NeurobiologvofAging.Vol. 13, pp. 413-419, 1992 0197-4580192 $5.00 + .

00
Printed in the U.S.A.All fightsreserved. Copyright© 1992 PergamonPressLtd.

Evidence for Changes in the Alzheimer's

Disease Brain Cortical Membrane
Structure Mediated by Cholesterol

R. P R E S T O N M A S O N * i l l #~, W I L L I A M J. S H O E M A K E R t ¶ , L Y D I A S H A J E N K O # ,
T I M O T H Y E. C H A M B E R S ~ A N D L E O G. H E R B E T T E * : ~ § #

Departments of Radiology,* Psychiatry, t Medicine,$ Biochemistry, § The Travelers Center on Aging, II

Neurobiology Laboratory,¶ Biomolecular Structure Analysis Center,# University of Connecticut Health Center,
Farmington, C T 06030

R e c e i v e d 24 J u n e 1991; A c c e p t e d 20 N o v e m b e r 1991

MASON, R. P., W. J. SHOEMAKER, L. SHAJENKO, T. E. CHAMBERS AND L. G. HERBETTE. Evidencefor changesin

the Alzheimer~" disease brain cortical membrane structure mediated by cholesterol composition. NEUROBIOL AGING
13(3)413-419, 1992.--Small angle X-ray diffraction analysis of Alzheimer's disease (AD) lipid membranes extracted from
cortical gray matter showed significant, reproducible structure changes relative to age-matched control samples. Specifically,
there was an average 4 A reduction in the lipid bilayer width and significant changes in the membrane electron density profiles
of AD cortical samples. There were no significant structure differences in the membrane bilayers isolated from an unaffected
region (cerebellum) of the AD brain. Lipid and protein analysis of 6 AD and 6 age-matched controls showed that the phos-
pholipid:protein mass ratio was unchanged but that the unesterified cholesterol:phospholipid(C:PL) mole ratio decreased by
30% in the AD temporal gyrus relative to age-matched controls. By contrast, the C:PL mole ratio in the cerebellum did not
change significantly. X-ray diffraction analysis of a cholesterol enriched AD sample demonstrated a virtual restoration of the
normal membrane bilayer width and electron density profile, suggesting that the cholesterol deficit played a major role in the
AD lipid membrane structure perturbation. Alterations in the composition and structure of the membrane bilayer may play
an important role in the pathophysiologyof AD by altering the activity and catabolism of membrane-bound proteins, including
the/3-amyloid precursor protein.

X-ray diffraction Alzheimer's disease Neural membrane structure and composition

A L Z H E I M E R ' S disease (AD) is a neurodegenerative disorder
( 1) characterized by progressive loss of higher intellectual func-
tion in the absence o f focal neurologic defects. A characteristic
neuropathological lesion in the cerebral cortex of patients with
AD is the neuritic plaque composed ofa/3-amyloid (#A) core
surrounded by a ring of largely abnormal neuritic processes.
Although the number of neuritic plaques generally correlate
with the degree of dementia, the mechanism underlying the
formation of these structures remains unknown.
The ~A peptide is a small fragment of a large, membrane-
bound amyloid precursor protein (APP) that is broadly distrib-
uted in the cerebral cortex (7,21). The 38--41 residue f3A por-
tion of APP appears to have 11 to 14 ofits amino acids buried
in the membrane and 28 amino acids lying just outside of the
membrane (8). Typically, APP is cleaved outside of the mem-
brane, thereby preventing complete/3A formation (8). Neither
the primary sequence nor the level of expression of APP ap-
pears to be altered in AD (22,23). Two questions then arise: (a)
how do proteases gain access to the APP695 intra bilayer cleav-
age site to form/3A in A D and (b) how does the hydrophobic
~A peptide dissociate from the membrane to accumulate in ex-
tracellular compartments? Previous biochemical studies have
suggested that the answer to these questions may be related to
changes in the composition and/or structure o f the AD neu-
ronal membrane bilayer (26). These membrane changes in
turn may affect the cleavage and release of the 13A fragment
from neural membranes.
To explore possible biophysical and chemical changes in
AD brain membranes, we examined reconstituted lipid mem-
brane bilayers originating from AD, Parkinson's disease (PD)
and age-matched control brains" temporal gyrus and cerebel-
lum using small angle X-ray diffraction. We correlated these
structural data with measurements of the protein, phospho-
lipid, and unesterified cholesterol content of the brain mem-
branes. The results of this study showed significant, reproduc-
ible changes in the structure and composition of membranes

tRequests for reprints should be addressed to R. Preston Mason, Ph.D., Biomolecular Structure Analysis Center, University of Connecticut
Health Center, Farmington, CT 06030.

