close

Вход

Забыли?

вход по аккаунту

?

Clinical and pathological correlates of apolipoprotein E 4 in Alzheimer's disease.

код для вставкиСкачать
Clinical and Pathological Correlates of
Apolipoprotein E 84 in Alzheimer's Disease
Teresa Gomez-Isla, MD,* Howard L. West, MPhil,* G . William Rebeck, PhD,* Sceven D. Harr, BA,*
John H. Growdon, MD,* Joseph J. Locascio, PhD,* Thomas T. Perls, MD,t$ Lewis A. Lipsitz, MD,?.
and Bradley T. Hyman, M D , PhD*
Inheritance of the apolipoprotein E (apoE) ~4 allele is associated with a high likelihood of developing Alzheimer's
disease (AD). The pathophysiologic basis of this genetic influence is unknown. We reasoned that understanding the
influence of apoE ~4 on the clinical course and neuropathological features of AD may provide tests of potential mechanisms. We carried out a prospective longitudinal study to compare the age of onset, duration, and rate of progression
of 359 AD patients to apoE genotype. Thirty-one of the individuals who died during the study were available for
quantitative neuropathological evaluation. Statistically unbiased stereological counts of neurofibrillary tangles (NFTs)
and AD deposits were assessed in a high-order association cortex, the superior temporal sulcus. Analysis o f clinical
parameters compared with apoE genotype showed that the € 4 allele is associated with an earlier age of onset but no
change in rate of progression o f dementia. Quantitative neuropathological assessment revealed that NFTs were strongly
associated with clinical measures of dementia duration and severity but not with apoE genotype. AP deposition, by
.
results indicate that
contrast, was not related to clinical features but was elevated in association with apoE ~ 4 These
apoE € 4 is associated with selective clinical and neuropathological features of AD and support hypotheses that focus
on an influence of apoE € 4 on amyloid deposition.
Gomez-Isla T, West H L , Rebeck GW, Harr SD, Growdon JH, Locascio JJ, Pcrls TT, Lipsitz LA,
Hyman B'l'. Clinical and pathological correlates of apolipoprotein E ~4
in Alzheimer's dkease. Ann Neurol 1996;39:62-70
There are three common alleles of the apolipoprotein
E (apoE) gene, ie, apoE ~ 2 apoE
,
~ 3 and
, apoE ~ 4 ,
with allele frequencies in the general population of approximately 0.08, 0.78, and 0.14, respectively [ I , 21.
T h e apoE ~4 allele is overrepresented in Alzheimer's
disease (AD) [3-101 and is accepted as a genetic risk
factor. T h e biological mechanism underlying this genetic risk is unknown. T w o hypotheses have emerged,
one being that apoE/AP binding leads to an influence
on senile plaques (9%)
[ 5 , 11-14], and the other being
that apoE/T interactions lead to an influence on neurofibrillary tangles (NFTs) [15, 161. In this context, we
studied apoE genotypes in a large group of prospectively followed AD patients, many of whom had also
undergone quantitative neuropathological examination
after death. W e found that, rather than uniformly
worsening all aspects of the disease, apoE € 4 selectively
affects specific aspects of the disease process; ie, apoE
~4 is associated with earlier age of dementia onset and
increased AP deposition. ApoE ~4 does not influence
rate of clinical progression or N F T formation. W e also
observed that there was a strong correlation between
NFT number and severity or duration of dementia,
whereas amyloid did not correlate with any clinical
measure of dementia.
T h e observation that there are heritable risk factors
for AD suggests that study of subpopulations defined
by genotype might help to shed light on the biological
mechanisms underlying genetic effects. Along these
lines, our current data favor an apoE €4 effect on AP
deposition and argue against hypotheses based on
a p o E / N F T interactions.
From the *Neiiroloby Service. Mas'ichusetts General Hospital.
tHebrew Rehabilitation Center for the Aged. 'ind fI)cacones Eldercare, 1)e'icoiicss Hospital, Boston. MA.
Received Jul 5, 1335, and in revised form Aug 7 and Sep 5. Acceptrd for publication Scp 7, 1335.
