Lipid alterations in lipid rafts from Alzheimer's disease human brain cortex

Abstract

Lipid rafts are membrane microdomains intimately associated with cell signaling. These biochemical microstructures are characterized by their high contents of sphingolipids, cholesterol and saturated fatty acids and a reduced content of polyunsaturated fatty acids (PUFA). Here, we have purified lipid rafts of human frontal brain cortex from normal and Alzheimer's disease (AD) and characterized their biochemical lipid composition. The results revealed that lipid rafts from AD brains exhibit aberrant lipid profiles compared to healthy brains. In particular, lipid rafts from AD brains displayed abnormally low levels of n-3 long chain polyunsaturated fatty acids (LCPUFA, mainly 22:6n-3, docosahexaenoic acid) and monoenes (mainly 18:1n-9, oleic acid), as well as reduced unsaturation and peroxidability indexes. Also, multiple relationships between phospholipids and fatty acids were altered in AD lipid rafts. Importantly, no changes were observed in the mole percentage of lipid classes and fatty acids in rafts from normal brains throughout the lifespan (24-85 years). These indications point to the existence of homeostatic mechanisms preserving lipid raft status in normal frontal cortex. The disruption of such mechanisms in AD brains leads to a considerable increase in lipid raft order and viscosity, which may explain the alterations in lipid raft signaling observed in AD.

Full text

Journal of Alzheimer’s Disease 19 (2010) 489–502 489
DOI 10.3233/JAD-2010-1242
IOS Press
Lipid Alterations in Lipid Rafts from
Alzheimer’s Disease Human Brain Cortex
Virginia Mart´
ına,b, Noem´
ı Fabeloa,b, Gabriel Santperec, Berta Puigc, Raquel Mar´
ına,d, Isidre Ferrercand
Mario D´
ıaza,b,∗
aInstituto de Tecnolog´
ıas Biom´
edicas, Universidad de La Laguna, Tenerife, Spain
bDepartamento de Biolog´
ıa Animal, Universidad de La Laguna, Tenerife, Spain
cInstitut Neuropatologia,Servei Anatomia Patologica, Hospital Universitari de Bellvitge, Universitat de
Barcelona, Hospitalet de Llobregat, CIBERNED, Spain
dDepartamento de Fisiolog´
ıa, Universidad de La Laguna, Tenerife, Spain
Accepted 31 July 2009
Abstract. Lipid rafts are membrane microdomains intimately associated with cell signaling. These biochemical microstructures
are characterized by their high contents of sphingolipids, cholesterol and saturated fatty acids and a reduced content of polyun-
saturated fatty acids (PUFA). Here, we have purified lipid rafts of human frontal brain cortex from normal and Alzheimer’s
disease (AD) and characterized their biochemical lipid composition. The results revealed that lipid rafts from AD brains exhibit
aberrant lipid profiles compared to healthy brains. In particular, lipid rafts from AD brains displayed abnormally low levels of
n-3 long chain polyunsaturated fatty acids (LCPUFA, mainly 22:6n-3, docosahexaenoic acid) and monoenes (mainly 18:1n-9,
oleic acid), as well as reduced unsaturation and peroxidability indexes. Also, multiple relationships between phospholipids and
fatty acids were altered in AD lipid rafts. Importantly, no changes were observed in the mole percentage of lipid classes and fatty
acids in rafts from normal brains throughout the lifespan (24–85 years). These indications point to the existence of homeostatic
mechanisms preserving lipid raft status in normal frontal cortex. The disruption of such mechanisms in AD brains leads to a
considerable increase in lipid raft order and viscosity, which may explain the alterations in lipid raft signaling observed in AD.
Keywords: Alzheimer’s disease, docosahexaenoic acid, human brain cortex, lipid rafts, membrane phospholipids, polyunsaturated
fatty acids
INTRODUCTION
Lipid rafts have been defined as cholesterol and sph-
ingolipid enriched membrane microdomains resistant
to solubilization by non-ionic detergents at low tem-
peratures. They may serve as platforms for intracel-
lular cell signaling by promoting protein-protein and
protein-lipid interactions [1,2]
Alterations in the molecular composition and cell
distribution of lipid rafts might have implications in
∗Correspondingauthor:Dr.MarioD´ıaz,U.D.I.Fisiolog´ıaAni-
mal,DepartamentodeBiolog´ıaAnimal,UniversidaddeLaLaguna,
38206 Tenerife, Spain. Tel.: +34 922318343; Fax: +34 922318342;
E-mail: madiaz@ull.es.
pathological events. There is increasing evidence that
lipid rafts may be targets of neurodegenerative diseases
such as Alzheimer’s disease (AD) [3,4]. AD, the most
common form of dementia, is a progressive degenera-
tive disease of the brain characterizedclinically by pro-
gressive loss of memory and cognitive function. Neu-
ropathologically,AD is characterized by senile plaques
and neurofibrillary tangles [5]. The main component
of senile plaques is amyloid, which consist mainly of
aggregated variants of amyloid β-protein (Aβ), a fam-
ily of 39–42 residue peptides formed by two sequen-
tial enzyme cleavage of the amyloid-βprotein precur-
sor (AβPP). AβPP is initially cleaved by β-secretase
followed by the subsequent intramembrane proteoly-
ses of the membrane bound C-terminal fragment cat-
ISSN 1387-2877/10/$27.50 2010 – IOS Press and the authors. All rights reserved

490 V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex
alyzed by γ-secretase. Lipid rafts may play an impor-
tant role in proteolytic processing and regulation of
AβPP cleavage, and recent reports have shown that
AβPPitselfis expressed inlipidrafts[6–8]. Inaddition,
AD-related components, such as the AβPP N-terminal
fragment [9], the Aβ-bearing C-terminal fragment pro-
ducedbyβ-secretase[10], α-secretase [11,12],BACE1
(β-secretase) [13,14], and PSEN1 (γ-secretase) [10,
15–17], as well as apolipoprotein E (ApoE) and tau
have been identified in lipid rafts of cultured cells and
mammalian brains.
Long chain polyunsaturated fatty acids (LCPUFA),
mainly docosahexaenoic acid (DHA; 22:6n-3), are par-
ticularly enriched in cell membrane phospholipids, es-
pecially in neural tissues [18,19]. Several studies have
shownthatthese fatty acidsare important for theproper
developmentandphysiologyofneuronal cells andtheir
deficiency has been associated with AD [20–25]. Also,
LCPUFAs such as DHA have the capacity to influence
plasma membrane organization and activity by modu-
lating the lipid composition and functionality of lipid
raft domains [26–28,30,31]. These observations sug-
gest that lipid rafts are likely molecular targets through
whichlongchain n-3 PUFAmodulatediversebiochem-
ical activities, and reduce the incidence and severity of
human diseases.
Considering the importance of lipid raft signaling in
the pathogenesis of AD and that these specific mem-
brane domains are putative targets for pharmacological
approaches in the preventionof the disease, the aim of
this study was to examine the lipid classes and fatty
acid composition of lipid rafts from frontal cortex of
human brains, one of the main areas affectedin AD, to
explore possible lipid profile alterations that could cor-
relatewith thisneuropathology. This isthefirstdetailed
biochemical study on lipid raft fatty acid composition
in human brains.
MATERIALS AND METHODS
Human brain tissue
Brain tissues were obtained from the Institute of
Neuropathology Brain Bank (Hospital Universitari de
Bellvitge, Spain) following the guidelines of the local
ethics committee. 10 patients had suffered from severe
(Global deterioration scale) dementia of Alzheimer
type. 20 cases were neurologically normal. The post-
mortem delays were between 3 and 18 h. Frontal cor-
tex tissue (cortex area VIII) were used for the isolation
of lipid rafts. Cases were divided into three categories:
AD group (patients with Alzheimer’s disease and av-
erage age 81.2 ±2.48 years), C >60 group (an age-
matched control obtained from subjects showing no le-
sions and average age 74.0 ±2.16 years old), and C
<60 group (samples from subjects showing no lesions
and average age 42.4 ±2.41 years). A summary of the
main clinical and neuropathologicalaspects of allcases
examined is shownin Table1.
Cases with and without clinical neurological disease
were processed in the same way following the same
sampling and staining protocols. At autopsy, half of
each brain was fixed in 10% buffered formalin, while
the other half was cut in coronal sections 1 cm thick,
frozen on dry ice, and stored at −80◦C until use. In
addition, samples of the frontal cortex werefixed in 4%
paraformaldehyde in phosphate buffer for 24 h, cry-
oprotected in 30% sucrose and frozen at −80◦C. The
neuropathological study was carried out on de-waxed
4µm-thick paraffin sections of the frontal cortex (area
8). The sections were stained with haematoxylin and
eosin, Kl¨
uver Barrera, and, for immunohistochemistry
to glial fibrillary acidic protein, CD68 and Licoper-
sicum esculentum lectin for microglia, Aβ-amyloid,
pan-tau, AT8 tau, phosphorylation-specifictau Thr181,
Ser202, Ser214, Ser262, Ser396 and Ser422, and αB-
crystallin, α-synuclein and ubiquitin. Following neu-
ropathological examination, 10 cases were categorized
as AD stages V/VIC of Braak et al. [32], modified
for paraffin sections [33]. The main clinical and neu-
ropathological findings in the present series are sum-
marized in Table 1.
Isolation of lipid rafts, non-raft fractions and
microsomes
Samples of frontal cortex grey matter were carefully
dissected out to avoid contaminationwith white matter.
Lipid raft fractions were isolated following Mukherjee
et al. [34] with slight modifications. Briefly, 0.1 g of
frontal cortex was homogenized in 8 volumes of buffer
A (50 mM Tris-HCl, pH 8.0, 10 mM MgCl2,20mM
NaF, 1 mM Na3VO4,5mMβ-mercaptoethanol, 1mM
PMSF) and a cocktail of proteases inhibitors (Roche
Diagnostics, Barcelona, Spain) containing 1% Triton
X-100 and 5% glycerol in aglass homogenizer grinder.
All steps in the protocol were performed on ice or in a
cold room at 4◦C. Tissue was then centrifuged at 500
xgfor 5 min and the supernatant was collected and
mixed in an orbital rotor for 1 h at 4◦C. About 800 µl
of sample was mixed with an equal volume of 80%

