from VAD rats that were not injected with radiolabeled ROL.
Tissues were homogenized and boiled. The boiled homogenate was incubated on ice in the presence of degassed methanol and methylene chloride, both containing 0.1% BHT.
[20-3H]atROL (NEN), 90% pure, 1 ␮Ci (60 ␮Ci/nmol) or
[20-3H]atRA (NEN), 92% pure, 1 ␮Ci (60 ␮Ci/nmol) was
added to the homogenate. The extraction procedure was
performed as described above.
HPLC. All HPLC was performed using a Waters 600 pump
and controller attached to a Waters 996 photodiode array
detector. A Foxy 200 fraction collector (Isco) was used to
collect fractions. Data were analyzed using Waters Millenium software. All solvents used in HPLC analysis were
purchased from Burdick and Jackson. Nonradiolabeled
atROL (Spectrum) and atRA (Spectrum), 1 ␮g of each, were
added to all samples before HPLC analysis. Organic extracts
were analyzed using a 250 ⫻ 4.6-mm, 7-␮m, Zorbax silica
column (Phenomenex). The normal-phase HPLC method was
adapted from Meyer et al. (20). A step gradient of solvent A
[hexane-isopropyl alcohol-acetic acid (200:0.5:0.5)] and solvent B (isopropyl alcohol containing 0.25% acetic acid) was
used. The exact gradient is shown in Figs. 1 and 2.
Reverse-phase HPLC was performed on the putative ROL
and RA compounds collected after normal-phase HPLC. This
method was similar to the one reported by Motto et al. (22).
A 250 ⫻ 4.6-mm Zorbax octadecyl silane (ODS) column (Phe-
E673
E674
RETINOIC ACID IN THE CNS
brain stem, hippocampus-cortex, cerebellum, remaining brain matter, and spinal cord, respectively (Table
1). These levels of RA are relatively higher than the
percentages of RA present in the organic extracts of
bone (5.95 ⫾ 0.66%), kidney (4.20 ⫾ 0.86%), liver
(9.41 ⫾ 1.96%), lung (5.45 ⫾ 0.45%), small intestine
(7.24 ⫾ 0.98%), skin (9.75 ⫾ 1.44%), spleen (7.46 ⫾
0.65%), and testis (2.31 ⫾ 0.66%) of rats injected with
1 ␮g [20-3H]atROL, as previously reported (27). Therefore, RA is not only the major metabolite of atROL in
the brain and spinal cord from these VAD rats 10 h
postinjection of a physiological amount of atROL; RA
comprises a relatively higher proportion of the retinoid
pool in the CNS than in other target tissues.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on September 29, 2016
Fig. 1. Normal-phase HPLC analysis of the organic extracts from the
cerebellum (black squares) and hippocampus-cortex (light gray
squares) from vitamin A-deficient (VAD) rats injected with 1 ␮g of
3.5 nmol (550,000,000 dpm) [20-3H]all-trans (at)-retinol (ROL). Dotted line (broken line) denotes the mobile phase gradient. The remaining brain and spinal cord samples from all 4 rats showed similar
chromatograms. Retinol was observed at 33 min in all tissue extracts
and in the retinol degradation control (dark gray squares). Retinoic
acid (RA) was detected between 5 and 12 min in all tissue extracts
but was not observed in the control. The peak eluted at 45 min was
present in the control and was not analyzed further. 9cROL, 9-cisROL. HRR, 14-hydroxy-retroretinol; DHT, 2-hydroxymethyl-3-methyl5-(2⬘-oxopropyl)-2,5-dihydrothiophene.
incubated on ice in the presence of organic solvents to
destroy the enzymatic activity responsible for the metabolism of atROL or atRA. Radiolabeled atROL or
atRA was added to the boiled homogenates, and extraction and HPLC analysis were performed. Two peaks of
radioactivity were observed in the ROL degradation
control experiment, one coeluting with ROL at 33 min
and the second at 45 min. Thus RA is not produced
through the degradation of ROL during the extraction
procedure. Two peaks of radioactivity were observed in
the RA degradation experiment, one coeluting with RA
at 8 min and the second eluting at 45 min. Reversephase HPLC analysis of the 45-min peak collected from
normal-phase HPLC demonstrated that the compounds eluted at 45 min during normal-phase HPLC
analysis were degradation products of ROL and/or RA
(data not shown).