Materials and Methods
W e enrolled 359 patients with a diagnosis of probable or
definite A D [17]. There were 137 ineii and 222 women with
a mean age of 77.8 years (rangc, 47-104 years) and a mean
age of onset o f 7 2 . 0 years (range, 46- 103 years). Blood samples wcrc collected from the 267 patients alive at the time
of the study and brain tissue from 92 definite A D patients
who had died at the time of this study. Sufficient tissue was
available in 31 of the 92 autopsied cases for detailed quanri-
Address correspondence to Dr Hym:in. Neurology Service. Warren
407. Massacliiisetth Gcneral Hospital, Boston. MA 011 14.
62 Copyright 0 1996 by the American Neurological Association
tative neuropathological studies. To test whether apoE genotype was correlated with significant differences in clinical
course, we charted behavioral abnormalities on the 153 AD
patients who underwent serial biannual evaluations. Average
follow-up time was 31.8 % 18.7 months (mean 5 SD). The
tests measured both cognitive and functional impairments,
the information, memory, and concentration (IMC) subtest
of the Blessed Dementia Scale (BDS) [18] score and the
Activities of Daily Living (ADL) [ 191 score, respectively. The
IMC-BDS is based on examination of the patient and detects
the presence and severity of dementia. Scores of 0 to 3 are
normal, whereas a score of 37 indicates maximal dementia
severity. The ADL scale is completed by family members at
clinical visits and contains 31 questions grouped in the following seven general areas: self-care activities, household care,
employment and recreation, shopping and money, travel,
communication, and social relationships. Independence on
each item is rated from 0 (normal) to 3 (fully dependent),
and overall scores range from 0 (fully independent) to 100
(maximally dependent). The rate of progression was calculated as the total change in the IMC-BDS and in the ADL
scores between the first and last test divided by the time
elapsed between these tests; preliminary analyses revealed that
the rate of these changes was linear with time over the course
sampled. Age of onset of dementia was determined by the
examining neurologist at the first clinic visit from patient
and family interview, by asking “when was the patient last
normal?”
As a control group, we examined 129 nondemented individuals, including 52 men and 77 women with a mean age
of 77.8 years (range, 42-102 years). Of the 129, 101 were
examined during life and had normal scores on the IMCBDS or the Folstein Mini-Mental State examination [20].
Twenty-eight other individuals had died by the time of the
study and were selected as controls because their brains
showed no neuropathologic evidence of AD or other neurodegenerative illness. Additional non-Alzheimer’s dementia
control groups consisted of tissues obtained from the Brain
Tissue Resource Center (McLean Hospital, Belmont, MA)
from individuals with the neuropathological diagnosis of
Pick’s disease (n = 31) or progressive supranuclear palsy
(PSP; n = 27).
ApoE genotyping was conducted by a polymerase chain
reaction-based approach [3, 51 by using D N A isolated from
blood or autopsy tissue. D N A samples were amplified in the
presence of 50 mM KCI, 10 mM Tris-HCI (pH 9.0), 0.1%
Triton X-100, 10% dimethyl sulfoxide, 1 mM MgCI2, 200
mM dNTPs, and 0.5 mM apoE-specific primers,
5’-TAA GCT TGG CAC G G C TGT CCA AGG A-3‘
5‘-ACA GAA TTC G C C CCG G C C TGG TAC ACT
G C C A-3‘
(35 cycles of 94”C, 30”/65”C, 3O”/7O0C, 1’30”). Amplified
DNA was digested with 3 units of the restriction endonuclease HhaI (37°C for 1 hour). DNA fragments were separated
by 8% nondenaturing polyacrylamide gel electrophoresis and
visualized with 1 pglml ethidium bromide.
The neuropathological diagnosis of AD (without addi-
tional abnormalities such as stroke, Parkinson’s disease, or
Lewy body disease) was established in all cases by the Massachusetts Alzheimer’s Disease Research Center Brain Bank using Khachaturian [21] criteria for the diagnosis of AD.
Thirty-one of the 92 autopsy cases were chosen for intensive
quantitative study based on the individual having been evaluated in one of the clinical units of the Massachusetts Alzheimer’s Disease Research Center, and on the availability of
technically adequate tissue. The right or left temporal lobe
(alternating among Massachusetts Alzheimer’s Disease Research Center Brain Bank cases) was fixed en bloc in paraformaldehyde-lysine-metaperiodate for 24 to 36 hours at
4°C and then cryoprotected in 15% glycerol/O.l M Trisbuffered saline (pH 7.4) at 4°C. Fifth-micrometer-thick frozen sections were prepared with a freezing sledge microtome
and individually stored in cryoprotectant at -80°C until
srained. Sections were immunostained with the monoclonal
antibody against AP, 10D5 [22] (courtesy of Dr Dale
Schenk, Athena Neuroscience, South San Francisco, CA),
or a monoclonal aptibody against phosphorylated Z, PHF1 (courtesy of Dr Peter Davies, Albert Einstein College of
Medicine, Bronx, NY) [23]. In both cases, detection was
via horseradish peroxidase-linked goat anti-mouse secondary
antibody (Jackson Immunoresearch, West Grove, PA) and
imrnunoreactivity visualized with 3,3’-diaminobenzidine.