V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex 491
Table 1
Summary of cases
Case Age Gender Postmortem Neuropathological Braak stage
(years) delay diagnosis
Control samples with age <60 (group C <60)
1 38 M 18h NL 0
2 49 M 7h35min NL 0
3 40 M 9h15min NL 0
4 39 M 3h30min NL 0
5 47 M 4h55min NL 0
6 45 F 14h40min NL 0
7 24 F 6h NL 0
8 46 F 14h5min NL 0
9 49 F 7h NL 0
10 47 F 9h35min NL 0
Control samples with age >60 (group C >60)
11 79 M 7h NL 0
12 85 M 5h45min NL 0
13 70 M 13h NL 0
14 78 M 2h15min NL 0
15 71 M 12h NL 0
16 82 F 11h NL 0
17 75 F 3h NL 0
18 66 F 8h NL 0
19 69 F 2h30min NL 0
20 65 F 4h NL 0
AD samples (group AD)
21 69 M 6h AD VC
22 93 M 7h20min AD VC
23 79 M 7h25min AD+AmA VC
24 73 M 2h30min AD VIC
25 86 M 4h15min AD+AmA VC
26 86 F 10h AD VC
27 83 F 5h AD VC
28 82 F 1h45min AD+AmA VC
29 72 F 9h30min AD VC
30 89 F 4h AD VC
M: male; F: female; NL: no lesions; AD: Alzheimer disease; AmA: amyloid
angiopathy. V/VI: refers to Braak and Braak stages of AD-related changes;
V/VI: neurofibrillary tangle distribution in the neocortex; C: large numbers of
senile plaques in the neocortex.
sucrose in buffer A and overlayed with 7.5 ml of a 35%
sucrose solution and 2.7 ml of a 15% sucrose solu-
tion in buffer A, in 10 ml ultracentrifuge tubes (Ultra-
clear, Beckman). Sucrose gradients were centrifuged
at 150,000 x gfor 18 h at 4◦C in a Beckman SW41Ti
rotor. Fractions of 2 ml were collected from the top
to the bottom and the final pellet, corresponding to the
precipitated detergent soluble fractions, i.e., non-rafts
fractions, were collected and resuspended in 200 µlof
buffer A and frozen until analyses.
Microsomal fractions of grey matter frontal cor-
tex samples were obtained by homogenization in RSB
buffer (10 mMTris-HCl, pH 8.0, 20 mM NaCl, 25 mM
EDTA and completeproteasesinhibitorcocktail),using
a Teflon-glass homogenizer grinder. The whole pro-
cedure was carried out at 4◦C. The tissue homogenate
was first centrifuged at 900 x gfor 15 min and the su-
pernatant was centrifuged at 10,000 x gfor 15 min to
sediment the mitochondrial fraction. A second super-
natant was collected, and centrifuged at 100,000 x g
at 4◦C in a Beckman SW55Ti rotor following standard
protocols to sediment microsomal fractions.
For the protein characterization of lipid rafts, non-
rafts and microsomes, samples were resuspended in
SDS loading buffer (625 mM Tris-HCl pH 6.8; 1%
SDS, 10% glycerol, 5% β-mercaptoethanol; 0.01%
bromophenol blue), boiled at 95◦C for 5 min, and
proceeded for SDS-PAGE and Western blotting. Sam-
ples were probed for the mouse anti-flotillin-1 anti-
body (610820, BD Biosciences,) and the monoclonal
anti-PrP antibody (S2022, Clone 3F4, Dako), both at
1:1000, the rabbit polyclonal anti-caveolin-1 (sc-894,
SantaCruzBiotech.,diluted1:200),andtherabbitpoly-
clonal anti-AβPP (ab17467, Abcam, diluted 1:500) to

492 V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex
identify raft-enriched fractions. The mouse monoclon-
al antibodies against the non-raft membrane and mi-
crosomal proteins α1subunit of the Na+/K+ATPase
(05-369,Upstate)and clathrin (C1860,Sigma Aldrich),
both diluted at 1:1000, were used as controls of lipid
rafts purity. The mouse anti-SOD1 antibody, raised
against a prokaryotic recombinantfusion protein corre-
sponding to the N-terminaldomains I toV of the Cu/Zn
superoxidedismutase (SOD-1) molecule(NCL-SOD1,
Novocastra Laboratories, diluted 1:1600), was used as
a cytosolic marker.
Lipid analyses
Total lipids from lipid rafts and non-raft fractions
were extracted with chloroform/methanol (2:1 v/v)
containing 0.01% of butylated hydroxytoluene (BHT)
asantioxidant[35]. Lipidclasses were separated froma
fraction of total lipid by one-dimensionaldouble devel-
opment high performance thin layer chromatography
(HPTLC), and were quantified by densitometry [36].
Equal amounts of total lipids (30 µg) were used in all
analyses.
Lipids from lipid rafts and non-raft fractions were
subjected to acid-catalyzed transmethylation with 1%
sulfuric acid (v/v) in methanol. The resultant fatty
acid methyl esters (FAME) were purified by thin lay-
er chromatography (TLC) [35]. FAME were separat-
ed and quantified by using a Shimadzu GC-14A gas
chromatograph equipped with a flame ionization de-
tector (250◦C) and a fused silica capillary column
SupelcowaxTM10 (30 m ×0.32 mm I.D.). Individual
FAME were identified by referring to authentic stan-
dards.
Unsaturation index was calculated as mini, where
miis the mole percentage and niis the number of
carbon-carbon double bonds of the fatty acid. The per-
oxidability indexwas calculated following Cosgrove et
al.[37]asmonoenoic*0,025+dienoic+trienoic
*2+tetraenoic*3+pentaenoic*6+hexaenoic*8.
Statistical analyses
Comparison between groups was assessed either by
one-way ANOVA followed by Tukey’s post-hoc test
orKruskal-WallisfollowedbyGames-Howellpost-hoc
test depending on the homocedasticity and normality
of experimental data. Data from univariate and bivari-
ate statistics are expressed as mean ±SEM. Statistical
significance is indicated in the figures and tables from
p<0.05. Pearson correlation coefficients were used
to express bivariate relationships between independent
variables (lipid parameters and age on one side, and
fatty acids and lipid classes on the other). Multivari-
ate statistics were performed using multivariate analy-
ses of variance (MANOVA) followed by discriminant
function analysis. Predictive variables were chosen ac-
cording to the number of cases in each group to ful-
fill the assumptions of discriminant analysis [38]. Da-
ta were arcsin transformed (percent lipid content) or
log-transformed (age) in order to attain the assumption
of normality. Individual canonical scores of each case
and centroids of each group were calculated using the
SPSS statistical package (version 15), and then plotted
in order to predict which group a particular individual
case belonged (Fig. 4).
RESULTS
Characterization of lipid rafts
Purity of isolated lipid rafts was confirmed by
demonstrating the enrichment of flotillin-1, the pro-
totipical raft marker protein in the corresponding frac-
tion (Fig. 1A). Similarly, the prion-relatedprotein PrP,
a protein known to be localized in neuronallipid rafts,
appeared concentrated in fraction 1 while the Cu/Zn
superoxide dismutase 1 (SOD-1), a cytosolic marker,
was completely excluded from this particular fraction.
To confirm whether lipid rafts (fraction 1) were
free of non-raft material, we performed additional im-
munoblotting experiments to compare the presence of
raft and non-raft membrane associated proteins be-
tween microsomes (M), non-raft fraction (NR), and
lipid raft fraction 1 (LR) (Fig. 1B). Results demonstrat-
ed the presence of flotillin-1 and caveolin-1, another
hallmark of lipid rafts, in fraction 1 and microsomes
that were not present in non-raft fractions. A similar
pattern was obtained for AβPP, known to be mainly
localized in neuronal lipid rafts. In contrast, the in-
tegral membrane Na+/K+ATPase α1subunit, and the
membrane vesicle coated protein clathrin, which are
non-raft and microsome associated, were not detected
in lipid rafts.
In order to demonstrate the purity of lipid rafts, we
performed additional analyses on the lipid profile of
lipid rafts and non-raft fractions. The results summa-
rized in Fig. 1C demonstrated that isolated lipid raft
fraction 1 exhibit significantly higher amounts of sph-
ingomyelin (SM, ∼11%), cholesterol (CHO, ∼35%),
and saturated fatty acids (∼50%) compared to non-