413
414
isolated from affected regions o f t b e AD brain. These changes
may play an important role in the pathophysiology of this dis-
ease.
MATERIALSAND METHOD
Samples of superior temporal gyrus and cerebellum were
snap frozen in liquid nitrogen ( - 7 0 ° C ) and coded by the staff
at the Institute for Biogerontology Research (Sun City, AZ).
The 8 controls had no history of dementia prior to death. The
diagnoses of the 9 AD patients were confirmed at autopsy. The
sex ratio (AD: 5 males; 4 females vs. PD: 5 males: 0 females vs.
C: 5 males; 4 females) were similar for the AD and control
groups. The postmortem delay (AD: 3.20 + 3.01 vs. PD: 3.25
+ 1.30 vs. C: 2.28 + .84 h) was not significantly different
among the 22 patients. For comparative studies, there were 6
AD, 5 PD, and 6 controls who had similar ages at the time of
death (AD: 74 + 6 years vs. Control: 74 + 4 years vs. PD: 76
+ 5 years, mean + SD). Stained sections taken from both tem-
poral gyrus and cerebellum were rated blindly by an experi-
enced neuropathologist for extent of cell loss and gliosis.
Isolation o/Brain Membranes
Gray matter from the superior temporal gyrus and cerebel-
lum was dissected from the frozen brain tissue. The material
was then weighed and homogenized in 30 volumes of 150 mM
Tris-HCI of ice cold buffer, pH 7.4 and homogenized for 30 s
using a Tekman homogenizer. The homogenate was subjected
to centrifugation at 20,000 × g for 25 rain at 4°C. The super-
natant was discarded and the final pellet was resuspended in 3
volumes of Tris buffer prior to lipid extraction.
Membrane protein was determined using bovine serum al-
bumin as standard (15). The phospholipid phosphorus content
of the samples was determined by previously described meth-
ods (3,5). Specifically, 50 ul of 72% percholoric acid was added
to 1-40 nmol ofphospholipid phosphorus in borosilicate tubes
and hydrolyzed for 1 h at 180°C in a sand block heater. The
following reagents were prepared for the phosphomolybdate
complex color development: solution A, 1.58% ascorbate; so-
lution B, 5.09 g a m m o n i u m molybdate, 33 ml concentrated
H2SO,, and 84 ml distilled water; solution C, 21.0 ml reagent
A, 2.0 ml reagent B. Reagent C (0.9 ml) is added to the cooled
percholoric acid digest, color developed at 80°C for 15 min,
and absorbance determined at 820 rim. A colorimetric assay
was used to measure total cholesterol in native membranes by
a modification of the cholesterol oxidase method (4,10). Cho-
lesterol was extracted from native membranes using a chloro-
form:methanol:water system (9). The efficiency of cholesterol
extraction was quantitated with [3HI cholesterol (specific activ-
ity of 93.8 Ci/mmol, NEN, Boston, MA) and determined to be
90%.
Electron Microscopy
Small drops (5ul) of the brain membrane homogenate ( 150
mM Tris, pH 7.4) were placed on Formvar--carbon-coated
specimen grids for one min. The grids were then blotted with
filter paper, and 5 ul of 1% phosphotungstic acid (pH 6.8) was
added. After 1 min., the grids were then blotted with filter
paper, and 5 ul of 1% phosphotungstic acid (pH 6.8) was added.
After 1 rain., the grids were blotted, air dried, and then exam-
ined and photographed in a Philips CM 10 tansmission elec-
tron microscope operated at 60 kV.
Stained sections taken from both temporal gyrus and cere-
bellum were rated blindly by an experienced neuropathologist
for extent of cell loss and gliosis. Half of the 6 AD cases ex-
MASON ET AL.
amined demonstrated a mild loss (approximately 10%) of neu-
rons and gliosis whereas only one case out of 6 age-matched
controls showed this same histological change. There was no
evidence of any histopathology in either the AD or control cer-
ebellum samples.
There were no significant changes in the choles-
terol:phospholipid (C:PL) mole ratios and membrane widths
within the control and AD subject groups related to these his-
tological differences, in addition, there was no correlation be-
tween these measurements and the cellular composition of all
twelve superior temporal gyrus samples, regardless of the clin-
ical diagnosis.
Transmission electron microscopy on negatively stained
brain preparations from both the cerebellum and superior tem-
poral gyrus (see Method section) showed a preponderance of
empty plasma membrane vesicles while myelin accounted for
less than 10% of the total membrane preparation. This micro-
scopic appearance was independent of either the brain location
for the tissue or clinical diagnosis.
Preparation ~?/Oriented Lipid Bilayers #~r X-ray D([lkaction
For X-ray diffraction studies, oriented lipid bilayers or mul-
tilayers were prepared from multilamellar vesicles (2,5).
Briefly, brain lipids dissolved in choloroform were dried down
as a thin film on the bottom ofglass 13 × 100 mm test tubes
by vigorous mixing with the vortex TM mixer. A specified vol-
ume of 0.5 mM HEPES. pH 7.3, 2mM NaCI was added to the
dried lipid preparation yielding a final lipid phosphorus con-
centration of 2 raM. The vesicles were then formed again by
vigorous mixing with the vortex TM mixer.
Preparation of multilayers for X-ray diffraction were pre-
pared as previously described (16). Briefly, 50 ul of the multi-
lamellar vesicle preparation (2 mM) were added to methyl-
methacrylate sedimentation cells containing an aluminum foil
substrate. The normal SW-28 rotor (Beckman Instruments,
Inc., Fullerton, CA) bucket caps were then replaced with "spin
dry caps" (containing a 100 um hole in the center) and the pel-
leted vesicles were spin dried at 65,000 g for 3 h under centri-
fuge vacuum. On completion of the spin-dry process, the sam-
ples were mounted on curved glass supports and rehydrated in
sealed brass canisters containing a saturated salt solution to de-
fine a specific relative humidity.
The salts used for controlling relative humidity for X-ray
samples were K2CO~, 45%; Mg(NO3)2, 55%, NaNO2, 66%; Di-
potassium salt of tartaric acid, 72%; (NH4)SO4, 81%; KBr, 84%;
ZnSO4, 93% and KNO3, 96%.
Small Angle X-ray DilJraction Data ('ollection and Reduction
Small angle X-ray diffraction studies were carried out by
aligning the samples at near grazing incidence with respect to
the X-ray beam (see Fig. 1). The radiation source was a colli-
mated, monochromatic X-ray beam (CuK,, x-ray, wavelength
= 1.54 A,) from an Elliot GX- 18 rotating anode X-ray gener-
ator (Marconi Avionics, Ltd., Borehamwood, UK). The ex-
perimental method utilized a single Franks' mirror defining a
line source where K,,~ and K~, are unresolved. Sample temper-
ature was controlled for the diffraction experiments.
The diffraction data from the multilamellar samples were
recorded on both Kodak DEF-5 film (Eastman Kodak Co.)
and a Braun position-sensitive I-D detector (Innovative Tech-
nologies, Inc., Newburyport, MA). Relative intensities for the
diffraction orders were obtained directly from digitized com-
puter plots of the detector data using an integration routine.
After repeated X-ray diffraction experiments with separate
.samples from control and AD tissue, the mean standard devi-
CHANGES IN AD BRAIN DUE TO CHOLESTEROL 415

