©Copyright 1993 by Humana Press Inc.

All rights of any nature whatsoever reserved.

1044-7393/93/1801-02-0123 $02.00

Regional Lipid Composition

in the Rat Brain
MIKULAS CHAVKO, l EDWIN M. NEMOTO,*. 2
AND JOHN A. MELICK 2
The Institute of Neurobiology, Slovak Academy of Sciences in Kosice,
Czechoslovakia; 2 Department of Anesthesiology and Critical Care
Medicine, University of Pittsburgh, School of Medicine, 1081 Scaife Hall,
Pittsburgh, PA 15261

Received January 17, 1992; Accepted June 4, 1992

ABSTRACT
The lipid composition of the brain is of great importance to its
metabolism and function. Although much research has been done on
regional brain lipid composition, studies usually suffer from limited
brain regions or from limited lipids analyzed. We modified a previously
described method for the separation of brain phospholipids and
glycolipids, improving the separation and sensitivity of the method.
Using this modified method, we measured the lipid composition of
the frontal and entorhinal cortices, the hippocampus, basal ganglia,
cerebellum, and medulla oblongata of five rats under nitrous oxide
analgesia. Total lipid content was highest (p<0.05) in the medulla
oblongata (111.0±6.0 mg/g wet brain, X±SD) followed by the hippo-
campus (72.6±2.8), cerebellum (62.7±4.6), basal ganglia (62.6±1.5),
frontal cortex (57.7±2.1), and entorhinal cortex (53.3±1.9). The areas
with higher total lipid content (p < 0.05) also had higher percentages
of cerebrosides (18.6±2.2 in the medulla oblongata vs 8.3±1.2 in the
frontal cortex) and 40 to 50% lower levels of phosphatidylcholine and
phosphatidylinositol. The relation between the ratio of cerebrosides
plus sulfatides to phosphatidyicholine and the total lipid content in-
dicates that differences in brain lipid composition between regions
are attributable to their relative gray/white matter content.
Index Entries: Cerebrosides; sulfatides; phospholipids; brain
regions.

*Author to whom all correspondence and reprint requests should be addressed.

Molecular and Chemical Neuropathology 123 Vol. 18, 1993

124 Chavko, Nemoto, and Melick

Abbreviations: CER, cerebrosides; CH, cholesterol; PC, phos-

phatidylcholine; PE, phosphatidylethanolamine; PI, phosphatidyl-
inositol; PS, phosphatidylserine; SPM, sphingomyelin; SU, sulfatides;
TL, total lipid.

INTRODUCTION
The lipid composition of the brain is of great functional and metabolic
importance. For example, phospholipids, a major constituent of biologi-
cal membranes, are involved in the regulation of membrane properties
related to structure, permeability, enzyme activities, and receptor affini-
ties (Cullis and Hope, 1985). Cholesterol, another major membrane con-
stituent, increases membrane stability and decreases permeability (Bloch,
1985). In neurons, whose functional activity is critically dependent on
membrane properties, functional differences might be reflected in dif-
ferences in membrane lipid composition. However, relatively few studies
have attempted to characterize the lipid composition of different brain
regions (O'Brien and Sampson, 1965; Svennerholm and Vannier, 1972;
Hoshi et al., 1973; Rao, 1977; Nonaka and Kishimoto, 1979; Ishibe and
Yamamoto, 1979; Macala et al., 1983; Okazaki et al., 1990; Soderberg et
al., 1990), partly because of the difficulty in quantitating lipids in small
amounts (20-30 mg) of brain tissue. In addition, most of the published
data provide information on only one or two lipid classes.
In this paper, our aim is to characterize the distribution of phospho-
lipids, cholesterol, and glycolipids in the cerebral cortex, hippocampus,
basal ganglia, cerebellum, and brain stem by a combination of column
chromatography, high performance thin layer chromatography (HPTLC),
and densitometry using a modification of a method previously described
by Macala et al. (1983).

MATERIALS AND METHODS

Materials
Phospholipid, cholesterol, and glycolipid standards were obtained
from Sigma Chemical Company (St. Louis, MO). Solvents of HPLC grade
were obtained from Fisher (Pittsburgh, PA), HPTLC plates (10x20 cm,
Silica gel 60, 0.20 mm) were obtained from Merck (Darmstadt, Germany),
and Bond-Elute NH 2 columns were obtained from Analytichem Interna-
tional (Harbor City, CA).