The superior temporal sulcus (STS) region was chosen for
morphometric analyses for several reasons. The STS is anatomically unique because it is one of only three areas of
association cortex identified in the monkey that receives afferent input from all sensory modalities [24]. The STS is
thus “higher order” association cortex and is known from
previous neuropathological studies to be severely and consistently affected in AD [25, 261. In addition, from a technical
point of view, the STS has clearly defined boundaries and
its structure is remarkably consistent across brains, decreasing
potential anatomical variability among brains.
Amyloid deposition was quantified using AP immunostaining and a Bioquant image analysis system [5, 271. Video
images were captured and a threshold optical density obtained, which discriminated staining from background. Manual editing of each field eliminated artifacts, separated contiguous stuctures, and deleted vessel-associated staining. A strip
of cortex 700 p m wide by the depth of gray matter, approximately 1 to 2 cm medial to the crown of the middle temporal
gyrus on the lower bank of the STS, was chosen in each
case for analysis. The “amyloid burden,” defined as the total
percentage of cortical surface area covered by SPs was calculated for each case.
NFTs were counted using the optical disector technique
and statistically unbiased stereological principles [28]. A strip
of cortex in the STS of the same dimensions and from the
same neuroanatomical location as described above was chosen for analysis. Total number of NFTs per section was ca1culated by determining NFT density within the counted volume and multiplying the result by the total STS volume per
section (measured area multiplied by section thickness) for
each case.
Subjects were divided into groups based on diagnosis and
genotype. Infrequent neuropathology subjects with an apoE
E2/2 (n = 0 AD), apoE &2/3 (n = 1 AD), or apoE ~ 2 / (n
4
Gomez-Isla et al: ApoE E4 in Alzheimer’s Disease
63
Table 1. Demographics, ApoE Genotypes, and Allele Frequencies in AD, Control, Pick j Disease, and PSI) Populations
AD (n
Sex (MIF)
Age
ApoE
414
314
313
214
213
212
Allele frequencies
€4
€3
€2
=
359)
Control (n = 129)
General
Population
Picks (n = 31)
PSP (n
=
27)
1371222
77.8 2 9.6
52177
77.8 ? 13.5
16115
68.3 -+ 13.3
2017
70.8 2 4.4
l6Y0 (59)
429’0 (152)
330/0 ( 1 1 8 )
3% ( 1 1 )
5% (17)
< I % (2)
<l%l ( 1 )
23% (30)
64% (82)
1.5% ( 2 )
1 1 % ) (14)
0
0
13%
68%
0
16%
3%
0
15% (4)
74% (20)
7% ( 2 )
4”/0 ( 1 )
0
0.3‘)’ ( p < 0.001)
0.56
0.04h ( p < 0.005)
0.13
0.8 1
0.06
0.14
0.78
0.08
(4)
(21)
(5)
(1)
0.06
0.83
0.1 1
0.11
0.83
0.06
The general population data are derived from the study of 1,209 individuals, aged 22 to 71 years [2].The control population consisted of
129 individuals, 42 to 102 years old; the AD population consisted of 359 individuals, onset at 46 to 103 years. In the AD population the
apoE &4 allele frequency was significantly higher vp < 0.001) and the apo E ~2 allcle frequency significantly lower (hp< 0.001) than in
the general population (x’ goodness of fit test).
ApoE = apolipoprorein E; A11 = Alzheimer’s disease; PSI’ = progressivc supranuclear palsy.
= 1 AD) genotype were not included in neuropathological
analyses due to small group sizes. Allele frequencies for AD
patients were compared with estimates of a normal populagoodness of fit procedure. A series
tion distribution b y a
of iinivariate and multivariate analyses of variance were conducted for both the clinical and neuropathological evaluations. For the investigation of clinical features, dependent
variables were rate of BDS and ADI. change, arid predictor
variables or covariates were apoE genotype, age at onset, duration of illness, sex of the patient, number of BDS t a t s
available, and first RDS score available (representing AD severity at start of study). Kaplan-Meier survival analyses were
used to relate age of onset to apoE genotype and compare
genders in terms of these relations. For the analyses of neuropathological factors, dependent variables were amyloid burden and total N F T count, whereas predictors or covariates
were apoE genotype, illness duration, and age of onset. All
clinical and morphological measurements were performed
without knowledge of genotype. Statistical analyses wcre performed using SAS software (Statistical Analysis Systems, Inc,
Cary, NC).