I
I
I
Film
ot

Electronic deteclo¢
I
I
Olf fr~cted x-rayS
It.._

1 t
X-¢ay
Curved X.ray beam
I
I
I
- i
Electron densily otol,le

generalOr sa~'q~e
','5..
I
I
I
I

FIG. 1. Summary of the conditions for small angle X-ray diffraction experiments. The experimental method utilized a mono-
chromatic copper K° X-ray source (wavelength = 1.54 ~,). The brain lipid multibilayer sample was placed on a curved glass
mount such that the plane of the membranes were oriented around an axis perpendicular to the incoming X-ray beam at discrete
angles. The sample diffraction was collected on film or a one-dimensional position-sensitive electronic detector. The diffraction
order intensities were then integrated from the films with a densitometer or directly from digitized computer plots of the elec-
tronic detector data.

ation of the normalized integrated intensities, i, for each of the
diffraction orders h (h = 1-6) was determined (see Tables 4
and 5 later).
Data reduction (background and other geometrical correc-
tions) has been described previously (I 6). The lamellar inten-
sity functions from the samples collected with the electronic
detector were corrected by a factor ofs = 2sin 0/;~, the Lorentz
correction.
To phase the lamellar reflections for each experiment, a hy-
dration series or swelling analysis was carried out (17). At least
three sets of intensity data at different relative humidities each
with unique unit cell repeat distances were used to assign an
unambiguous phase combination to the experimentally ob-
tained structure factors.
The phased structure factors (presented in Tables 4 and 5)
were calculated directly from the experimentally obtained in-
tensity data corresponding to the measured D spaces. For fur-
ther analysis of the data, the continuous structure factor func-
tion for both AD and control data sets were sampled at the
same D space. The result of this more stringent analysis still
yielded a significant difference in the lipid bilayer electron den-
sity profile structures consistent with Fig. 2 and lends further
support to the observed structural differences in the AD versus
control brain lipid membrane bilayer as reported in this article.
RESULTS
Analysis of Membrane Protein, Cholesterol
and Phospholipid Content
The protein, phospholipid, and cholesterol content of brain
membranes from the gray matter of superior temporal gyrus
and cerebellum were measured and compared. In Table I, the
C:PL mole ratios for control and AD brains are summarized.
The results of the analysis demonstrated a significant decrease
(p < 0.01, Wilcoxon two-sample rank test, two-tailed) of 30%
in the C:PL mole ratio of AD temporal gyrus when compared
with control samples. By contrast, the C:PL mole ratio in the
cerebellum did not change significantly. Moreover, the mass
ratio of protein to phospholipid was not significantly perturbed
in both the temporal gyrus and cerebellum of the AD and con-
trol tissue. The C:PL mole ratio values of fresh AD temporal
gyrus (patient age at death, 73) prepared as synaptoneuro-
somes (SNM) was similar to those obtained from frozen tissue
and homogenized. As a further control, we analyzed brain
samples from 5 age-matched, Parkinson's disease (PD) brains.
The C:PL mole ratios from the temporal gyrus (0.57 + 0.10)
and cerebellum (0.48 + 0.06) of the Parkinson's disease brains
were not significantly different from the controls.
Correlation Between Age and Membrane Cholesterol Content
There was a significant correlation between age and the
C:PL mole ratio in the temporal gyrus but not the cerebellum
of both AD and control samples (Table 2). The average age of
the 8 control and 9 AD patients was 75 + 9 years. A Pearson
correlation coefficient (r) of 0.72 (p < 0.025) and 0.78 (p <
0.01 ) was calculated for 8 control and 9 AD samples, respec-
tively. The slope of the change (age versus C:PL mole ratio)
was 0.01 for both control and AD samples. In these same sam-
TABLE 1
C H O L E S T E R O L T O PHOSPHOLIPID M O L E R A T I O IN A D A N D
C O N T R O L B R A I N ( M E A N + SD)*
BrainArea Control(n = 6) AD(n = 6)
Sup. Temporal Gyrus 0.66 + 0.05° 0.46 _+ 0.08b
Cerebellum 0.45 + 0.08 0.50 _+ 0.08
a versus b, p < .01, Wilcoxon two sample rank test (two-tailed).
*Patient ages at time of death were: 74 + 4 years (control) and 74 +
6 years (AD).
The individual C:PL (age) values for the temporal gyrus were AD:
0.46(68); 0.52(69); 0.38(71 ); 0.34(73); 0.54(78); 0.53(85).
Control: 0.64(67); 0.62(73); 0.68(73): 0.72(76); 0.71(77); 0.59(77).
416 M A S O N ET AL.