The putative RA and ROL compounds were collected
and reanalyzed using reverse-phase HPLC to confirm
their identities. The putative RA coeluted with nonradiolabeled RA at 64 min upon reverse-phase HPLC
analysis (Fig. 2A). The putative ROL peak was observed at the same retention time, 51 min, as nonradiolabeled ROL (Fig. 2B). Therefore, RA is the major
metabolite of atROL present in the CNS of VAD rats
10 h posttreatment with a physiological amount of
atROL.
Quantitation of the radioactivity demonstrated that
RA comprised 17.5 ⫾ 1.7, 27.2 ⫾ 2.9, 11.1 ⫾ 0.7, 19.8 ⫾
1.6, and 23.5 ⫾ 0.8% of the total organic extract from
AJP-Endocrinol Metab • VOL
Fig. 2. Reverse-phase HPLC analysis of the putative retinoic acid (A)
and retinol (B) peaks collected from normal-phase HPLC of the
organic extracts from the cerebellum (black squares) and hippocampus-cortex (gray squares). The chromatograms from these two tissue
extracts are representative of the chromatograms collected from the
remaining tissue extracts from all 4 rats. Retinoic acid was detected
in all putative retinoic acid peaks eluting as the major peak upon
reverse-phase HPLC analysis at 64 min (A). The presence of retinol
in the tissue extracts was confirmed by detection of a peak of
radioactivity eluting at 51 min upon reverse-phase HPLC (B).
282 • MARCH 2002 •
www.ajpendo.org
E675
RETINOIC ACID IN THE CNS
Table 1. Percentages of organic phase radioactivity
from brain and spinal cord extracts detected as
retinoic acid and retinol
Brain stem
Hippocampus-cortex
Cerebellum
Remaining brain matter
Spinal cord
Retinoic Acid
Retinol
17.5 ⫾ 1.7
27.2 ⫾ 2.9
11.1 ⫾ 0.7
19.8 ⫾ 1.6
23.5 ⫾ 0.8
51.3 ⫾ 2.4
43.1 ⫾ 6.0
49.7 ⫾ 3.1
46.3 ⫾ 2.9
60.9 ⫾ 4.8
Data represent the average percentage of the radioactivity from
the organic extracts of brain and spinal cord found as retinoic acid or
retinol ⫾ SD.
Table 2. Amounts of retinoic acid and retinol detected
in tissue extracts
Brain
Spinal cord
Bone
Small intestine
Skin
Testis
Lung
Spleen
Liver
Kidney
Retinoic Acid
Retinol
2.78 ⫾ 0.34
1.19 ⫾ 0.27
0.29 ⫾ 0.05*
1.28 ⫾ 0.15†
0.40 ⫾ 0.03*
0.46 ⫾ 0.12†
1.37 ⫾ 0.31‡
0.64 ⫾ 0.08†
3.10 ⫾ 0.67
12.9 ⫾ 4.3
7.40 ⫾ 0.89
3.08 ⫾ 0.20
3.40 ⫾ 0.25
3.84 ⫾ 0.35
1.86 ⫾ 0.18
14.8 ⫾ 2.0
3.40 ⫾ 0.25
7.02 ⫾ 0.81
13.9 ⫾ 1.2
116 ⫾ 19.4
Data represent the average amounts of retinoic acid or retinol/g
tissue (means ⫾ SD) from 4 vitamin A-deficient rats injected with
3.52 nmol [20-3H]all-trans-retinol. * P ⬍ 0.001, † P ⬍ 0.005, and
‡ P ⬍ 0.01 compared with the amount of retinoic acid detected in the
brain.