x’
Results
T h e apoE ~4 allele was overrepresented in AD patients
compared to the general population by about threefold
(0.39 vs 0.14, X’ goodness of fit statistic = 376.8, p
< 0.001) (Table 1 ) . T h e apoE ~4 allele frequency in
our control group (0.13) is similar to the allele frequency of 0.14 observed in the general population [ l ,
21. T h e apoE ~2 allele was underrepresented i n AD
relative t o the distribution for the general population
(x’ goodness of fit statistic = 15.8, p < 0.001) (see
Table I ) . As dddirional controls, we studied two groups
of individuals with neurodegenerative diseases other
than AD. Both rhe group of neuropathologically con-
64
Annals of N e u r o l o g
Vol 39
No 1 January 1996
Table 2. ApoE Genotypes and Age o f Onset o f Dementia in
AD
ApoE Genorype
n
Age of Onset
(Mean
SD)
€414
€314
€313
€214
€213
€212
54
142
107
11
13
2
67.1
72.3
73.9
69.4
73.4
69.0
+
7.6
2 8.3
? 11.7
f 11.0
? 12.9
? 9.9
-C
There is R significant difference in age of onset of dernenria aniong
genotype groups ( p < 0.001). Fisher PLSD t u t showed painvise
differences ( p < 0.05) between 2 / 3 and 414, between 3 / 3 and 31
4 or 4/4, and between 3/4 and 414.
ApoE = apolipoprotein E; AD = Alzheimer’s disease.
firmed Pick’s disease and of PSP had apoE ~4 allele
frequencies indistinguishable from the control population (see Table 1).
There was a progressive decrease in age of dementia
onset in AD with inheritance of apoE ~4 compared
with apoE ~ 3 1 3individuals (Table 2). Age of dementia
onset declined significantly as the number of apoE ~4
alleles increased ( p < 0.0001 for linear contrast apoE
~ 3 1 3to apoE G I 4 t o apoE ~ 4 1 4 )T. h e age of onset for
apoE ~ 2 1 3(73.4 2 12.9 years, n = 13) was statistically
different from only the a p o E ~ 4 1 4subgroup ( p <
0.05). Kaplan-Meier curves of the proportion of the
group remaining unaffected [2?] were plotted to further evaluate the relationship a m o n g genotype, gender,
a n d age of onset of dementia (Fig I). For m e n and
women combined (n = 303), as well as for women
alone (n = 193), the curves for all pairs of genotypes
(ie, apoE €313, apoE €314, and apoE €414) were significantly different from one another (each, p < 0.01).
In men (n = IIO), apoE €414 was significantly different from apoE ~ 3 / (3p < 0.05) and apoE ~ 3 / (4 p <
0.005), but apoE €313 and apoE €314 were not significantly different from each other.
Mean IMC-BDS scores worsened by 3.5 points per
year and the ADL scores by 10.6% per year in the AD
group as a whole, although cognitive and behavioral
scores varied among individual patients. Based on these
measures, apoE genotype did not affect rate of dementia progression following AD onset. Rates of change
on both the BDS (Fig 2A) and the ADL scale (Fig 2B)
were similar among AD patients grouped according to
apoE genotype. N o differences among apoE groups
were found after adjusting for covariates of age of onset, duration, number of tests, and first BDS score.
Analysis of the 10% of individuals with the fastest or
slowest rates revealed that genotypes were not differentially distributed even in these cases at the extremes of
progression variability.
We also reviewed the estimated duration of illness
as available in the clinical population and found, for
the apoE €3.3 to apoE €3.4 to apoE €4.4 groups, there
was a significant increase in duration of illness (Fig 3;
p < 0.01). Further evaluation showed that there
tended to be a negative correlation between age of onset and duration; ie, people with earlier onset tend to
have longer durations. An analysis of covariance
showed that the significant genotype effect for age of
onset remained when duration or age were covaried,
but the gentoype effect for duration was no longer significant when age of onset was covaried. A model consistent with these findings is that the apoE €4 allele
leads to an earlier age of onset, which is associated, not
surprisingly, with longer durations. The findings for
genotype differences in duration therefore appear to be
an indirect consequence of the age of onset effect.
The brains of 31 AD patients were studied using
quantitative neuropathological measures of SP density,
amyloid burden, and number of NFTs in the STS.
There was a significant relationship between amyloid
deposition and apoE genotype, with the average amyloid burden increasing steadily from apoE €313 to apoE
€314 to apoE €414 in AD brains ( p < 0.05) (Fig 4 ) .
A post hoc test showed that the increase from apoE
€313 to apoE €414 was linear and highly significant (F
= 8.24; df = 1, 26; p < 0.01). ApoE genotype accounted for 26% of the variance in amyloid burden.