TABLE 2 FABLE 4
CORRELATION COEFFICIENTS (PEARSON) BETWEtN AGE AND NORMALIZED STRUCTUREFACTORS FROM CONTROL AND AD
THE CHOLESTEROL TO PHOSPHOLIPID MOLE RATIO IN AD AND SUPERIOR TEMPORAl, GYRUS LIPID MEMBRANES (MEAN +_ SD)
CONTROL BRAIN
Control(n = 3) AD(n = 4)
Brain Area Control (n = 8) AD (n = 9)
1= .9727 z .0028 ~ I = -.9853 z .0024 h
Sup. Temporal G.vrus +0.72* ~ 0.78"l- 2 = .1615 4 .(X)49' 2 = -.0744 + .0144 ~
Cerebellum 0.27 - 0.07t 3 = +.0105 _~ .0148 ~ 3 = .t .0454 ± .0132 f
4 = .1591 + .0119 ~ 4 = -.1316 + .0124 h
The range in patients" ages was 20 years (55-77) tot control and 23
5 = +.0355 ~ .0049 5 = t.0272 ± .0265
years (68-91 ) for AD.
*p < .025, Student's one-tailed t test. i p < .01, Student's one-tailed 6 = -.(KI86 +_ .0185 6 = -.0338 + .0193
ttest.~n = 8 a versus b. p < .01, Student's two-tailed t test. c versus d, p < .01.
Student's two-tailed t test. e versus f. p < .05, Student's two-tailed t
tcst. g versus h, p < . 10, Student's two-tailed t test.
pies, the mass ratio o f p h o s p h o l i p i d to protein was similar for
control (0.70 + 0.05), A D (0.69 + 0.05), a n d P D patients
(0.78 + 0.06. n = 5) a n d did not correlate significantly with T h e difference in the m e a n s o f the D-space from the control
age. By contrast, the r values were - 0 . 2 7 (ns) a n d - 0 . 0 7 (ns) a n d A D samples in T a b l e 3 was 3.9 A. T h u s , the a d d i t i o n o f
in the c e r e b e l l u m o f control a n d A D samples, respectively. free cholesterol to A D m e m b r a n e s from the t e m p o r a l gyrus
was able to a c c o u n t for the difference in the D-space between
Small A ngh, X-ray Scattering o/Reconstituted Membranes control a n d A D m e m b r a n e s given in T a b l e 3.
M e m b r a n e multilayers o f reconstituted lipids from the su- T h e effect o f cholesterol o n m e m b r a n e structure was further
perior t e m p o r a l gyrus a n d c e r e b e l l u m gave clearly defined, re- d e m o n s t r a t e d in the electron density profiles. Diffraction pat-
producible, diffraction orders. T h e multilayer samples gave a terns were a n a l y z e d as described in the M e t h o d section. A rep-
single lattice diffraction p a t t e r n over a range of h u m i d i t i e s resentative X-ray diffraction p a t t e r n from A D a n d control
from 45% to 98% relative h u m i d i t y a n d low t e m p e r a t u r e s (5°C multilayer samples is s h o w n in Fig. 1. T h e square root o f these
to 15"C). For example, at 72% relative h u m i d i t y a n d 9"C, six intensities or s t r u c t u r e factors are s u m m a r i z e d in Tables 4 a n d
s h a r p lamellar diffraction orders were observed with a n average 5. T h e positive or negative signs c o r r e s p o n d to the phase o f the
unit cell repeat d i s t a n c e (the d i s t a n c e between the center o f t h e structure factor. C o n s i s t e n t with the D-space results, the nor-
interbilayer w a t e r space from o n e m e m b r a n e to the next re- malized s t r u c t u r e factors from control a n d A D m e m b r a n e
ferred to as the D-space for the m e m b r a n e ) o f 59 to 63 A, de- samples were significantly different from samples o b t a i n e d