AJP-Endocrinol Metab • VOL
Table 3. Amounts of retinoic acid transported to
tissues from the blood
Brain
Liver
Testis
Spleen
Lung
ng of RA/g
of tissue
pmol of RA/g
of tissue
0.87 ⫾ 0.06
8.74 ⫾ 0.38
2.13 ⫾ 0.25
1.45 ⫾ 0.15
0.61 ⫾ 0.02
2.9 ⫾ 0.2
29.1 ⫾ 1.3*
7.1 ⫾ 0.8†
4.8 ⫾ 0.5†
2.1 ⫾ 0.1†
Data represent the average amount of retinoic acid (RA) (means ⫾
SD) detected in tissues harvested from 4 vitamin A-deficient rats
treated with 0.6 ␮g [20-3H]all-trans-retinoic acid 10 h posttreatment.
* P ⬍ 0.001 and † P ⬍ 0.005 compared with the amount of RA
detected in the brain.
282 • MARCH 2002 •
www.ajpendo.org
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on September 29, 2016
The femtomoles of RA per gram tissue were calculated for the entire brain and spinal cord (Table 2) from
each rat. The level detected in the brain was 2.78 ⫾
0.34 fmol RA/g tissue and that in the spinal cord was
1.19 ⫾ 0.27 fmol RA/g tissue. The amount of RA detected in the brain was higher than the amounts previously reported for bone (0.29 ⫾ 0.05 fmol/g), small
intestine (1.28 ⫾ 0.15 fmol/g), skin (0.40 ⫾ 0.03 fmol/g),
testis (0.46 ⫾ 0.12 fmol/g), lung (1.37 ⫾ 0.31 fmol/g),
and spleen (0.64 ⫾ 0.08 fmol/g; see Ref. 27). Only the
liver and kidney contained more RA per gram tissue
than brain (3.10 ⫾ 0.67 and 12.90 ⫾ 4.30 fmol/g,
respectively; see Ref. 27). The liver and kidney also
contained far more ROL per gram tissue than brain
(13.9 ⫾ 1.2 and 116.0 ⫾ 19.4 compared with 7.40 ⫾
0.89 fmol ROL/g tissue in the brain; Table 2).
atRA enters the brain from the bloodstream. To determine whether the relatively high levels of RA detected in the brain were the result of preferential
transport of circulating RA to the brain, VAD rats were
injected with 0.6 ␮g (116 ␮Ci) of [20-3H]atRA intrajugularly. Nine hours post-RA injection, the rats were
injected through the tail vein with 1.5 mg (1.5 ␮Ci) of
[14C-carboxyl]inulin. Inulin is a polysaccharide used to
measure extracellular fluid volume in a number of
different tissues. The compound does not cross the
blood-brain barrier but is cleared rapidly from the body
by the kidneys (21). After inulin injection (30 min), the
rats were anesthetized. A 30-min time point was chosen to maximize [14C]inulin circulation throughout the
body while minimizing the renal clearance of inulin.
Blood was collected, and the rats were perfused
through the heart with 80 ml PBS. Both [3H]atRA
(168.5 ⫾ 14.9 nCi/ml) and [14C]inulin (8.2 ⫾ 0.6 nCi/ml)
were detected in the serum. The two compounds were
also detected in solubilized brain (178.0 ⫾ 11.5 and
0.11 ⫾ 0.03 nCi inulin/g tissue).
Because inulin does not cross the blood-brain barrier
(20), detection of inulin in the solubilized brain shows
that some blood remains in the tissue after perfusion
with PBS. Therefore, the [3H]RA measured in the solubilized brain is a combination of RA present in the
brain tissue and RA present in the blood remaining in
the tissue after perfusion with PBS. The volume of
blood remaining in the brain after perfusion was calculated by dividing the amount of inulin in the brain by
the concentration of inulin in the blood. By multiplying
the volume of blood in the brain after perfusion by the
concentration of [3H]RA in the blood, the amount of
[3H]RA resulting from blood contamination in the
brain can be subtracted from the total amount of
[3H]RA measured in the brain. Therefore, 0.87 ⫾ 0.6 ng
RA/g tissue or 2.9 ⫾ 0.2 pmol RA/g tissue was present
in the brain of these rats 9.5 h after injection of a
physiological amount of atRA (Table 3).
atRA is not preferentially transported to the brain
compared with other target tissues. Although atRA was
transported to the brain from the blood, atRA was also
transported to other target tissues through the bloodstream. Liver, testis, spleen, and lung from the VAD
rats given [3H]RA and [14C]inulin were solubilized, and
inulin and RA were quantitated. Both compounds were
detected in the tissues, demonstrating that blood remained in these tissues after perfusion with PBS.