Covarying with duration of illness did not affect this
result, suggesting that length of disease does not influence amyloid burden. In further analyses, we found
that neither duration of illness nor last BDS score correlated with amyloid burden for the AD group as a
whole nor within any of the apoE genotype subgroups
(Fig 5 ) .
In contrast to amyloid deposition, the mean number
of NFTs did not differ among genotype groups (F =
0.88; df = 2, 25; NS). N F T numbers in the STS correlated closely with duration of illness (Fig 6 (v = 0.75,
p < 0.001) and with last BDS score (v = 0.62, p
< 0.05). All three genotype groups showed a similar
relationship of duration with N F T formation; ie, there
was no significant interaction of apoE genotype with
NFTs (F = 0.08; df = 2, 22; NS; see Fig 6).
Discussion
The results of our analyses confirm prior reports that
apoE €4 is overrepresented in AD patients compared
with nondemented normal control subjects and patients with other neurodegenerative diseases (see Table
1) [6, 30-341. We anticipated that such a powerful
risk factor for AD would produce a malignant clinical
course. We estimated rate of progression by a measure
of cognitive function based on examining the patient
(BDS) and by a measure of functional activity based
on information related by a caregiver (ADL). By both
measures, the results of our study clearly rejected the
hypothesis that apoE €4 would be associated with a
more rapidly progressive course. Despite the association
of apoE €4 with high risk of developing AD and relatively younger age of onset, there were no associations
between apoE genotypes and the rates of dementia progression. ApoE €4 was associated with a longer duration of illness, b u t our analysis suggests that this simply
reflects the earlier onset rather than a direct effect of
apoE €4. Corder and associates [35] found that survival
from onset decreased with increasing age of onset but
was unrelated to apoE €4 dose. Thus, apoE genotype
does not appear to affect the course of the disease once
the AD process has begun. Our results are in direct
contrast to the recent study of Frisoni and colleagues
[36], who report a slower progression for individuals
with the apoE €4 allele. Their study, of 62 patients,
used a retrospective assessment of rate of progression
by estimating a Mini-Mental State examination score
at baseline based on age and education level, then estimating duration of illness and calculating a yearly rate
of change. By contrast, our study assessed more than
150 patients who were followed prospectively for an
average of more than 30 months, with rates of progression calculated directly.
Even though apoE genotype does not influence the
course of dementia, we did find that having the apoE
€4 allele is associated with a significantly lower age of
onset of dementia in a clinic-based population, in accord with other observations [4, 8, 29, 37, 381. It has
recently been suggested that men and women differ in
risk for developing AD if they have only one copy of
the apoE €4 allele [39]. Specifically, inheritance of one
Gomez-Isla et al: ApoE
&4 in Alzheimer’s Disease 65
1 . 0 +-------**---I
I
I
I
I
2*000*11
000111
2** 1
0011
P
I
1001
T
I
22**11
1 . 0 +------**---I
I
o 0.8 +
1001
P
I
2 0
0
1
i
1
0
1
n 0.6
+
rt
I
Il
I
2
I
f
a
f
1!
I
e
l
c
l
1
*
I1
n
I
1
22
n
1
0
1
l
50
60
80
70
90
j E 1 O0-0
2
l
C
I
d
11
1111
I
e
1
I
0.0
11
1
22
1
I
I
0
00
00
11
00
11
0-0
2---211
I
221-11 00
2
1-l*
2
I
II
+
+---+-----+--~-~+-----+-----+----+----+----+
+---+-..---+-----+-----+-----+----+-------+-----+
40
+
1
t 0.2 +
e
l
i1 O:0
11 00
22
1
0-0
222 11
0
22 111 00
22
11-100
2
110----0
0.0 +
I
0.4
f
f
2
I
I
1
I
I
I
I
**
22 100
2 1 1
I 1100
22 1 0
221 1
I
"
100
11000
I
I
2 **o
2-2 10
2 1*0
I
I
I
a
2
:2i;
I
I l l
I 10
221*0
t 0.2 +
e
l
d
I
22**
22**1
o 0.6 +
1*0
22 11
0.4 +
n
1
I
0
r
t
i
2 2 00
1
u
*o
I
22 10
2 1*
1
1
P
o--*
I
o 0.8 +
22 1*0
1
I
r
I
I
P
**
2 2I
2000-*11
00*0*11
I
100
50
40
110
60
70
80
90
100
13.0
Age at Onset of Dementia
Age at onset of Dementia
A
B
1.0 + - - - - - - - * * - - a
1*--0*1
I
**0-1
I
2* 11
I
P
I
0
0.8 +
P
I
0
1
r
t
i
I
I
r
I
00 11
20*011
20-*
I
I
I
l
22
0.6 +
n
l
I
*
I
l
0
*1
00
22 **
22 **
I
I
I I1
2 2 *1
2 2 11
u
0.4 +
n
1
a
f
f
e
j 011
l
l
I001
2 111
c
!