p e n d i n g o n the b r a i n area from which the lipids were extracted from the s u p e r i o r t e m p o r a l gyrus b u t not the cerebellum.
(Table 3). Interestingly, the average variance in the 12 D-space C o m p a r i s o n o f m e m b r a n e electron density profiles calculated
values from the different regions o f the b r a i n was only 2% or from the structure factors revealed large differences in the
1.2 ,~. T h u s , the r e c o n s t i t u t e d lipids p r o d u c e d a m e m b r a n e m e m b r a n e bilayer width a n d h y d r o c a r b o n core electron den-
s t r u c t u r e that was highly preserved u n d e r b o t h n o r m a l a n d sity (Fig. 2).
pathological c o n d i t i o n s .
T h e r e were significant differences ( p < 0.01, S t u d e n t ' s two- I)ISCUSSION
tailed t test) in the m e m b r a n e bilayer width or D-space o f re- T h i s study describes structural changes in lipid m e m b r a n e s
c o n s t i t u t e d lipid m e m b r a n e s from A D t e m p o r a l gyrus but not r e c o n s t i t u t e d from the gray m a t t e r o f distinct brain regions
c e r e b e l l u m c o m p a r e d to c o n t r o l s (Table 3). T h e r e a p p e a r e d to from A D a n d control patients. Figure 2 a n d T a b l e 3 d e m o n -
be a c o r r e l a t i o n between t h e C : P L c o n t e n t o f the lipid extracts strate differences in the electron density profiles a n d d i m e n -
a n d D-space value that was i n d e p e n d e n t o f the b r a i n location sions o f A D versus control m e m b r a n e s . T h e b i o c h e m i c a l basis
o f the sample. T o test this correlation, free cholesterol was for these differences may, in part, be related to the significant
a d d e d to the organic p h a s e o f a n A D lipid extract from the su- decrease in the C : P L mole ratios m e a s u r e d in the lipid extracts.
perior t e m p o r a l gyrus prior to vesicle f o r m a t i o n . T h e a m o u n t F o r example, Tables 1 a n d 3 show a relationship between the
o f free cholesterol a d d e d b r o u g h t the A D C : P L mole ratio to a C : P L m o l e ratios a n d the m e m b r a n e D-space i n d e p e n d e n t o f
level similar to that o f the control. Multilayers were t h e n pre- the source o f the lipids. A d d i n g cholesterol to the A D superior
pared for X-ray diffraction. Analysis o f the diffraction p a t t e r n t e m p o r a l gyrus samples, to levels similar to control, was able
s h o w e d a n increase in the A D m e m b r a n e D-space by 3.9 A
from 58.3 A to 62.2 A following the a d d i t i o n o f cholesterol.
TABLE 5
NORMALIZED STRUCIURE FACTORS FROM CONTROL AND AD
TABLE 3 CEREBELLUM LIPID MEMBRANES(MEAN _+ SD)
MEMBRANE WIDTH OF CONTROL AND AD RECONSTITUTED LIPID Control(n = 3) AD(n = 3)
MEMBRANES(MEAN _+ SD)
I = -.9782 # .0065 1= -.9691 ± .0091
Brain Area Control (n = 3) A D (n = 4)
2 = -.1114 ± .0315 2 = -.1415 ± .0330
Sup. Temporal Gyrus 62.6 ± 1.22A" 58.7 ± 0.4A b 3 = ~.0342 ± .0119 3 = +.0088 ± .0124
Cerebellum 59.1 ± 0.9A 58.4 ± 1.8Ac 4 = -.1517 ± .0244 4 = -.1848 ± .0238
5 = +.0531 ± .0178 5 = +.0577 ± .0018
a versus b, p < .01, Student's two-tailed t test. 6 = -.0433 ± .0185 6 = -.0477 ± .(X)56
¢n=3.
CHANGES IN AD BRAIN DUE TO CHOLESTEROL 417