V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex 493
Fig. 1. (A) Western blot characterization of frontal cortex lipid rafts. Illustrated corresponds to a subject from group C>60. Lipid raft resident
proteins as flotillin-1 and PrP are found mainly in fractions 1 and 2 whereas cytosolic protein SOD-1 is found in soluble fractions 5 and 6. (B)
Comparison of protein markers present in microsomal fraction (M), non-raft fractions (NR) and lipid rafts (LR) extracted from a subject from
group C >60. Equal amounts of total protein were used for M, NR and LR samples. Immunoblotted proteins were the lipid raft hallmarkers
caveolin-1 (cav-1) and flotillin-1, amyloid-βprotein precursor (AβPP), also known as lipid raft resident, and non-raft membrane associated
proteins Na+/K+ATPase α1subunit (ATPase) and clathrin. (C) Summary of lipid analyses of lipid rafts and non-rafts fractions from group C
>60 (n=10). **, * statistically different from non-raft with p<0.01 and p<0.05, respectively. SM: sphingomyelin, CHO: cholesterol, PC:
phosphatidylcholine, PS: phosphatidylserine, PI: phosphatidylinositol, PG: phosphatidylglycerol, PE: phosphatidylethanolamine.
raft fractions. Within saturates, palmitic (16:0) and
stearic (18:0) fatty acids accounted for more than 90%
of saturates (see Table 3). In contrast, the contents
of phosphatidylcholine (PC), phosphatidylserine (PS),
phosphatidylinositol (PI), phosphatidylglycerol (PG),
and phosphatidylethanolamine (PE) were significant-
ly smaller in lipid rafts compared to non-raft frac-
tions(Fig. 1C). Taken together, these experimentaldata
demonstrate that the protein and lipid profiles of frac-
tion 1 correspond to the expected features of highly
purified lipid rafts.
Detailed analyses of fatty acids revealed that lipid
rafts exhibited significant contents of monoene fatty
acids(15–18%)andn-3 long chain polyunsaturatedfat-
ty acids (n-3 LCPUFA) (5–7%), specifically oleic acid
(18:1n-9) and docosahexaenoic acid (DHA; 22:6n-3),
while eicosapentaenoic acid (EPA, 20:5n-3) levels (the
other essential n-3 LCPUFA) were negligible (Table 3).
Arachidonic acid (20:4n-6) was also present in signifi-
cant amounts in lipid rafts from all groups. These fatty
acids are preferentially esterified on PE, PS, and PI.
Accordingly,our analyses revealed significant levels of
PE, PS, and PI, representing nearly 20%, 6.5%, and
2.5%, respectively, of the total phospholipids present
in the lipid rafts of the three groups (Table 2). Thus, the
lipid composition of human cortex lipid rafts closely
resembles that previously reported in cell membranes
from different sources [1,4,27].
Lipid rafts from control subjects
Analyses performed in control samples (C <60 and
C>60 groups) revealed no differences on either lipid
classes or fatty acid composition of lipid rafts in the
whole range of ages examined (24–85 years), suggest-
ing a considerable stability in lipid raft lipid biochem-
istry throughout the lifespan. Specifically, no differ-
ences were observed in the levels of CHO, PE, and SM

494 V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex
Table 2
Lipid class composition of brain cortex lipid raft samples from groups C<60,
C>60 and AD
C<60 C >60 AD
Sphingomyelin 11.87 ±1.37 11.43 ±1.44 11.71 ±1.48
Phosphatidylcholine 4.94 ±0.45 5.35 ±0.49 4.64 ±0.45
Phosphatidylserine 6.53 ±0.29 6.82 ±0.50 6.59 ±0.47
Phosphatidylinositol 2.13 ±0.11 2.16 ±0.17 3.05 ±0.62
Phosphatidylglycerol 0.68 ±0.11 0.70 ±0.10 1.03 ±0.33
Phosphatidylethanolamine 20.16 ±0.66 20.97 ±0.57 19.35 ±0.79
Sulphatides 10.07 ±0.75 10.56 ±0.68 9.37 ±0.57
Cerebrosides 4.82 ±0.87 5.10 ±0.79 4.43 ±0.75
Cholesterol 35.41 ±1.54 33.04 ±1.18 36.40 ±1.51
Free Fatty Acids 2.11 ±0.18 2.17 ±0.28 1.70 ±0.41
Sterol esters 1.23 ±0.52 1.68 ±0.60 1.72 ±0.80
Neutral Lipids 38.75 ±0.99 36.90 ±1.41 39.83 ±1.58
Polar Lipids 61.22 ±0.98 63.09 ±1.41 60.17 ±1.58
Phospholipid/Cholesterol 1.01 ±0.06 1.11 ±0.07 0.98 ±0.08
Results are expressed as mole % and represent means ±SEM.
Age (years)
20 30 40 50 60 70 80 90
mole% Cholesterol
0
10
20
30
40
50
Age (years)
20 30 40 50 60 70 80 90
mole% Sphingomyelin
0
10
20
30
40
50
Age (years)
30 40 50 60 70 80 90
mole% DHA
0
5
10
Age (years)
30 40 50 60 70 80 90
mole% AA
0
5
10
DHA=5,684+0,0111*Age
R
2
=0,023
AA=3,965-0,0058*Age
R
2
=0,020
CHO=36,58-0,058*Age
R
2
=0,051
SM=8,2057+0,0016*Age
R
2
=0,00004
Fig. 2. Regression analyzes for Cholesterol (CHO), sphingomyelin (SM), docosahexaenoic acid (DHA), and araquidonic acid (AA) contents as
a function of age in lipid rafts from control subjects. Linear regression equations and determination coefficients are indicated.
lipid classes, nor in the percentages of AA and DHA
fattyacids of lipidraft from control samplesas function
of age (24–85 years) (Fig. 2). In addition, no gender
differences were detected for any lipid class or fatty
acid in any group (not shown).
Correlation analyses of all lipid variables showed
that most bivariate relationships were similar between
groups C >60 and C <60. Only slight differences
were detected for some lipid classes such PE (r=
−0.862, p<0.005), sultatides (r=−0.863, p<
0.005), and cerebrosides (r=−0.877, p<0.005),
that were negatively correlated to SM in C <60 but