Amounts of RA transported to the tissues were calculated as described above for the brain samples. The
liver contained the highest amount of RA [8.74 ⫾ 0.38
ng RA (29.1 ⫾ 1.3 pmol)/g tissue (Table 3)]. The lowest
levels of RA were found in the lung [0.61 ⫾ 0.02 ng RA
(2.1 ⫾ 0.1 pmol)/g tissue]. The levels of RA in testis and
spleen were 2.13 ⫾ 0.25 ng RA (7.1 ⫾ 0.8 pmol)/g tissue
and 1.45 ⫾ 0.15 ng RA (4.8 ⫾ 0.5 pmol)/g tissue,
respectively (Table 3). More RA was present in liver,
E676
RETINOIC ACID IN THE CNS
testis, and spleen than was present in the brain 10 h
postinjection of atRA. This suggests that either more
atRA was transported to the liver, testis, and spleen
than to the brain or that the liver, testis, and spleen
may clear RA less efficiently than the brain. If the
latter was true, it seems unlikely that relatively high
levels of RA would be detected in the brains of VAD
injected with a physiological amount of atROL, as
demonstrated in the initial experiment. Thus RA enters the brain but is not preferentially transported
from the blood to the brain compared with other target
tissues.
DISCUSSION
AJP-Endocrinol Metab • VOL
282 • MARCH 2002 •
www.ajpendo.org
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on September 29, 2016
Vitamin A is important for the proper formation and
patterning of the CNS during embryogenesis (14, 18).
This is especially clear in hindbrain development
where embryos from VAD mothers suffer defects in
hindbrain patterning (11, 28). This phenotype can be
rescued by treating the mothers with vitamin A or with
pharmacological doses of atRA (30), the bioactive metabolite of vitamin A responsible for cellular growth
and differentiation. Although RA is essential for embryological development, a teratogenic phenotype similar to the phenotype observed in VAD embryos is also
observed in embryos from mothers fed an excess of
vitamin A (5). Thus it is imperative that the synthesis
and degradation of RA in the developing CNS be
tightly regulated. The identification of a retinaldehyde
dehydrogenase, RALDH2, and isolation of an RA catabolic enzyme, P-450RAI-1, in embryonic tissues provide a potential mechanism for the regulation of RA
levels during development (19, 29).
The function of RA and conversion of ROL into RA in
the adult brain are not well understood. It seems likely
that vitamin A plays a role in neurological functions of
adult animals, since retinoid-binding proteins have
been identified in the adult brain and spinal cord. Both
cellular ROL-binding protein and ROL-binding protein
have been identified in cells comprising the blood-brain
barrier, suggesting a mechanism by which vitamin A
may enter the brain (17). The nuclear retinoid receptors RAR and RXR have been characterized in the
adult brain (4). Mice in which the genes encoding
RAR-␤-RXR-␤, RAR-␤-RXR-␥, and RXR-␤-RXR-␥ were
disrupted showed impaired locomotion (12). This locomotor defect was not because of abnormalities in the
skeletal muscle or peripheral nervous system but was
the result of a defect in the dopamine signaling pathway. A putative retinoic acid response element has
been characterized in the promoter encoding the D2
dopamine receptor, providing a possible link between
the retinoid signaling pathway and the dopamine signaling pathway (24). RA is not only required for the
development of cultured, nerve growth factor-supported, embryonic chick sympathetic neurons, but RA
is essential for the maintenance of neurites (23). This
suggests that RA may play a role in the survival of
developed neurites such as those found in the adult
CNS. The presence of RA itself in adult brain further
supports a physiological role for vitamin A in the adult
CNS. RA is produced from ROL in crude adult rabbit
brain homogenate (8) and has been detected in brain
from adult song birds (6). These studies suggest that
vitamin A exerts its actions on the CNS through the
binding of RA to nuclear receptors and subsequent
control of gene transcription.