1.0

I AD t*
~ , S A M P L E - - - ~ ~
~ ~ I~-CONTROL
~ ,~ SAMPLE
0.0

£D

,,i

-I.0

) . . . . , . . . . i . . . . . . . . , . . . . , - - - ,
-30.0 -20.0 -I0.0 010 t0.0 20.0 30.O
X (ANGSTROMS)

FIG. 2. Comparison of a typical electron density profile from the superior temporal gyms
of an AD (dashed curve) and age-matched control (solid curve) subject. The two peaks of
higher electron density correspond to the electron-dense, phosphate headgroup and the elec-
tron density minima at the center of the figure corresponds to the terminal methyl groups
at the center of the membrane bilayer. The lipid bilayer width (i.e., D-space) of the AD and
control membranes were 57.8A and 62.2A, respectively. The C:PL mole ratios for these
samples were 0.38:1 (AD) and 0.64:1 (control).

to correct for the D-space difference (see Fig. 3). This is consis-
tent with the effect ofcholesterol on ordering the acyl chain re-
gion (i.e., decreasing trans-gauche isomerizations) in the mem-
brane bilayer hydrocarbon core resulting in an increase in the
overall membrane width.
The C:PL mole ratio is the main physiological determinant
of membrane fluidity. Increases in membrane fluidity have
been related to a decrease in ligand binding to B-adrenergic, se-
rotonin, and opiate receptors in mouse brain membranes
( 12,13,14). In our laboratory, we have observed that changes
in the C:PL mole ratio affect the partitioning of certain drug
molecules into the membrane. The partition coefficient for
nimodipine, a 1,4-dihydropyridine Ca +2 channel blocker cur-
rently being tested for the treatment of AD, increased three-
fold as the C:PL mole ratio decreased by 10 mol% (unpub-
lished data). Nimodipine localizes near the hydrocarbon core/
water interface, a region which overlaps the location ofcholes-
terol (11). Thus, changes in membrane composition, particu-
larly cholesterol, may strongly affect both endogenous (e.g.,
neurotransmitters, hormones, second messengers) and exoge-
nous (e.g., drug) regulation of brain cell function.
This study also noted a significant correlation between
membrane C:PL mole ratios and age in the superior temporal
gyrus (Table 2). The rate of membrane cholesterol increase was
very similar in the AD and control cortical tissue; but the ab-
solute levels were significantly lower in the AD cases. Increases
in the C:PL mole ratio have also been reported in human
whole brain extracts (19), platelets (6), and lymphocytes (18)
as a function of age. There was also a significant correlation be-
tween the cholesterol:protein mole ratio but not the phospho-
lipid:protein ratio, in the superior temporal gyrus as a function
ofage. By contrast, there were no significant changes in the cer-
ebellum C:PL mole ratio or cholesterol:protein mass ratios as
a function ofage.
Finally, changes in the composition and structure ofthe AD
brain membrane may affect the cleavage of the amyloid pre-
cursor protein (APP), resulting in elevated levels of/~A. Mu-
tational analysis studies suggest that the cleavage site for ~-am-
yloid is buried in the membrane (8,22). However, alterations
in the AD membrane structure (e.g., a decrease in the D-space)
may alter the tertiary structure of APP and increase the prob-
ability for proteases togain access to this cleavage site. For ex-
ample, the average 4 A decrease in the membrane bilayer D-
space could uncover amino acid residues otherwise inaccesible
to certain extracellular proteases. Figure 4 shows a molecular
model based on the X-ray diffraction data to illustrate how al-
terations in membrane structure may result in exposure of the
BA cleavage site. The putative HA site indicated in the figure as
an " X " is buried within the lipid bilayer under normal choles-
terol conditions (Frame A) but is exposed in AD brain mem-
418 M A S O N ET AL.