0.2 +
e
d
l
l
t
2;!1
1
I 1 1
220-*-0
2
1 00
I
0.0
I
1-10--0
1
0-0
22 11
000
22
!
I
22
2
+
11---1
1
o----o
0
+---+-----+-----+-----+-----+-----+-----+---~-+
40
50
60
70
80
90
100
110
A g e at Onset of Dementia
c
Fig 1. Kaplan-Mmr curves o f tha propoYtton unaffected by dementia of men and women (A), women (B), and men (C). 0
apolzpoprotein E ( q o E ) €4 alleles; 1 = 1 apoE &4alleles; 2 = 2 apoE &4alleles. See text f o r detatls.
66 Annals of Neurology Vol 39 No 1 January 1996
= 0
h
20 +
T
6!
5
0
3
18
4
I
I
I
0
I
I
I
0
+
16 +
0
0
0
I
a2
0
c)
14 +
2 3
a2
D
”e 2
u
A
* 1
rn
0
T
3.3
3.4
I
I
I
10 +
I
I
I
I
N
n
0
12 +
R
cd
E
0
8 +
P a 0
*.2
3.3
3.4
4.4
6 +
I
Apo E Genotype
I
I
A
4 +
I
I
I
2 +
T
o +
4.4
GENOTYPE
Fig 3. “Box-Whisker Plots ” disphying diitributions of scores
f i r duration of illness by apolipoprotein E genotype. The bottom and top edges of the box are the 25th and 75th percentiles, respective&, with “whiskers” extending as f a r as the data
extend, up to 1.5 interquartile ranges from the box; values
more extreme up to 3 interquartile ranges are indicated with
*.2
3.3
3.4
4.4
a zero, and scores still more extreme with an asterisk. The
central horizontal line represents the sample median, and the
plus sign (+) the sample mean.
Apo E Genotype
B
Fig 2. Lack of effect of apolipoprotein E (apoE) genoype on
rate of clinical progression, measured as change on the Blessed
Dementia Scale (BDS) (A) and Activities of Daily Living
(ADL) (B) score (both measured in points per year). All
patients who had at least three scores obtained approximately
6 months apart (average fallow-up, 31.8 months) were analyzed. Groups consisted of 13 apoE E -/2 individuals (includindividuals, and 6
ing 2 apoE ~ 2 / 2individuals, 5 apoE ~ 2 / 3
apoE E 3 4 individuals), 45 apoE &3/3 individuali, 63 apoE
~314individuals, and 32 apoE ~ 4 / individuali
4
f i r the BDS
and 13, 40, 59, and 29, respectivebi, .fir ADL scores. No signgcant differences among genogpes were present.
copy of the apoE ~4 allele confers risk in women but
not in men, although this result was not seen in another population [40]. Comparison of the KaplanMeier curves of the proportion unaffected from men
or women in our study shows that the risk for AD
increases in women from apoE &3/3 to apoE ~ 3 1 4to
,
apoE ~ 4 1 4 ,but the risk for AD was not different in
men between apoE ~ 3 1 3and apoE ~ 3 1 4 ,with both
differing from apoE r 4 / 4 (see Fig 1). Caution must be
used in interpreting these dara, however, because we
studied fewer men than women, and the great majority
of our older AD patients were women. Populationbased epidemiological studies will be needed to determine whether the apoE ~44associatedrisk of developing AD differs between men and women.
Our neuropathological findings dovetail with the absence of apoE effects on the rate of dementia progression. Amyloid deposition occurs early in the brains of
AD patients [26]. Further, point mutations in the amy-
Gomez-Isla
et
al: ApoE
~4 in Alzheimer’s Disease 67
20000
1
h
%
10
5
a
8
m
6
I
N=30
R=0.75
NFTs
2
N=29
10000 -
E"
4
3.3
3.4
4.4
3
2
0
A
2.3
2.4
4
2
5
0
3.3
3.4
10
Duration
4.4
Apo E Genotype
1s
20
(years)
Fiz 6. Duration of dementia correlates stron,& with neurofibrillmy tangle (NFT] formation in the superior temporal SuLcus. Alzheimer j disease duration is signtJicantly correlated
with NFT numbers jor all groups together (r = 0.,75,
p <
0.001). Moreover, each of the three genotype group! showed
similar slopes that were not signzj5cantly different from one
another. Cases described in Figure 2, except one apolipopro4 had technically inadequate immunostaining
tein E ~ 3 / case
I_
Fig 4. Apolipoprotein E (apoE) ~4 is associated with
increased AP. Data represent m a n -C SEM values for 10
apoE &3/3, 12 apoE &3/4,and 7 apoE &4/4brains. Amyloid
burden, the percentage of cortical area occupied by AP,
increases signijicaritly with inheritance of more apoli € 4
alleles (F = 8.24; df = 1,26; p < 0.01).