Surgical Procedures
Male Wistar rats (350-450 g) were intubated under 4% halothane
anesthesia, immobilized with 0.3 mg pancuronium bromide (Elkin Sinns,

Molecular and Chemical Neuropathology Vol. 18, 1993

Regional Brain Lipids 125
Cherry Hill, NJ), and ,their lungs mechanically ventilated (Model 680
Rodent Respirator, Harvard Apparatus, South Natick, MA) with 0.5%
halothane/70% N 20/30% 0 2 . The'femoral artery and vein were cannulated
for measurement of blood pressure, PaCO 2 , Pa0 2, and pH, and for
measurement of fluid and drug infusions, respectively. A plastic funnel
fitted onto the skull was used for in situ freezing of the brain with liquid
N 2 (Ponten et al., 1973). After topical application of lidocaine jelly (2%) to
all surgical wounds, halothane was discontinued for 30 min, after which
time liquid N 2 was poured into the funnel to freeze the brain in situ with
blood pressure maintained at 80-100 mm Hg by titrated intravenous infu-
sion of 0.01 mg/mL norepinephrine (Sanofi Winthrop Pharmaceuticals,
New York, NY) to prevent postmortem changes in brain lipids.

Sample Dissection and Extraction

The brain was dissected at -15°C in a refrigerated chamber. Tissue
samples (20-30 mg) were obtained from the frontal cortex (A28-31), the
basal ganglia (A28-31), the hippocampus (A20-23), the entorhinal cortex,
the medulla oblongata, and the cerebellar cortex and placed into pre-
weighed test tubes containing 2.5 mL chloroform/methanol (1:1) and pro-
cessed in a Polytron homogenizer (Brinkman, Inc., Westbury, NJ) for 3
min. Lipid analyses, described briefly below, were done as described by
Macala et al. (1983).

Column Chromatography
The columns (Bond-Elute NH 2 ) were equilibrated with 20 mL chloro-
form/methanol/0.8M sodium acetate (30:60:8 [v/v]) and washed with 50
mL chloroform/methanol/water (30:60:8 [v/v]). The brain samples were
applied to the column and the neutral lipids eluted with 15 mL chloro-
form/methanol/water (30:60:8). Acidic lipids were eluted with 15 mL
chloroform/methanol/0.8M sodium acetate (30:60:8). Recovery of neutral
and acidic lipids from the columns was estimated at about 95% using
phosphatidylcholine (PC) and phosphatidylserine (PS) standards.

HPTLC
One to six micrograms of each lipid and four different standard mix-
tures were spotted on HPTLC plates prewashed with chloroform/metha-
nol/water (60:35:8 [v/v]) activated at 115°C for 15 min, and cooled in a
1

vacuum desiccator. Neutral lipids were developed with two consecutive

runs in two solvent systems: chloroform/methanol/water (65:25:2.5 [v/v])
and hexane/diisopropylether/acetic acid (65:35:2 [v/v]). Acidic lipids were
developed in chloroform/methanol/acetic acid/formic acid/water (35:15:6:2:1
[v/v]). The plates were dipped into 3% cupric acetate (w/v) and 8% phos-
phoric acid (v/v) and heated in an oven at 180°C for charring of neutral
(Fig. 1A) and acidic (Fig. 1B) lipids.

Molecular and Chemical Neuropathology Vol. 18, 1993

126 Chavko, Nemoto, and Melick

B s ss ¤ V
Al

® d i^ SU
_
"^-^
—
ps
^— — oe ® .. — — P1

1
Fig. 1. Representative chromatograms from densitometric quantitation
of (A) neutral and (B) acidic lipids from rat brain cortex. In A and B, the two out-
side lanes are standards 1-4 and the other lanes are rat brain samples.

Densitometry
TLC plates were scanned using a Bio-Rad-620 reflectance densitometer
(Bio-Rad, Richmond, CA) and areas computed using the Bio-Rad 1-D
Analyst software. Standard calibration curves were generated and curvi-
linear regression equations constructed. The amount of lipid in each
sample was calculated from the regression equation.

Statistical Analysis
Statistical significance was determined by one-way analysis of
variance and t-test. All data are expressed as X± SD.