>..
[--,

z
f.~
?-~
z O0
o
~.,

"! .0

-~;o ~;.0 -t;., o?o ,*'.0 z,., 3,0

X (ANGSTROHS)

FIG. 3. (A) Comparison of lipid bilayer electron density profiles from cholesterol enriched A D (dashed curve) and control (solid curve) superior
temporal gyrus samples at 93% relative humidity and q*C. Following cholesterol enrichment of the AD sample to a C:PL mole ratio of ~0.7: I,
the D-space increased from 58.3~, to 62.2.,~. The control C:PL ratio was unchanged (D-space = 62.40A). (B) The profiles were correlated with
a space-filled model of a membrane bilayer showing the location of the cholesterol molecules.

b r a n e s ( F r a m e B) w h o s e lower cholesterol c o n t e n t m a y pro-
duce the decreased m e m b r a n e bilayer width. F u r t h e r structure
analysis o n the A P P p r o t e i n will d e t e r m i n e where the/3A cleav-
age sites are with respect to the m e m b r a n e in A D tissue.
ACKNOWLEDGEMENTS
We thank R. W. Besdinc, MD, for continual discussions relating
these data to the neuropathology of AD. We would also like to thank
Mr. H. Grazioso for his excellent technical assistance and M. Grunnet,
MD, Department of Pathology. for analyzing the histology of the .sam-
pies. Joseph Rogers, Ph.D., of the Institute tbr Biogerontology Re-
search (Sun City, AZ) provided frozen samples of human brain tissue
for these studies. C. Hamann, MD. from St. Francis Medical Center
(Hartford, CT) assisted in providing fresh brain material for this pro-
ject.
This work was supported by the American Health Assistance
Foundation and the John A. Hartford Foundation. R. P. M. is a John
A. Hartford Foundation Gerontology Scholar. LGH is an Established
Investigator of the American Heart Association. The Biomolecular
Structure Analysis Center acknowledges support from R JR Nabisco
Inc., the Patterson Trust Foundation. and the State of Connecticut's
High -lechnology Programs.

B
Cleavage Site Cleavage Site "~
Buried ~--Amyloid Fragment Exposed ~ ~_.~^myloid Fragment

ALZHEIMER'5
CONTROL MEMBRANE MEMBRANE

Average D~Spac e Average D-Space

62.6~ 5s.7~.