I
h
0
e,
3F4
12
N=31
R=0.004
NS
8
s
2
4
4
I
I
.,
0
?
10
Duration
15
20
(years)
Fig 5. No correlation is seen between duration o f dementia
and amyloid burden in the sziperior temporal sulcus (STS).
There was no significant correlation of illness duration with
A/3 deposition when data collapsed across genoppes (r =
0.004) or viewed as individual genoppe groups. Cases
described in Figure 2 plus I apolipoprotein E (apoE) €21.3
and I apoE ~ 2 / 4brains.
loid precursor protein (APP) gene on chromosome 21
are sufficient to cause AD in a small number of families
[41].However, in the vast majority of AD cases without abnormalities in the gene encoding APP, the link between the presence or amount of amyloid in the brain
and clinical measures of dementia is tenuous [42],and
numerous studies have failed to find correlations between
dementia severity and amyloid deposition [26, 27, 43,
441. Instead, the amyloid burden appears to reach a steady
state in both AD [27]and Down's syndrome (trisomy21)
ofadvanced age [45],implying the presence of a clearance
68 Annals of Neurology Vol 39 No 1 January 1996
mechanism for AD or AP aggregates. We postulate that
apoE may play a role in clearing AP from the brain because
apoE is an amphipathic protein known to bind to hydrophobic and amphipathic substances (including AP [7,
14,461) and direct them to specific receptors for clearance
[5,471.
In contrast to amyloid, indices of neuronal dysfunction and death do correlate significantly with clinical
manifestations of AD dementia. Arriagada and coworkers [26] found that the number of cortical NFTs
wdS positively correlated with both duration of illness
and with the IMC-BDS score. Terry and colleagues
[48] and DeKosky and associates [49] reported that
the critical lesion underlying dementia in AD was a
decrease in the number of synapses and found a significant correlation between worsening of the IMCBDS score and a decrease in the number of synapses
in frontal lobe of AD brains. Thus, both SPs and NFTs
are characteristic AD lesions; but only indices of neuronal damage such as NFTs bear a proximal relationship to clinical measures of AD severity. Based on these
facts, it would have been surprising if apoE genotype
had influenced the rate of dementia progression given
an influence on amyloid deposition but not on NFT
formation.
Strittmatter and co-workers [ 151 postulated that
apoE interacts with T, and apoE4 (compared with
apoE3) impedes normal microtubule function and fails
to protect T against the hyperphosphorylation associated with NFT. If this hypothesis were correct, we predict that apoE ~4 would be accompanied by increased
NFT formation. Since NFTs correlate with duration
of illness and apoE ~4 is associated with younger onset
and longer durations, it seemed possible that NFT
counts would be higher with apoE ~4 via an indirect
interaction. Two studies using the Braak and Braak
staging scheme, which did not take varying duration
into account, found increased NFTs with apoE ~4 [50,
5 11. Our covariant analysis asked whether NFTs correlated with apoE genotype independently of duration
of illness, and we found no evidence for such a correlation. This lack of correlation among NFTs and apoE
genotypes argues against the idea that apoE4 protein
is directly or indirectly involved in NFT formation.
Our findings provide a framework for generating
and testing hypotheses regarding the biological mechanisms that underlie apoE’s effects. Our studies confirm
the observations that the primary clinical correlates of
inheritance of apoE ~4 are increased risk for AD and
earlier age of dementia onset [3-10, 371. The primary
neuropathological correlate of apoE ~4was increased
cortical AP deposition rather than an influence of
NFTs. The strong correlation of apoE ~4 with increased amyloid burden confirms several reports that
suggest that apoE ~4 is associated with AP deposition
in AD [5, 52-55], trisomy 21 (Down’s syndrome) [45,
551, amyloid angiopathy [56],diffuse Lewy body disease [54],and normal aging [57].
These data are most consistent with hypotheses implicating apoE/AP interactions [5, 12, 131. In accord
with this formulation, a lower age of onset and increased AP deposition are the apparent consequences
of alterations in APPl AP metabolism in both trisomy
21 (Down’s syndrome) and familial AD associated with
APP mutations [41, 58-60]. We postulate that apoE
~ 4 ’ smain biological effect in AD is to influence APP
metabolism or AP deposition. The amount of AP deposited could be influenced by either enhanced production, enhanced aggregability, or diminished clearance of AP. We postulate that the apoE4 protein
enhances the initial steps of amyloid deposition in AD,
perhaps by altering clearance efficiency [5, 551.