V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex 495
Table 3
Fatty acid composition of brain cortex lipid raft samples from groups C <60, C >60 and
ADs
C<60 C >60 AD
15 : 0 1.09 ±0.21 0.91 ±0.14 2.01 ±0.69
16 : 0 24.84 ±1.13 24.10 ±1.31 23.62 ±1.17
16 : 110.98 ±0.10 b 1.08 ±0.08 b 2.33 ±0.57 a
16 : 4 4.17 ±0.14 4.23 ±0.13 4.07 ±0.27
18 : 0 21.81 ±0.39 22.05 ±0.43 21.01 ±0.89
18:1 n-9 17.66 ±0.51 a 17.64 ±1.10 a 15.15 ±0.53 b
18:1 n-7 4.48 ±0.28 4.99 ±0.32 6.26 ±0.94
18 : 2 n-6 0.98 ±0.09 0.85 ±0.09 0.81 ±0.12
20 : 121.15 ±0.18 1.08 ±0.19 0.98 ±0.19
20 : 4 n-6 3.95 ±0.23 3.70 ±0.18 3.30 ±0.37
22 : 2 n-6 0.56 ±0.07 b 0.68 ±0.09 b 1.24 ±0.21 a
22 : 5 n-6 0.75 ±0.09 a 0.46 ±0.06 b 0.52 ±0.07 ab
22 : 6 n-3 6.53 ±0.38 a 6.87 ±0.34 a 4.91 ±0.55 b
Totals
Saturates 49.15 ±1.28 48.58 ±1.72 48.48 ±1.28
n-9 18.73 ±0.67 a 18,87 ±1.15 ab 15.98 ±0.62 b
n-3 6.70 ±0.40 ab 7,11 ±0.37 a 5.16 ±0.64 b
n-6 6.91 ±0.32 6,35 ±0.24 6.55 ±0.29
n-3 LCPUFA 6.70 ±0.40 ab 7.11 ±0.37 a 5.16 ±0.64 b
n-3/n-6 0.98 ±0.07 ab 1.12 ±0.05 a 0.77 ±0.08 b
18:1/n-3 LCPUFA32.76 ±0.26 2.57 ±0.22 3.39 ±0.42
Saturates/n-3 7.52 ±0.43 b 6.98 ±0.38 b 11.18 ±1.19 a
Saturates /n-9 2.66 ±0.15 2.72 ±0.25 3.07 ±0.18
Unsaturation index 109.42 ±2.26 a 110,34 ±2.41 a 95.75 ±4,01 b
Peroxidability index 85.73 ±3.60 a 86.51 ±3.36 a 66.81 ±5,90 b
Results are expressed as mole % and represent means ±SEM. Values in the same row bearing
different superscript letters are significantly different (p<0.05). Totals include some minor
components not shown. 1Contains n-9 and n-7 isomers. 2Contains n-11 and n-9 isomers.
318:1n-9/n-3 LCPUFA.
not in C >60 groups. Similarly, CHO was negatively
correlated to PE in C <60 group (r=−0.693, p<
0.05) but not in C >60 group.
Given the similarity in the lipid content of lipid rafts
fromcontrol brains, datafrom both groups werepooled
together and reanalyzed for multiple relationships be-
tween lipid classes and major fatty acids. The anal-
yses revealed positive significant correlations for PC,
PS, and PI versus DHA (r=0.593, p<0.01; r=
0.714, p<0.001 and r=0.703, p<0.005 for PC,
PS and PI, respectively) (Figs 3A and 3B). Another
important association was observed for PS and PI with
AA (r=0.670, p<0.005 and r=0.465, p<0.005
for PS and PI, respectively) (Figs 3C and 3D). Interest-
ingly, PE was not correlated to DHA or AA. Among
saturates, stearic acid (18:0) was positively correlated
to PS (r=0.565, p<0.01, Fig. 3E) and negatively
correlated to PE (r=−0.516, p<0.05) but uncorre-
lated to PC or PI. With regards to palmitic acid (16:0),
no significant correlations were detected for any of the
phospholipids analyzed. Taken together these relation-
ships might suggest that both 18:0 and DHA esterify
PS in lipid rafts from control brains. Unlike whole cell
membrane, PE seemed to be not associated with DHA
in lipid rafts microdomains. It was also evident that a
negative relationship between oleic acid (18:1n-9) and
saturates (r=−0.877, p<0.001) exists (Fig. 3F)
indicating the negative relationship between both fatty
acids in settling lipid raft fluidity.
Lipid rafts from AD brains
The proportion of phospholipids and cholesterol in
the lipid rafts from frontal cortex of AD brains was
similar and not significantly different from that of lipid
rafts from healthy subjects (Table 2). Also, there were
nosignificantchanges of flotillin-1contentin lipid rafts
from AD as compared to age-matched controls, a find-
ing that may reflect that the presence and formation of
rafts is unaffected in AD.
Nevertheless, despite the absence of changes in lipid
classes composition from lipid rafts, there was a sig-
nificant reduction of n-9, n-3, and n-3 LCPUFA in
lipid rafts isolated from AD subjects (Table 3). AD-
induced alterations of lipid raft n-3 and n-3 LCPUFA
composition can be entirely attributed to the depletion
of DHA levels, which represent more than 90% of the

496 V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex
PI (mole %)
0246
DHA (mole %)
0
2
4
6
8
10
Controls
Alzheimer’s
PS (mole %)
0 5 10 15
DHA (mole%)
0
2
4
6
8
10
PS (mole %)
0 5 10 15
AA (mole %)
0
1
2
3
4
5
6
7
PI-SLS vs AA-SLS
18:0 (mole %)
15 20 25
18:1n-9 (mole%)
10
15
20
25
30
PS (mole %)
2,5 5,07,5 10,0
18:0 (mole %)
15
20
25
PI (mole %)
0246
AA (mole %)
0
1
2
3
4
5
6
A B
C D
E F
Controls
Alzheimer’s
Controls
Alzheimer’s
Controls
Alzheimer’s
Controls
Alzheimer’s
Controls
Alzheimer’s
8
8
Fig. 3. Regression analyses for a subset of phospholipids and fatty acids in control ( )andAD(•)lipidrafts.A,B:linearrelationshipsbetween
PS (A) and PI (B) versus DHA; C,D: linear relationships between PS (C) and PI (D) versus AA; E: linear relationship between PS and palmitic
acid (18:0); F: linear relationship between oleic acid (18:1n-9) and stearic acid (18:0). Correlation coefficients and statistical significances are
indicated in the text. Units are expressed as mole percentage for all variables.
total n-3 fatty acids in all groups. Thus, in the AD
group the DHA content was 28% lower than in the age-
matched control group. Appreciable reductions of mo-
noene 18:1n-9 and 20:4n-6 were also observed in lipid
rafts from AD brains, yet for the later differences were
notsignificant. Also, AD lipidraftsexhibitedincreased
saturates/n-3 ratio and reduced unsaturation index (that
could be entirely attributable to the reductions in n-3
LCPUFA and n-9 monounsaturatedfatty acids) as well
as a significant reduction in the peroxidability of mem-
brane lipids. Interestingly, no significant differences
were found between females and males for any lipid
class or fatty acid in AD lipid rafts (not shown).
Analysisof correlation betweenall variables showed
that some of the existent relationships between lipid
classes and fatty acids were altered or disappeared in
ADlipid rafts. Thus, the positiverelationshipsbetween
PI and DHA or AA (Figs 3B and 3D) and between PS