Cytoplasmic Side
Cytopltsmle Side

FIG. 4. Molecular model based on the X-ray diffraction data to rationalize how changes in AD membrane structure and
composition may result in exposure of the ~A cleavage site. The ~A cleavage site indicated by an X is buried within the lipid
bilayer under normal cholesterol conditions (control membrane), (A) but is exposed in AD brain membranes (B) where a
depressed cholesterol content results in a decreased membrane bilayer width.
C H A N G E S IN A D B R A I N D U E T O C H O L E S T E R O L
REFERENCES
I. Alzheimer, A. Ueber eine eigenartige Erkrankung der Hirnrinde.
Allg. Z. Psychiatr. Psych-Gerichtl. Med. 64:146-148; 1907.
2. Bangham, A. D.; Standish, M. M.; Watkins, J. C. Diffusion of uni-
valent ions across the lamellae of swollen phospholipids. J. Mol.
Biol. 13:238; 1965.
3. Chen, P. S., Jr.; Toribara, T. Y.; Warner, H. Microdetermination
of phosphorus. Anal. Chem. 28:1756-1758; 1956.
4. Chester, D. W.; Tortelotte, M. E.; Molchior D. L.: Romano, A. H.
The influence of saturated fatty acid modulation of bilayer phys-
ical state on cellular and membrane structure and function.
860:383-398; 1986.
5. Chester, D. W.; Herbette, L. G.; Mason R. P.: Joslyn, A. F.; Trig-
gle, D. J.; Koppel, D. E. Diffusion of dihydropyridine calcium
channel antagonists in cardiac sarcolemmal lipid multibilayers.
Biophys. J. 52:102 I- 1030; 1987.
6. Cohen, B. M.; Zubenko, G. S. Aging and the biophysical charac-
teristics of cell membranes. Life Sci. 34:1403-1409; 1985.
7. Dyrks, T.; Weidemann, A.; Multhaup, G.; Salbaum. J. M.; Le-
maire, H-G.: Kang, J.; Buller-Hill, B.; Masters C. L.; Beyreuther,
K. Identification, transmembrane orientation and biogenesis of
the amyloid A4 precursor of AIzheimer's disease. EMBO J. 7:949-
957; 1988.
8. Esch, F. S.; Keim, P. S.; Beattie, E. C.; Blacher, R. W.: Culwell, A.
R.; Oltersdorf, T.; McClure, D.; Ward, P. J. Cleavage ofamyloid
peptide during constitutive processing of its precursor. Science.
248:1122-1124; 1990.
9. Folch, J.; Lees, M.; Sloane Stanley, G. A. A simple method for the
isolation and purification of total lipides from animal tissues. J.
Biol. Chem. 226:497-509: 1957.
10. Heider, J. G.: Boyett, R. L. The picomole determination of free
and total cholesterol in cells in culture. J. of Lipid Res. 19:514-
518; 1978.
I I. Herbette, L. G.; Chester, D. W.; Rhodes, D. G. Structural analysis
ofdrug molecules in biological membranes. Biophys. J. 49,91-94;
1986.
12. Heron, D. S.; Hershkowitz, M.; Shinitzky M.; Samuel, D. 1980.
The lipid fluidity of synaptic membranes and the binding of se-
rotonin and opiate ligands. In Littauer, U.Z.; Dudai, Y.; Silman,
I.; Teichberg V.I.; Vogel Z. Eds. Neurotransmitters and their re-
ceptors. New York: Wiley: 1980: 125-138.
13. Heron, D. S.: Shinitzky, M.; Hershkowitz M.; Samuel, D. Lipid
fluidity markedly modulates the binding of serotonin to mouse
brain membranes. Proc. Natl. Acad. SCi. 77:7463-7467; 1980.
419
14. Hershkowitz, M.; Heron, D.; Samuel D.; Shinitzky, M. The mod-
ulation of protein phosphorylation and receptor binding in syn-
aptic membranes by changes in lipid fluidity: Implications for
aging. Prog. Brain Res. 56:419-434; 1982.
15. Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall. R. J. Pro-
tein determination with the phenol reagent. J. Biol. Chem.
193:265-275; 1951.
16. Mason, R. P.; Gonye, G. E.; Chester D. W.; Herbette. L. G. Par-
titioning and location of bay K 8644, 1,4-dihydropyridine Ca "2
channel agonist, in model and biological lipid membranes. Bio-
phys. J., 55:769-778; 1989.
17. Moody, M.F. X-ray diffraction pattern of nerve myelin: A method
tbr determining the phases. Science. 142:1173- I 174; 1963.
18. Rivnay, B.; Bergman, S.; Shinitzky M.; Gloverson. A. Correla-
lions between membrane viscosity, serum cholesterol, lympho-
cyte activation and aging in man. Mech. Aging Dev. 12:119-126:
1980.
19. Rouser, G.: Kitchensky, G.; Yamamoto A.; Baxter, C. F. Lipids
in the nervous system of different species as a function of age. Adv.
Lipid Res. 10:261-360; 1972.
20. Khachaturian, Z. S. Diagnosis of AIzheimer's disease. Arch. Neu-
rol. 42:1097-1105; 1985.
21. Shivers, B. D.; Hilbich, C.; Multhaup, G.; Salbaum, M.; Be-
yreuther, K.; Seeburg, P. H. Alzheimer's disease amyloidogenic
glycoprotein: expression pattern in rat suggests a role in cell con-
tact. EMBO J. 7:1365-1370; 1988.
22. Sisodia, S. S.; Koo, E. H.; Beyreuther, K.; Unterbeck, A.; Price, D.
L. Evidence that ~A Protein in Alzheimer's disease is not derived
by normal processing. Science. 248:492-495; 1990.
23. Tanaka, S.; Nakamura, S.; Ueda, K.; Kameyama, M.; Shiojiri, S.;
Takahashi, Y.; Kitaguchi, N.; Ito. H. Three types ofamyloid pro-
tein precursor mRNA in human brain: Their differential expres-
sion in alzheimer's disease. Biochem. Biophys. Res. Commun.
157:472-479:1988.
24. Weidemann, A.; Konig, G.: Bunke, D.; Fischer, P.; Salbaum, J.
M.; Masters C. L.; Beyreuther, K. Identification, biogenesis, and
localization of precursors of Alzheimer's disease A4 amyloid pro-
tein. Cell. 57:115-126: 1989.
25. Yankner, B. A.; Duffy L. K.: Kirschner. D. A. Neurotrophic and
neurotoxie effects of amyloid B protein: Reversal by tachykinin
neuropeptides. Science. 250:279-282; 1990.
26. Pettegrew, J. W. Molecular insights into Alzheimer's disease. An-
nals NY Acad. Sci. 568:5-28; 1989.