RESULTS
The total lipid content of the medulla oblongata was 1.5 to 2 times
higher than that of the other regions examined (Table 1). The total lipid
content in other brain regions varied from 73 to 53 mg/g wet brain in the

Molecular and Chemical Neuropathology Vol. 18, 1993

Regional Brain Lipids 127

Table 1
Lipid Composition of Rat Brain Regionsa

TL SPM PC PE CER CH PI PS SU

F. Cortex 57.7 1.54 16.08 14.92 4.82 12.46 1.20 4.62 2.08
±2.1 ±0.21 ±1.54 ±1.30 ±0.76 ±0.96 ±0.18 ±0.61 ±0.61
E. Cortex 53.3 1.78 15.98 13.64 3.96 10.60 1.12 4.56 1.64
±1.90 ±0.25 ±1.09 ±0.30 ±0.82 ±0.84 ±0.13 ±0.30 ±0.38
Hippocampus 72.6 2.08 17.72 17.10 9.50 16.84 0.90 5.34 3.16
±2.8 ±0.38 ±0.82 ±1.06 ±1.94 ±1.79 ±0.16 ±0.80 ±0.51
Basal 62.6 1.94 16.86 14.86 7.24 13.50 1.08 4.76 2.36
ganglia ±1.5 ±0.32 ±0.69 ±0.77 ±0.78 ±0.75 ±0.08 ±0.25 ±0.55
Cerebellum 62.7 2.40 17.34 13.96 8.16 12.84 1.14 4.26 2.56
±4.6 ±0.48 ±1.08 ±0.38 ±1.95 ±0.70 ±0.11 ±050 ±0.74
Medulla 111.0 3.90 20.22 25.48 20.72 24.48 1.24 6.66 8.24
oblongata ±6.0 ±0.53 ±1.09 ±1.75 ±3.42 ±1.67 ±0.18 ±0.52 ±0.79
Abbreviations: TL=total lipid, SPM sphingomyelin, PC= phosphatidylcholine,
PE = phosphatidylethanolamine, CE = cerebrosides, CH = cholesterol, PI = phosphatidyl-
inositol, PS = phosphatidylserine, SU = sulfatides.
a Results are expressed in mg/g wet brain X±SD (n=5).

following descending order: hippocampus> cerebellum = basal ganglia>

frontal cortex> entorhinal cortex.
Most striking is that except for phosphatidylinositol (PI), the medulla
had the highest concentration of each of the lipids compared to other
brain regions (Table 1). PC and phosphatidylethanolamine (PE) were the
largest phospholipid components at about 15 mg/g brain; in the medulla,
they ranged between 20 and 25 mg/g brain, respectively. They were
followed by PS, sphingomyelin, and PI in quantity. PC and especially PE
were higher in the hippocampus than in both cortical regions, the basal
ganglia, or the cerebellum. PS and PI were generally similar between the
different brain regions.
Cholesterol was two times higher in the medulla than other brain
regions. At 16 mg/g brain in the hippocampus, it was higher than in the
other regions except for the medulla. Sulfatides were in the range of 1 to 2
mg/g brain except for the medulla at 8 mg/g brain followed by the hippo-
campus, at about 3 mg/g brain.
Cerebroside content varied between brain regions (Table 1). It was
highest in the medulla oblongata (p< 0.01 compared to all other regions),
2 to 4 times higher than in the hippocampus, cerebellum, and basal
ganglia, all of which were similar to each other. Cortical regions had the
lowest levels (p<0.05, compared with all other regions).
The molar PE/PC ratio was 1.26 in the medulla oblongata, 0.96 in the
hippocampus, 0.86 and 0.89 in enthorhinal and frontal cortices, respec-
tively, and 0.81 in the cerebellum (Table 2).

Molecular and Chemical Neuropathology Vol. 18, 1993

128 Chavko, Nemoto, and Melick
Table 2
Molar Ratios of PE to PC and Cerebrosides
Plus Sulfatides to PC in Rat Brain Regionsa
Region PE/PC (CE + SU)/PC
Frontal cortex 0.89±0.05 0.43±0.07
Entorhinal cortex 0.86±0.06 0.35±0.08
Hippocampus 0.96±0.05 0.72±0.12
Basal ganglia 0.88±0.07 0.60±0.07
Cerebellum 0.81±0.05 0.68±0.05
Medulla oblongata 1.26±0.07 1.45 ±0.23
Abbreviations: PE = phosphatidylyethanolamine,
PC = phosphatidylcholine, CE = cerebrosides, SU = sulfatides.
Values are means ± SD (n = 5).

The molar ratio of cerebrosides plus sulfatides to PC was character-

istic of the brain regions examined. In general, the higher the total lipid
concentration in a given brain region, the higher the ratio (Table 2).