V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex 497
and 18:0 (Fig. 3E) observed in control brains vanished
in AD rafts. The same stands for the relationship be-
tween oleic acid (18:1n-9) and saturates, 16:0 or 18:0
(Fig. 3F). On the other hand, unlike control rafts, PE
appeared to be positively correlated to AA and DHA
(r=0.871, p<0.005 and r=0.692, p<0.05 for
AA and DHA, respectively) while PI was negatively
related to stearic acid (r=−695, p<0.05) in AD
microdomains (not shown).
A deeper insight into the lipid alteration of AD lipid
rafts was obtained by using discriminant analysis. Our
resultsshowedthat the first canonicalfunctionaccount-
edforthegreat majority of the variationbetweengroups
(91.8%) while the second canonical variable accounted
only for 8.2%. The variables which showed the highest
absolute correlation with respect to every discriminate
functionwere age,n3/n6 ratio, saturates/n3 ratio,DHA,
peroxidability index, 18:1n-9, PI and PE. The 1st dis-
criminantfunction(mainlydeterminedby age variable)
clearly separates C <60 group from the rest of the
groups, while the 2nd function (defined by n3/n6 ratio,
saturates/n3 ratio, DHA, peroxidability index, 18:1n-9,
PI and PE) separates C >60 and AD groups (Fig. 4A).
Sequential Chi-square test revealed that the 1st dis-
criminant function contributes to the discrimination of
the groups to a large extent (χ2=74.06, p<0.001),
and according to the structure coefficients was mainly
determined by age variable. In order to test whether
the lipid composition per se could be used to identify
groups, we performed additional analyses without the
contribution of age. In this case, the contribution to
overall variance of the first and second canonical func-
tions was 77.5% (χ2=33.70, p<0.01) and 22.5%
(χ2=9.84, p>0.2), respectively. The 1st discrim-
inant function clearly separates AD group from con-
trol groups, while the 2nd function roughly separates
C>60 and C <60 groups (Fig. 4B). Structure coeffi-
cients revealed that the n3/n6 ratio, saturates/n3 ratio,
DHA and n-9 fatty acids (mainly 18:1n-9) determine
the 1st discriminant function, while the 2nd function
was defined by the 22:5n-6 content.
DISCUSSION
Our findings provide a new view of lipid rafts
in human brain cortex as liquid-ordered (lo) mem-
brane microdomains enriched in flotillin, choles-
terol, sphingolipids, and saturated fatty acids but al-
so containing small amounts of unsaturated fatty
acids(specifically monounsaturated and n-3 LCPUFA),
A
B
Fig. 4. Scatter plot of the first and second canonical variables in
the discriminant function analyses for C >60, C <60 and AD
groups. (A) including and (B) excluding age variable from analyses.
Centroids for each group are represented as white filled squares. For
details and interpretation see Results and Discussion sections.
phosphatidylserine, phosphatidylinositol, and phos-
phatidylethanolamine. Among n-3 LCPUFA contents
in lipid rafts, our analyses showed that docosahex-
aenoic acid accounted for more than 95% of total n-3
LCPUFA,while eicosapentaenoic acid(EPA, 20:5n-3),
the other important n-3 highly unsaturated fatty acid
in brain lipids, were negligible. Also, signaling pre-
cursor arachidonic acid (20:4n-6) was also present in

498 V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex
significant amounts in lipid rafts (around 3.5%) from
all groups and docosapentaenoic acid (DPA, 22:5n-
6), produced by further elongation and desaturation
of arachidonic acid in neural cells, represent less than
0.8% of total fatty acids. These significant amounts of
unsaturated fatty acids are reflected in the unsaturation
index (average 109.88), which apparently contradicts
the common view of lipid rafts as highly saturated mi-
crodomains[1]. However, increasing evidence indicate
that lipid rafts are heterogeneous both in terms of their
protein and lipid contents, and can be localized to dif-
ferentregionsof the cell[2,39]. Indeed,severalauthors
havefoundhighlevelsofunsaturatedfattyacids in lipid
rafts isolated from different cell types, includingnerve
cell membranes [40,41]. Specifically, polyunsaturated
lipids representing 3.4% of the total lipid were found in
lipid rafts from rat brain [42], and much higher levels
of polyunsaturated lipids (13.3% of overall lipid) have
been found in lipid rafts from RBL-2H3 cells [40]. The
consequences of appreciable amounts of LCPUFA in
lipid rafts are that these microdomains may exist in a
more loosely packed disordered state that phase sepa-
rate within the membrane due to their high cholesterol
and sphingomyelin contents [1]. In this sense, it has
been suggested that higher levels of polyunsaturated
lipids lead to an imperfect ordering of rafts lipids that
might allow accommodating transmembrane polypep-
tide helices that rafts would normally exclude [40].
Comparison of lipid rafts from control samples (C
<60 and C >60 groups) revealed no differences in
the levels of CHO, PE, PS, andSM lipid classes, or in
the percentages of saturates, monoenes, AA, and DHA
fatty acids in the range of ages analyzed (24–85years).
This is interesting since changes in whole brain lipid
composition as a function of age have been reported
in normal subjects [20,43]. Thus, in the frontal cortex
and hippocampus, PE and PC concentrations decrease
by about 30% in the healthy elderly comparedto young
adults [20]. Also, DHA contents in the main brain
phospholipids (PC and PE) have been reported to be
reduced in older compared to young subjects [43,44].
In agreement with our findings, studies performed on
synaptosomal lipid rafts have shown that SM levels do
not differ in mice from different groups of age [29].
Therefore, according to our present data, increasing
age does not alter lipid raft composition in control sub-
jects, suggesting the existence of homeostatic mecha-
nismswherebynervecellstend to maintain the physico-
chemical structure of lipid rafts.
Correlation analyses of lipid classes and fatty acids
in the lipid rafts of control subjects also revealed that
DHA was associated to PC, PS, and PI, rather than to
PE. It was also evident the strong relationship between
PS and 18:0, which is in good agreement with the no-
tion that 18:0 and DHA esterify PS in lipid rafts at po-
sitions sn-1 and sn-2 of the phospholipid backbone, re-
spectively [45]. In fact, such molecular association has
been shown to represent the most abundantform of PS
in neural membranes from human cortex [45], rat cor-
tex [46], rat hippocampus [47], olfactorybulb[48], and
photoreceptor discs [49]. In relation to the presence of
this molecular species of PS containing DHA in lipid
rafts,severalstudieshavedemonstratedits involvement
in the modulation of PI3K/Akt and Raf-1 signaling
pathways in neural cells [50,51]. The relationship be-
tween PS and DHA is complex and it appears that neu-
ronal survival inducedby DHA (and to a lesser extend
docosapentaenoic acid 22:5n-6) depends on its capac-
ity to increase PS in neural membranes [51]. Addi-
tionally, DHA is the precursor molecule of (10,17S)-
docosatriene (neuroprotectin D1 or NPD1), an oxy-
genated product of DHA which play an important role
insupportingneuronalsurvivalunderpathologicalcon-
ditions,includingAβ−42 -inducedneurotoxicity,byin-
ducing neuroprotective and antiapoptotic gene expres-
sion [52].
Our analyses revealed no differences in the con-
tents of flotillin-1, phospholipids, sphingomyelin, or
cholesterol in lipid rafts from AD frontal cortex com-
pared to healthy brains, an observation that reflects
that the presence and formation of rafts is unaffect-
ed in AD. In agreement with our findings, a recent
study by Molander-Melin and colleagues [3] reported
no changes in the recovery or compositionof the major
membrane lipids, glycerophospholipid, CHO, and SM
in lipid rafts in the frontal cortex of AD brains. These
observations are of critical importance since choles-
terol has been shown to interact with Aβin a recip-
rocal manner: Aβimpacts on cholesterol metabolism
and modifications of cholesterollevels alter Aβexpres-
sion [53]. In fact, some have suggested that elevated
cholesterol levels representa risk factor for AD, and a
linkage of cholesterol and AD has been associated to
the occurrence of ApoE4 [54,55]. However, it seems
now clear that rather than the bulk brain cholesterol
levels, changes in cholesterol nerve cell membrane do-
mains, i.e., alterations in the transbilayer distribution of
cholesterol, in particular in the exofacial leaflet of the
membrane, could act to promotesynthesis of Aβandto
catalyze the fibrillogenesis of soluble Aβin AD [53].
On the other hand, a complex relationship has been
proposed to exist between PSEN1 and membrane lipid