This study demonstrates that RA is not only present
throughout the brain and in the spinal cord of adult
VAD rats given a physiological (1 ␮g, 3.52 nmol)
amount of atROL, but it is detected at levels that are
relatively higher than those detected in other target
tissues. RA comprised the highest proportion of the
organic phase radioactivity in the hippocampus-cortex,
and the lowest percentage of RA was detected in the
cerebellum. The relatively low levels of RA in the
cerebellum compared with other areas of the brain may
be because of an abundance of P-450RAI-2, a novel
RA-inducible, RA-catabolizing, cytochrome P-450 enzyme found in the adult brain, primarily in the cerebellum (29). The high expression of this RA-catabolizing enzyme in the cerebellum provides a mechanism
for enhanced RA clearance from that area of the brain.
The relatively high proportion of RA in the hippocampus-cortex found in this study supports the previous
work linking RA and proper motor skills (12), because
motor skills map to areas of the cortex.
The levels of RA detected in the brain are higher
than the percentages of RA present in the organic
extracts of other target tissues from rats injected with
1 ␮g (3.52 nmol) [20-3H]atROL, as we previously reported (27). Not only is RA detected in relatively higher
proportions in the brain, but the femtomoles of RA per
gram tissue detected in the brain are relatively high
compared with the amounts we previously reported
from other target tissues (27). Only the liver and kidney displayed higher levels of RA than the brain. Both
the liver and kidney contain much higher levels of ROL
than the brain, so relatively high levels of RA are
present in the brain even though there is a relatively
low abundance of ROL, the metabolic starting point in
the synthesis of RA, in the brain. Thus it seems possible that RA is either synthesized more efficiently in the
brain than in other target tissues or is transported
preferentially from the blood to the brain. Preferential
transport of circulating RA to the brain and spinal cord
would result in the relatively high levels of RA observed.
RA circulates through the bloodstream bound to serum albumin at very low levels under normal physiological conditions (7). Previous work demonstrated
that RA is transported to a variety of tissues in VAD
rats given small doses of RA (25). The highest levels of
RA were detected in the liver, kidney, and intestine,
whereas the lowest levels were detected in the testis
and fat pads. The brain was not analyzed in the study.
RA is detected in the brain of vitamin A-sufficient rats
treated with pharmacological amounts of RA (15).
However, it remained unclear whether RA, given at
physiological levels, could enter the brain of VAD rats.
Our work demonstrates that RA does enter the brain of
RETINOIC ACID IN THE CNS
We thank Dr. Margaret Clagett-Dame for valuable suggestions
and input, Dr. Lori Plum and Cindy Rohde for technical assistance
with the RA transport experiment, and Connie Smith for assistance
with the atROL metabolism experiment.
This work was supported in part by funds from the National
Foundation for Cancer Research, the Wisconsin Alumni Research
Foundation, and Chemistry-Biology Interface Training Grant no. 5
T32 GM-08505.
REFERENCES
1. Abu-Abed S, Dolle P, Metzger D, Beckett B, Chambon P,
and Petkovich M. The retinoic acid-metabolizing enzyme,
CYP26A1, is essential for normal hindbrain patterning, vertebral identity, and development of posterior structures. Genes Dev
15: 226–240, 2001.
2. Allenby G, Bocquel MT, Saunders M, Kazmer S, Speck J,
Rosenberger M, Lovey A, Kastner P, Grippo JF, Chambon
P, and Levin AA. Retinoic acid receptors and retinoid X receptors: interactions with endogenous retinoic acids. Proc Natl Acad
Sci USA 90: 30–34, 1993.
AJP-Endocrinol Metab • VOL
3. Bligh EG and Dyer WJ. A rapid method of total lipid extraction
and purification. Can J Biochem Physiol 37: 911–917, 1959.
4. Chambon P. A decade of molecular biology of retinoic acid
receptors. FASEB J 10: 940–954, 1996.