DISCUSSION
Although degradation of membrane lipids is believed to be negligible
even hours after death (O'Brien and Sampson, 1965), rapid accumulation
of brain free fatty acids (FFA) after decapitation ischemia (Bazan, 1970;
Shiu et al., 1981) indicates that membrane lipid degradation occurs rapidly
postmortem. It contradicts the suggestion that hydrolysis products of
lipids, including FFA, were similar in postmortem and surgical biopsy
specimens (O'Brien and Sampson, 1965). The quantity of FFA accumulat-
ing 1 h postmortem is a small fraction of the total available FFA pool.
Nevertheless, its accumulation indicates rapid hydrolysis of membrane
lipids, which may be important compared with levels reported for differ-
ent human brain regions from postmortem tissue (Soderberg et al., 1990).
Our chromatography and densitometry methods combined two mod-
ifications of the separation and quantitation techniques developed by
Macala et al. (1983). First, we separated neutral from acidic lipids using
prepacked Bond-Elute NH, columns instead of DEAE-Sephadex columns,
thus eliminating the time-consuming need to prepare the columns. The
recovery of neutral and acidic lipids and their separation on the column
was estimated with a mixture of PC and PS. No detectable cross-contam-
ination of either fraction occurred, and recovery was about 95%. Macala
et al. (1983) and Wood et al. (1989) reported some loss of acidic phospho-
lipids during separation on the DEAE-columns and during a partitioning
step to remove sodium acetate. Despite this potential loss, the PS levels
they reported were about two times higher than those reported by Norton
and Poduslo (1973).

Molecular and Chemical Neuropathology Vol. 18, 1993

Regional Brain Lipids 129
Second, we used two solvent systems to separate neutral from acidic
lipids. The use of chloroform/methanol/water instead of the mixture used
by Ando et al. (1980) gave better resolution of the neutral lipids, especially
PE and the cerebrosides. The solvent system for the acidic lipids was as
described by Macala et al. (1983).
The only previous study that evaluated brain regions in similar detail
to this study (e.g., gray and white matter, caudate, hippocampus, pons,
cerebellum, and medulla oblongate) was Soderberg et al.'s (1990) study
in human autopsy material. Cholesterol and PL were the only primary
lipids quantitated. The total PL content they reported for the different
brain regions were clearly lower than our values in rats. For example, in
the rat hippocampus, we found total PL values in the order of 43 mg/g
wet brain, whereas they reported values between 17 and 23 mg/g wet
brain. A difficulty in comparing our findings with those of previous
reports is that the lipid values were often presented as molar percentages
of total lipid without the total lipid values.
The total lipid level in the cerebral cortex in our study was about 55
mg/g wet brain corresponding to a value of 59 mg/g for gray matter in the
human brain. Except for sphingomyelin, the lipid composition of the
cerebral cortex was very similar to that found in human gray matter
(Norton, 1976). There were no significant differences between the frontal
and entorhinal cortices in any of the lipids analyzed. Comparable values
in the literature are unavailable.
The regional distribution of lipids in the rat brain generally varied in
accordance with the relative distribution of gray or white matter. Lipid
levels were lowest in the cortices, increasingly higher in the basal ganglia
and cerebellum, and followed by the hippocampus and medulla oblongata.
A similar relative distribution of phospholipids was reported in the
human brain (Soderberg et al, 1990). There was a good correlation between
total lipid and cerebroside content in the various brain regions, namely,
regions with higher total lipid content also had the higher cerebroside
content.
The cerebrosides and sulfatides are deposited in the brain at the same
rate as myelin (Norton and Poduslo, 1973; Suzuki, 1975; Nonaka and
Kishimoto, 1979). Although these lipids appear to be ubiquitous in the
nervous system, they are excellent markers for myelin. Therefore, the
regional quantities of lipids in the rat brain reflect primarily the mye-
linated fiber content within the various brain regions. This notion is sup-
ported by the molar ratio of galactolipids to PC. Like the cerebrosides,
sulfatides appear to have a similar relationship to total lipids (Nonaka and
Kishimoto, 1979).
Regional variations in PC and PI appear to be inversely related to the
levels of cerebrosides and total lipids: The greater the amount of total
lipids and cerebrosides, the smaller the PC and PI content. These findings
are consistent with the low levels of PC and PI found in isolated myelin
compared with other membranes (Norton and Autillio, 1966). The regional

Molecular and Chemical Neuropathology Vol. 18, 1993

130 Chavko, Nemoto, and Melick
distribution of lipids in the rat brain is primarily a result of differences in
myelin content. However, as indicated by the ratio of PE to PC, this
dependence may not be absolute. The PE to PC ratio does not correspond
with the galactolipid to PC ratio and therefore may reflect differences in
the lipid composition of nonmyelin membranes.