V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex 499
micro-environment [56]. These authors have shown
that brain membranes from mice expressing a human
wild-type PSEN1 transgene are less fluid and contain
higher cholesterol and sphingomyelin levels, suggest-
ing that interaction of PSEN1 with lipids directly af-
fects the fluidity of brain membranes [56].
Notwithstanding the absence of changes in lipid
classes, our data showed for the first time that AD lipid
rafts exhibited significant reductions in DHA (andcon-
sequently in n-3 fatty acids and n-3 LCPUFA, where
docosahexaenoic acid represents about90% of their to-
tal contents) when compared with age-matched con-
trols. These findings are in consonance with the obser-
vations in different brain areas including frontal cortex
and hippocampus from AD patients, where DHA in
main DHA-containing phospholipids (PE and PS) are
notably reduced [20–22,25,43]. The relevance of these
observations is outstanding given the extremely low
capacity of human brain to synthesize DHA through
desaturation/elongation processes of its n-3 LCPUFA
precursor α-linolenic acid [57]. As a result, depletion
of DHA in neural phospholipids cannot be mitigated
by compensatory metabolic pathways even after large
α-linolenic acid dietary intakes [43,57].
Interestingly,DHA reduction in lipid rafts from AD
brains was not accompanied by parallel increases of
docosapentaenoic acid (DPA, 22:5n-6, Table 2) or neg-
atively correlated to it (not shown), which is in contrast
with what has been observed in rat brain microsomes
in animals receiving n-3 deficient diets [58], where
thesubstitution of 18:0/22:6n-3with18:0/22:5n-6like-
ly provides an alternative mechanism for maintaining
membrane fluidity and interactions with other mem-
brane components, given the structural similarities of
the two molecular species [45].
We have also observed that, in absolute terms, AA
wasslightlyreducedinAD lipid rafts,butmoreinterest-
ingly, that the positive significant correlation between
PI and 20:4n-6 demonstrated in age-matched controls,
completely disappeared in AD lipidrafts. Given that PI
is considered to be the main AA-containing phospho-
lipid in neural membranes and that phosphoinositide
metabolites have been linked to synaptic survival, plas-
ticity, and long-term potentiation [59,60], alterations
of PI molecular composition at the raft microdomains
likely commit nerve cells to abnormal intracellular sig-
naling underlying synaptic dynamics in AD. In agree-
ment, reduced levels of PI-derived oleic and arachi-
donic acids have been reported to be significantly de-
creased in whole cell membrane in the hippocampus of
AD subjects [22].
Significant reductions of monoene 18:1n-9 were al-
so detected in lipid rafts from AD brains. One addi-
tional important observation derived from correlation
analyses performed here is that the significant nega-
tive relationship between 18:0 and 18:1n-9 observedin
lipid rafts from healthy brains vanished in AD samples
(Fig. 3). Such relationship is physico-chemically rele-
vant since the ratio between both fatty acids represents
an evolutionary conserved mechanism to preserve the
homeoviscous state of cell membranes in response to
different forms of physical and/or chemical stress [61,
62]. It can be hypothesized that the reduction in 18:1n-
9 together with disappearance of the relationship be-
tween 18:0 and 18:1n-9 in AD lipid rafts might indi-
cate a down-regulation of ∆9-desaturase, the enzyme
responsible for the synthesis of n-9 monounsaturated
fatty acids from their saturated precursors [63], which
wouldresultinaconcomitant loss of the ability to adjust
lipid raft physical order. This interesting hypothesis is
being currently assessed in brain samples fromhumans
and AβPP/PS1 transgenic mice in our laboratory.
AD lipid rafts also displayed increased saturates/n-3
ratio and reduced unsaturation index, which indicates
that raft microdomains from AD brain cortex are no-
tably more viscous and liquid-ordered than in the age-
matched control group. Alternatively, the reduction
of DHA and 18:1n-9 contents in raft domain phos-
pholipids in the AD group would alter the structure
of lipid rafts compared to normal subjects. Indeed, it
has been experimentally demonstrated that n-3 LCP-
UFA, mainlyDHA, are incorporated into both choles-
terol and sphingolipid-rich detergent-resistant liquid-
ordered (lo) and liquid disordered (ld) plasma mem-
brane microdomains in many cell types [26–28,64],
but the poor affinity of DHA and perhaps other long
chain PUFA for cholesterol provides a lipid-driven
mechanism for lateral phase separation of cholesterol-
and sphingolipid-rich lipid microdomainsfrom the sur-
rounding ldphase in model membranes [65,66] alter-
ing the size, stability, and distribution of cell surface
lipid microdomains such as rafts. Furthermore, it has
been proposed that microdomain PUFA impoverish-
ment may have profoundconsequences in the dynam-
ic partitioning of acylated proteins, membrane order
andfluidity,phasebehavior,elasticcompressibility, ion
permeability, fusion, rapid flip-flop, receptor binding
and resident protein function, thereby altering signal
transduction events [31,65,67].
Another consequence of the reduction in the unsat-
uration index of AD lipid rafts is the reduction in the
peroxidability of membrane lipids. It has been demon-

500 V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex
strated that polyunsaturated fatty acids are very sus-
ceptible to the oxidation induced by free radicals, gen-
erating specific reactive aldehydes, such as malondi-
aldehyde or 4-hydroxynonenal [37,68]. The impor-
tant PUFA content in brain tissue and its high oxy-
gen consumption support the possible significance of
lipid peroxidation-derivedprocesses in brain aging and
AD pathogenesis [69,70]. The significant reduction of
LCPUFA and peroxidability and unsaturation indexes
observedinlipidraftsfromAD brains is consistentwith
a progressive generation of aldehyde reactive species
and other lipoperoxides during the development of AD
pathology. Such generation of LCPUFA-related re-
active species is likely to be buffered in control aged
brains, as revealed by comparison with C <60 con-
trol brains (Table 3), which agrees with the notion that
mechanisms involved in cellular antioxidant defense
must have been depressed in late phases of AD [71].
Finally,consideringthatthedifferencesbetweennor-
mal and AD lipid rafts involved not only individual
lipid parameters but also a number of linear relation-
ships between them, we performed discriminant func-
tionanalyses to check whether lipid rafts from different
groups could be defined from a multivariate approach.
Our analyses were conclusive and revealed that, in-
dependently of a priori conditioning factors (age and
Braak stage), control and AD groups could be resolved
by means of two canonical functions (see Fig. 4B). The
first of these discriminant functions (defined by pre-
dictive variables n3/n6 ratio, saturates/n3 ratio, DHA
and18:1n-9)clearlyseparatedADfrom control groups,
while the second function (defined by 22:5n-6) seem-
inglyseparated groups C >60and C <60. We cancon-
clude from these analyses that lipid biochemical com-
position of lipid raft per se can be used as predictive
tool to determine the presence of AD pathology and, to
a lesser extent, to establish the influence of aging.
In summary,our present results provide the first de-
tailed view of the fatty acid composition of lipid rafts
isolated from frontal cerebral cortex of human brains.
The observations demonstrate the presence of signifi-
cant amounts of monounsaturated (especially 18:1n-9)
and polyunsaturated fatty acids (DHA and AA), in the
biochemical composition of lipid raft and point to the
existence of homeostatic mechanisms preserving lipid
raft status in normal frontal cortex. The disruption of
such mechanisms in AD brains alters lipid raft com-
position and physico-chemical properties, which may
explain the abnormal lipid raft signaling processes ob-
served in AD.
ACKNOWLEDGMENTS
This work was funded by grants SAF2007-66148-
C02-02 (Spanish Ministry of Education and Science),
PI08/0582 (Spanish Ministry of Health, Instituto de
Salud Carlos III), and supported by the EuropeanCom-
missionundertheSixth FrameworkProgramme(Brain-
Net Europe II, LSHM-CT-2004-503039). We thank
PULEVA BIOTECH (Spain) for collaborating in the
development of lipid strategies in the present project.
We are indebted to Dr. Miguel Molina for his generous
aid and helpful comments on the interpretation of mul-
tivariate analyses. This work is dedicated to the memo-
ry of Ignacio J. Lozano Soldevilla, a beloved colleague
who always brought jollity tothe art of producing sci-
ence, even at the twilight of his life.
Authors’ disclosures available online (http://www.j-
alz.com/disclosures/view.php?id=121).
REFERENCES
[1] Brown DA, London E (2000) Structure and function of
sphingolipid- and cholesterol-rich membrane rafts. J Biol
Chem 275, 17221-17224.
[2] Pike LJ (2003) Lipid rafts: bringing order to chaos. J Lipid
Res 44, 655-667.
[3] Molander-Melin M, Blennow K, Bogdanovic N, Dellheden
B, Mansson JE, Fredman P (2005) Structural membrane al-
terations in Alzheimer brains found to be associated with re-
gional disease development; increased density of gangliosides
GM1 and GM2 and loss of cholesterol in detergent-resistant
membrane domains. J Neurochem 92, 171-182.
[4] Michel V, Bakovic M (2007) Lipidrafts in health and disease.
Biol Cell 99, 129-140.
[5] Duyckaerts C, Dickson DW (2003) Neurodegeneration: The
molecular pathology of dementia and movement disorders.
ISN Neuropath. Press, Basel.
[6] Lee SJ, Liyanage U, Bickel PE, Xia W, Lansbury PT Jr,
Kosik KS (1998) A detergent-insoluble membrane compart-
ment contains A beta in vivo.Nat Med 4, 730-734.
[7] Hayashi H, Mizuno T, Michikawa M, Haass C, Yanagisawa K
(2000) Amyloid precursor protein in unique cholesterol-rich
microdomains different from caveolae-like domains. Biochim
Biophys Acta 1483, 81-90.
[8] Ehehalt R, Keller P, Haass C, Thiele C, SimonsK (2003) Amy-
loidogenic processing of the Alzheimer beta-amyloid precur-
sor protein depends on lipid rafts.J Cell Biol 160, 113-123.
[9] Tikkanen R, Icking A,Beicht P, Waneck GL, Volker H (2002)
The receptor-bound N-terminal ectodomain of the amyloid
precursor protein is associated with membrane rafts. Biol
Chem 383, 1855-1864.
[10] VetrivelKS,ChengH,KimSH,ChenY,BarnesNY,ParentAT,
Sisodia S, Thinakaran G(2005) Spatial segregation of gamma-
secretase and substrates in distinct membrane domains. J Biol
Chem 280, 25892-25900.
[11] Ikezu T,Trapp BD, Song KS, Schlegel A, Lisanti MP, Okamo-
to T (1998) Caveolae, plasma membrane microdomains for
alpha-secretase-mediated processing of the amyloid precursor
protein. J Biol Chem 273, 10485-10495.