5. Collins MD and Mao GE. Teratology of retinoids. Annu Rev
Pharmacol Toxicol 39: 399–430, 1999.
6. Denisenko-Nehrbass N, Jarvis E, Scharff C, Nottebohm F,
and Mello CV. Site-specific retinoic acid production in the brain
of adult songbirds. Neuron 27: 359–370, 2000.
7. De Ruyter MG, Lambert WE, and De Leenheer AP. Retinoic
acid: an endogenous compound of human blood. Unequivocal
demonstration of endogenous retinoic acid in normal physiological conditions. Anal Biochem 98: 402–409, 1979.
8. Dev S, Adler AJ, and Edwards RB. Adult rabbit brain synthesizes retinoic acid. Brain Res 632: 325–328, 1993.
9. Dowling JE and Wald G. The biological function of vitamin A
acid. Proc Natl Acad Sci USA 46: 597–608, 1960.
10. Duester G. Families of retinoid dehydrogenases regulating vitamin A function. Eur J Biochem 267: 4315–4324, 2000.
11. Durston AJ, Timmerman JP, Hage WJ, Hendriks HF, de
Vries NJ, Heideveld M, and Nieuwkoop PD. Retinoic acid
causes an anteriorposterior transformation in the developing
central nervous system. Nature 340: 140–144, 1989.
12. Krezel W, Ghyselinck N, Sarnad TA, Dupe V, Kastner P,
Borrelli E, and Chambon P. Impaired locomotion and dopamine signaling in retinoid receptor mutant mice. Science 279:
665–667, 1998.
13. Kurlandsky SB, Gamble MV, Ramakrishnan R, and
Blaner WS. Plasma delivery of retinoic acid to tissues in the rat.
J Biol Chem 270: 17850–17857, 1995.
14. Langman J and Welch GW. Effect of vitamin A on development of the central nervous system. J Comp Neurol 128: 1–16,
1966.
15. Le Doze F, Debruyne D, Albessard F, Barre L, and Defer
GL. Pharmacokinetics of all-trans-retinoic acid, 13-cis-retinoic
acid, and fenretinide in plasma and brain of rat. Drug Metab
Dispos 28: 205–208, 2000.
16. MacCrehan WA and Schonberger E. Reversed-phase high
performance liquid chromatographic separation and electrochemical detection of retinol and its isomers. J Chromatogr Sci
417: 65–78, 1987.
17. MacDonald PN, Bok D, and Ong DE. Localization of cellular
retinol-binding protein and retinol-binding protein in cells comprising the blood-brain barrier of rat and human. Proc Natl Acad
Sci USA 87: 4265–4269, 1990.
18. Maden M, Gale E, and Zile M. The role of vitamin A in the
development of the central nervous system. J Nutr 128: 471S–
475S, 1998.
19. McCaffery P and Drager UC. High levels of a retinoic acidgenerating dehydrogenase in the meso-telencephalic dopamine
system. Proc Natl Acad Sci USA 91: 7772–7776, 1994.
20. Meyer E, Lambert WE, and De Leenheer AP. Simultaneous
determination of endogenous retinoic acid isomers and retinol in
human plasma by isocratic normal-phase HPLC with ultraviolet
detection. Clin Chem 40: 48–50, 1994.
21. Morrison AB. The distribution of intravenously-injected inulin
in the fluid of the nervous system of the dog and rat. J Clin Invest
38: 1769–1777, 1959.
22. Motto MG, Facchine KL, Hamburg PF, Burinsky DJ, Dunphy R, Oyler AR, and Cotter ML. Separation and identification of retinoic acid photoisomers. J Chromatogr Sci 481: 255–
262, 1989.
23. Plum LA and Clagett-Dame M. All-trans-retinoic acid stimulates and maintains neurite outgrowth in nerve-growth factorsupported developing chick embryonic sympathic neurons. Dev
Dyn 205: 52–63, 1996.