ACKNOWLEDGMENTS
This work was supported in part by the American Heart Association
(Western Pennsylvania Affiliate), the University Anesthesiology and Crit-
cal Care Medicine Foundation, and the U.S. Public Health Service, Grant
No. 5 RO1 H L27208-09. The authors gratefully acknowledge the editorial
assistance of Lisa Cohn and the technical assistance of Zachery Pantazes.

REFERENCES
Ando S., Kon K., and Tanaka Y. (1980) High-precision TLC-densitometry of
membrane lipids, in Membrane Fluidity: Biophysical Techniques and Cellular
Regulation (Kates M. and Kuksis A., eds.), pp. 43-55. Humana, Clifton, NJ.
Bazan N. G. (1970) Effects of ischemia and electroconvulsive shock on free fatty
acid pool in the brain. Biochim. Biophys. Acta 218, 1-10.
Bloch K. (1985) Cholesterol: evolution of structure and function, in Biochemistry of
Lipids and Membranes (Vance D. E. and Vance J. E., eds.), pp. 1-24. Benjamin/
Cummings, Menlo Park, CA.
Cullis P. R. and Hope M. J. (1985) Physical properties and functional roles of
lipids in membranes, in Biochemistry of Lipids and Membranes (Vance D. E.
and Vance J. E., eds.), pp. 25-72. Benjamin/Cummings, Menlo Park, CA.
Hoshi M., Williams, M., and Kishimoto Y. (1973). Characterization of brain cere-
brosides at early stages of development in the rat. J. Neurochem. 21, 709-712.
Ishibe T. and Yamamoto A. (1979) Regional differences in lipid composition in
rabbit nervous tissue. J. Neurochem. 32, 1665-1670.
Macala L. J., Yu R. K., and Ando S. (1983) Analysis of brain lipids by high per-
formance thin-layer chromatography and densitometry. J. Lipid Res. 24,
1243-1250.
Nonaka G. and Kishimoto Y. (1979) Levels of cerebrosides, sulfatides, and galac-
tosyl diglycerides in different regions of rat brain. Biochim. Biophys. Acta 572,
432-441.
Norton W. T. (1976) Formation, structure and biochemistry of myelin, in Basic
Neurochemistry (Siegel G. J., Albers R. W., Katzman R., and Agranoff B. W.,
eds.), pp. 74-99. Little, Brown, Boston, MA.
Norton W. T. and Poduslo S. E. (1973) Myelination in rat brain: changes in mye-
lin composition during brain maturation. J. Neurochem. 21, 759-773.
Norton W. T. and Autilio L. A. (1966) The lipid composition of purified bovine
brain myelin. J. Neurochem. 13, 213-222.

Molecular and Chemical Neuropathology Vol. 18, 1993

 Regional Brain Lipids 131
O'Brien J. S. and Sampson E. L. (1965) Lipid composition of the normal human
brain: gray matter, white matter and myelin. J. Lipid Res. 6, 537-544.
Okazaki M., Sakamoto H., Ohtsuki-A., and Oguchi K. (1990) Changes in brain
lipid composition in thamine deficient rats. lap. J. Pharmacol. 54, 171-178.
Ponten U., Ratcheson R. A., Salford L. G., and Siesjo B. K. (1973) Optimal freez-
ing conditions for cerebral metabolites in rats. J. Neurochem. 21, 1127-1138.
Rao P. S. (1977) Fatty acids of cerebrosides in different regions of the developing
foetal brain. Lipids 12, 335-339.
Shiu G. K., Nemmer J. P., and Nemoto E. M. (1981) Barabiturate attenuation of
brain free fatty acid liberation during global ischemia. J. Neurochem. 37,
1448-1456.
Soderberg M., Edlund C., Kristensson K., and Dallner G. (1990) Lipid composi-
tions of different regions of the human brain during aging. J. Neurochem. 54,
415-423.
Suzuki K. (1975) Sphingolipids of the nervous system, in The Nervous System
(Tower D. B., ed.), pp. 483-491, Raven, New York.
Svennerholm L. and Vanier M. T. (1972) The lipid distribution in the human
nervous system. II. Lipid composition of human fetal and infant brain. Brain
Res. 47, 457-468.
Wood W. G., Cornwell M., and Williamson L. S. (1989) High performance thin-
layer chromatography and densitometry of synaptic plasma membrane
lipids. J. Lipid Res. 30, 775-779.

Molecular and Chemical Neuropathology Vol. 18, 1993