V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex 501
[12] Marlow L, Cain M, Pappolla MA, Sambamurti K (2003) Beta-
secretase processing of the Alzheimer’s amyloid protein pre-
cursor (APP). J Mol Neurosci 20, 233-239.
[13] Riddell DR, Christie G, Hussain I, Dingwall C (2001) Com-
partmentalization of beta-secretase (Asp2) into low-buoyant
density, noncaveolar lipid rafts. Curr Biol 11, 1288-1293.
[14] Cordy JM, Hussain I, Dingwall C, Hooper NM, Turner AJ
(2003) Exclusively targeting beta-secretase to lipid rafts by
GPI-anchor addition up-regulates beta-site processing of the
amyloid precursor protein. Proc Natl Acad Sci USA 100,
11735-11740.
[15] Vetrivel KS, Cheng H, Lin W, Sakurai T, Li T, Nukina N,
Wong PC, Xu H, Thinakaran G (2004) Association of gamma-
secretase with lipid rafts in post-Golgi and endosome mem-
branes. J Biol Chem 279, 44945-44954.
[16] Hattori C, Asai M, Onishi H, Sasagawa N, Hashimoto Y, Sai-
do TC, Maruyama K, Mizutani S, Ishiura S (2006) BACE1
interacts with lipid raft proteins. J Neurosci Res 84, 912-917.
[17] Hur JY, Welander H, Behbahani H, Aoki M, Franberg J, Win-
blad B, Frykman S, Tjernberg LO (2008) Active γ-secretase
is localized to detergent-resistant membranes in human brain.
FEBS J 275, 1174-1187.
[18] Uauy R, Hoffman DR, Peirano P, Birch DG, Birch EE (2001)
Essential fatty acids in visual and brain development. Lipids
36, 885-895.
[19] Lane RM, Farlow MR (2005) Lipid homeostasis and
apolipoprotein E in the development and progression of
Alzheimer’s disease. J Lipid Res 46, 949-968.
[20] Soderberg M, Edlund C, Kristensson K, Dallner G (1991)
Fatty acid composition of brain phospholipids in aging and in
Alzheimer’s disease. Lipids 26, 421-425.
[21] Soderberg M, Edlund C, Alafuzoff I, Kristensson K, Dallner
G (1992) Lipid composition in different regions of the brain
in Alzheimer’s disease/senile dementia of Alzheimer’s type. J
Neurochem 59, 1646-1653.
[22] Prasad MR, Lovell MA, Yatin M, Dhillon H, Markesbery
WR (1998) Regional membrane phospholipid alterations in
Alzheimer’s disease. Neurochem Res 23, 81-88.
[23] Hashimoto M, Hossain S, Shimada T, Sugioka K, Yamasaki H,
Fujii Y, Ishibashi Y, Oka JI, Shido O (2002) Docosahexaenoic
acid provides protection from impairment of learning ability in
Alzheimer’s disease model rats.J Neurochem 81, 1084-1091.
[24] Calon F, Lim GP, Yang F, Morihara T, Teter B, Ubeda O,
Rostaing P, Triller A, Salem N. (2004) Docosahexaenoic acid
protects from dendritic pathology in an Alzheimer’s disease
mouse model. Neuron 43, 633-645.
[25] Young G, Conquer J (2005) Omega-3 fatty acids and neu-
ropsychiatric disorders. Reprod Nutr Dev 45, 1-28.
[26] Fan YY, McMurray DN, Ly LH, Chapkin RS (2003) Dietary
(n-3) polyunsaturated fatty acids remodel mouse T-cell lipid
rafts. J Nutr 133, 1913-2190.
[27] Ma DW, Seo J, Davidson LA, Callaway ES, Fan YY, Lup-
ton JR, Chapkin RS (2004a) n-3 PUFA alter caveolae lipid
composition and resident protein localization in mouse colon.
FASEB J 18, 1040-1042.
[28] Ma DW, Seo J, Switzer KC, Fan YY, McMurray DN, Lup-
ton JR, Chapkin RS (2004b) n-3 PUFA and membrane mi-
crodomains: a new frontier in bioactive lipid research. J Nutr
Biochem 15, 700-706.
[29] Igbavboa U, Eckert GP, Malo TM, Studniski AE, Johnson
LNA, Yamamoto N, Kobayashi M, Fujita SC, Appel TR,
M¨
uller WE, Wood WG,Yanagisawa K (2005) Murine synap-
tosomal lipid raft protein and lipid composition are altered by
expression of human apoE 3 and 4 and by increasing age. J
Neurol Sci 229-230, 225-232.
[30] Chen W, Jump DB, Esselman WJ, Busik JV (2007) Inhibi-
tion of cytokine signaling in human retinal endothelial cells
through modification of caveolae/lipid rafts by docosahex-
aenoic acid. Invest Ophthalmol VisSci 48, 18-26.
[31] Chapkin RS, Wang N,Fan YY, Lupton JR, Prior IA (2008)Do-
cosahexaenoic acid alters the size and distribution of cell sur-
face microdomains. Biochim Biophys Acta 1778, 466-471.
[32] Braak H, Braak E, Peter J, Morrison JH (1999) Neurodegen-
erative and age-related changes in structure and function of
cerebral cortex, in Cerebral cortex, (Peters A., Morrison J.
H., eds), Vol. 14. Kluwer Academic/Plenum Publishers, Dor-
drecht.
[33] Braak H, Alafuzoff I, Arzberger T,Kretzschmar H, Del Tredici
K (2006) Staging of Alzheimer disease-associated neurofibril-
lary pathology using paraffin sections and immunocytochem-
istry. Acta Neuropathol 112, 389-404.
[34] Mukherjee A, Arnaud L, Cooper JA (2003) Lipid-dependent
recruitment of neuronal Src to lipid rafts in the brain. J Biol
Chem 278, 40806-40814.
[35] ChristieWW(1982)LipidsAnalysis.PergamonPress, Oxford.
[36] Olsen RE, Henderson RJ(1989) The rapid analysis of neutral
and polar marine lipids using double development HPTLC and
scanning densitometry. J Exp Mar Biol Ecol 129, 189-197.
[37] Cosgrove JP, Church DF,Pryor WA (1987) The kinetics of the
autoxidation of polyunsaturated fatty acids. Lipids 22, 299-
304.
[38] Huberty CJ (1994) Applied discriminant analysis. Wiley-
Interscience. New York.
[39] Pike LJ (2004) Lipid rafts: heterogeneity on the high seas.
Biochem J 378, 281-292.
[40] Fridriksson EK, Shipkova PA, Sheets ED, Holowka D, Baird
B, McLafferty FW (1999) Quantitative analysis of phos-
pholipids in functionally important membrane domains from
RBL-2H3 mast cells using tandem high-resolution mass spec-
trometry. Biochemistry 38, 8056-8063.
[41] Pike LJ, Han X, Chung KN, Gross R (2002) Lipid rafts are
enriched in arachidonic acid and plasmenylethanolamine and
their composition is independent of caveolin-1 expression: a
quantitative electrospray ionization/mass spectrometric anal-
ysis. Biochemistry 41, 2075-2088.
[42] Brugger B, Graham C, Leibrecht I, Mombelli E, Jen A,
Wieland F, Morris R (2004) The membrane domains occu-
pied by glycosylphosphatidylinositol-anchored prion protein
and Thy-1 differ inlipid composition. J Biol Chem 279, 7530-
7536.
[43] Plourde M, Fortier M, Vandal M, Tremblay-Mercier J, Free-
mantle E, B´
egin M, Pifferi F, Cunnane SC (2007) Un-
resolved issues in the link between docosahexaenoic acid
and Alzheimer’s disease. Prostaglandins Leukot Essent Fatty
Acids 77, 301-308.
[44] Horrocks LA, Farooqui AA (2004) Docosahexaenoic acid in
the diet: its importance in maintenance and restoration of neu-
ral membrane function. Prostaglandins Leukot Essent Fatty
Acids 70, 361-372.
[45] Mozzi R, Buratta S, Goracci G (2003) Metabolism and func-
tions of phosphatidylserine in mammalian brain. Neurochem
Res 28, 195-214.
[46] Little SJ, Lynch MA, Manku M, Nicolaou A (2007) Docosa-
hexaenoic acid-induced changes in phospholipids in cortex
of young and aged rats: a lipidomic analysis. Prostaglandins
Leukot Essent Fatty Acids 77, 155-162.