24. Samad TA, Krezel W, Chambon P, and Borrelli E. Regulation of dopaminergic pathways by retinoids: activation of the D2
receptor promoter by members of the retinoic acid receptorretinoid X receptor family. Proc Natl Acad Sci USA 94: 14349–
14354, 1997.
25. Smith JE, Milch PO, Muto Y, and Goodman DS. The plasma
transport and metabolism of retinoic acid in the rat. Biochem J
132: 821–827, 1973.
282 • MARCH 2002 •
www.ajpendo.org
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on September 29, 2016
VAD adult rats. However, much higher amounts of RA
were detected in the liver, spleen, and testis, and
similar amounts of RA were detected in the lung compared with the brain, suggesting that RA is not transported preferentially to the brain.
The relatively low amounts of RA transported to the
brain of VAD rats given a physiological amount of RA
are somewhat unexpected, since a previous study demonstrated that the majority of RA present in the brain
and liver is derived from circulating RA, whereas the
testis obtains very little RA from the circulation (13).
The aforementioned experiments were performed using vitamin A-sufficient rats that were infused with a
continuous amount of [3H]atRA. We used VAD rats
given a physiological amount of atRA. Thus it seems
that vitamin A status and the amount of RA given to
rats may affect tissue-specific transport of RA.
Because RA is not transported preferentially to the
brain, it seems likely that the formation of RA from
ROL is more efficient in the CNS than in other areas of
the body. A number of enzyme(s) responsible that convert atROL to atRA have been identified in various
adult, mammalian tissues, including liver, skin, eye,
adrenal gland, and reproductive tissue and embryos
(10). However, very little is understood about the metabolic pathway that converts atROL to atRA in the
adult CNS. Many of the enzymes credited with the
bioactivation of vitamin A lack substrate specificity,
and their physiological relevance is unclear (10).
This work clearly demonstrates that RA is present
throughout the brain and spinal cord of adult VAD rats
treated with a physiological amount of atROL. Not only
is RA detected throughout the brain and spinal cord, it
is present at relatively higher levels than in other
target tissues, suggesting a vital need for RA in the
CNS of VAD adult rats. Although RA can enter the
brain, it is not transported preferentially to the brain.
Thus it seems likely that RA is synthesized more efficiently in the CNS than in other target tissues. Ultimately, the CNS may be an excellent place to isolate
the enzyme(s) comprising the metabolic pathway that
converts atROL to atRA.
E677
E678
RETINOIC ACID IN THE CNS
26. Thompson JN, Howell JM, and Pitt GAJ. Vitamin A and
reproduction in rats. Proc Royal Soc Biol 159: 510–535, 1964.
27. Werner E and DeLuca HF. Metabolism of a physiological
amount of all-trans-retinol in the vitamin A-deficient rat. Arch
Biochem Biophys 393: 262–270, 2001.
28. White JC, Highland M, and Clagett-Dame M. Abnormal
development of the inuatrial venous valve and posterior hindbrain may contribute to late fetal resorption of vitamin A-deficient rat embryos. Teratology 62: 374–384, 2000.
29. White JA, Ramshaw H, Taimi M, Stangle W, Zhang A,
Everingham S, Creighton S, Tam SP, Jones G, and Petkovich M. Identification of the human cytochrome P450,
P450RAI-2, which is predominantly expressed in the adult cerebellum and is responsible for all-trans-retinoic acid metabolism.
Proc Natl Acad Sci USA 97: 6403–6408, 2000.
30. White JC, Shankar VN, Highland M, Epstein ML, DeLuca
HF, and Clagett-Dame M. Defects in embryonic hindbrain development and fetal resorption resulting from vitamin A deficiency in
the rat are prevented by feeding pharmacological levels of all-transretinoic acid. Proc Natl Acad Sci USA 95: 13459–134564, 1998.
31. Yamamoto M, Drager U, Ong DE, and McCaffery P. Retinoid-binding proteins in the cerebellum and choroid plexus and
their relationship to regionalized retinoic acid synthesis and
degradation. Eur J Biochem 257: 344–350, 1998.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on September 29, 2016
AJP-Endocrinol Metab • VOL
282 • MARCH 2002 •
www.ajpendo.org