502 V. Mart´
ın et al. / Lipid Rafts in AD Frontal Cortex
[47] Murthy M, Hamilton J, Greiner RS, Moriguchi T, Salem NJr,
Kim HY (2002) Differential effects of n-3 fatty acid deficien-
cy on phospholipid molecular species composition in the rat
hippocampus. J Lipid Res 43, 611-617.
[48] Hamilton L, Greiner R, Salem NJr., Kim HY (2000) n-3 fat-
ty acid deficiency decreases phosphatidylserine accumulation
selectively in neuronal tissues. Lipids 35, 863-869.
[49] Aveldano MI, Bazan NG (1983) Molecular species of phos-
phatidylcholine, -ethanolamine, -serine, and -inositol in mi-
crosomal and photoreceptor membranes of bovine retina. J
Lipid Res 24, 620-627.
[50] Kim HY, Akbar M, Lau A (2003) Effects of docosapentaenoic
acid on neuronal apoptosis. Lipids 38, 453-457.
[51] Akbar M, Calderon F, Wen Z, Kim HY (2005) Docosahex-
aenoic acid: A positive modulator of Akt signaling in neuronal
survival. Proc Natl Acad SciUSA102, 10858-10863.
[52] Lukiw WJ, Cui JG, Marcheselli VL, Bodker M, Botkjaer A,
Gotlinger K, Serhan C N, Bazan N (2005) A role for do-
cosahexaenoic acid-derived neuroprotectin D1 in neural cell
survival and Alzheimer disease. J Clin Invest 115, 2774-2783.
[53] Gibson W, Eckert GP, Igbavboa U, M¨
uller WE (2003) Amy-
loidbeta-proteininteractionswithmembranesandcholesterol:
causes or casualties of Alzheimer’s disease. Biochim Biophys
Acta 1610, 281-290.
[54] Cedazo-M´
ınguez A (2007) Apolipoprotein E and Alzheimer’s
disease: molecular mechanisms and therapeutic opportunities.
Cell Mol Med 11, 1227-1238.
[55] Donahue JE, Johanson CE (2008) Apolipoprotein E, amyloid-
beta, and blood-brain barrier permeability in Alzheimer dis-
ease. J Neuropathol Exp Neurol 67, 261-270.
[56] Eckert GP, M¨
uller WE (2009) Presenilin 1 modifies lipid raft
composition of neuronal membranes. Biochem Biophys Res
Commun 382, 673-677.
[57] Plourde M, Cunnane SC (2007) Extremely limited synthesis
of long chain polyunsaturates in adults: implications for their
dietary essentiality and use as supplements. Appl Physiol Nutr
Metab 32, 619-634.
[58] Garc´
ıa MC, Ward G, Ma YC, Salem N, Kim HY (1998) Effect
of docosahexaenoic acid on the synthesis of phosphatidylser-
ine in rat brain microsomes and C6 glioma cells. J Neurochem
70, 24-30.
[59] Kelly A, Lynch MA (2000) Long-term potentiation in dentate
gyrus of the rat is inhibited by the phosphoinositide 3-kinase
inhibitor, wortmannin. Neuropharmacology 39, 643-651.
[60] Osborne SL, Meunier FA, Schiavo G (2001) Phosphoinosi-
tides as key regulators of synaptic function. Neuron 32, 9-12.
[61] McDonald AG (1988) Application of the theory of homeo-
viscous adaptation to excitable membranes: pre-synaptic pro-
cesses. Biochem J 256, 313-327.
[62] Pamplona R, Barja G, Portero-Ot´
ın M (2002) Membrane fat-
ty acid unsaturation, protection against oxidative stress, and
maximum life span: a homeoviscous-longevity adaptation?
Ann N Y Acad Sci 959, 475-490.
[63] Nakamura MT, Nara TY (2004) Structure, function, and di-
etary regulation of delta6, delta5, and delta9 desaturases. Annu
Rev Nutr 24, 345-376.
[64] Fan YY, Ly L, Barhoumi R, McMurray DN, Chapkin RS
(2004) Dietary docosahexaenoic acid suppresses T cell protein
kinase C theta lipid raft recruitment and IL-2 production. J
Immunol 173, 6151-6160.
[65] Stillwell W, Wassall SR (2003) Docosahexaenoic acid: mem-
brane properties of a unique fatty acid. Chem Phys Lipids 126,
1-27.
[66] Shaikh S. R., Dumaual A. C., Castillo A., LoCascio D., Sid-
diqui R. A., Stillwell W. and Wassall S. R. (2004) Oleic and
docosahexaenoic acid differentially phase separate from lipid
raft molecules: a comparative NMR, DSC, AFM, and deter-
gent extraction study. Biophys J 87, 1752-1766.
[67] Shaikh SR, Caffrey M, Stillwell W, Cherezov V, Wassall
SR (2003) Interaction of cholesterol with a docosahexaenoic
acid-containing phosphatidylethanolamine: trigger for mi-
crodomain/raft formation? Biochemistry 42, 12028-12037.
[68] Esterbauer H, Schaur RJ, Zollner H(1991) Chemistry and bio-
chemistry of 4-hydroxynonenal, malonaldehyde and related
aldehydes. Free Radic Biol Med 11, 81-128.
[69] Aksenov MY, Aksenova MV, Butterfield DA, Geddes JW,
Markesbery WR (2001) Protein oxidation in the brain in
Alzheimer’s disease. Neuroscience 103, 373-383.
[70] Pamplona R, Dalf´
o E, Ayala V, Bellmunt J, Prat J, Ferrer I,
Portero MJ (2005) Proteins in human brain cortex are modi-
fied by oxidation, glycoxidation, and lipoxidation: effects of
Alzheimer’s disease and identification of lipoxidation targets.
J Biol Chem 280, 21522-21530.
[71] Shi Q, Gibson GE (2007) Oxidative stress and transcriptional
regulation in Alzheimer disease. Alzheimer Dis Assoc Disord
21, 276-291.