Supported by N I H grants AG08487,
AG12406.
P50 AG05134, and
We thank Beth Souza for assistance with database management. We
thank the Brain ’Tissue Resource Center, McLean Hospital (grant
no. M H / N S 31862), and the Massachusetts Alzheimer Disease Research Center Brain Bank (Dr E. T. Hedley-Whyte, Director) for
brain tissue specimens.
References
1. Mahley RW. Apolipoprotein E: cholesterol transport protein
with expanding role in cell biology. Science 1988;240:622630
2. Ordovas JM, Litwack-Klein L, Wilson PWF, et al. Apolipoprotein E isoform phenotyping methodology and population
frequency with identification of apoEl and apoE5 isoforms. J
Lipid Res 1987;28:371-380
3. Saunders AM, Strittmatter WJ, Schmechel D, et al. Association
of apolipoprotein E allele ~4 with late-onset familial and sporadic Alzheimer’s disease. Neurology 1993;43:1467- 1472
4. Poirier J, Lhvignon J. Bouchillier 0 , et al. Apolipoprotein E polymorphism and Alzheimer’s disease. Lancet 1993;3
5. Rebeck GW, Reiter JS, Strickland DK, Hyman
protein E in sporadic Alzheimer’s disease: allelic variation and
receptor interactions. Neuron 1993;11:575-580
6. Mayeux R, Stern Y.Ottman R, et al. The apolipoprotein E4
allele in patients with Alzheimer’s disease. Ann Neurol 1993;
34:752-754
7. Strittmatter WJ, Saunders AM, Schniechel D, et al. Apolipoprotein E: high-avidity binding to P-amyloid and increased frequence of type 4 allele in late-onset familial Alzheimer disease.
Proc Natl Acad Sci USA 1993;90:1977-1981
8. Tsai M-S, Tangalos EG, Petersen KC, et al. Apolipoprotein E:
risk factor for Alzheimer’s disease. Am J Hum Genet 1994;54:
643-643
9. Chartier-Harlin M-C, Parfitt M, Legrain S, et al. Apolipoprotein E, ~4 allele as a major risk factor for sporadic early and
late-onset forms of Alzheimer’s disease: analysis of the 19ql3.2
chromosomal region. H u m Mol Genet 1994;3:569-574
10. Hendrie HC, Hall KS, Hui S, et al. Apolipoprotein E genotypes and Alzheimer’s disease in a community srudy of elderly
African Americans. Ann Neurol 1995;37:118-120
11. Strittmatter WJ, Weisgraber KH, Huang D, et al. Binding of
human apolipoprotein E to PA4 peptide: isoform specific effects and implications for late onset Alzheirner disease. Proc
Natl Acad Sci USA 1993;90:8098-8102
12. Ma J , Yee A, Brewer AYH Jr, et al. Amyloid-associated proteins
a,-antichymotrypsin and apolipoprotein E promote assembly of
Alzheimer P-protein into filaments. Nature 1994;372:92-94
13. Wisniewski T , Frangione B. Apolipoprotein E: a pathological
chaperone protein in patients with cerebral and systemic amyloid. Neurosci Lett 1992;135:235-238
14. Wisniewski T, Golabek A, Matsubara E, et al. Apolipoprotein
E: binding to soluble Alzheimer’s P-amyloid. Biochem Biophys
Res Commun 1993;192:359-365
5. Strittmatter WJ, Weisgrabber KH, Goedert M, et al. Hypothesis: microtubule instability and paired helical filament formation in the Alzheimer disease brain are related to apolipoprotein
E genotype. Exp Neurol 1994;125:163-171
6. Han S-H, Einstein G, Weisgraber KH, et al. Apolipoprotein
E is localized to the cytoplasm of human cortical neurons: a
light and electron microscopic study. J Neuropathol Exp NCLIrol 1994;53:535-544
17. McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA
group under the auspices of Department of HHS Task Force
on Alzheimer’s Disease. Neurology 1984;34:939-944
18. Blessed G , Tomlinson BE, Rorh M. The association between
quantitative measures of dementia and of senile change in the
cerebral grey matter of elderly subjects. Br J Psychiatry 1968;
114:797-811
19. Weintraub S. The record of independent living: an informantcompleted measure of activities of daily living and behavior in
elderly patients with cognitive impairment. Am J Alzheimer
Care 1986;1:35-39
20. Folstein M, Folstein S, McHugh P. “Mini-mental state”: a
practical method for grading che cognitive state of patients for
the clinician. J Psychiatr Kes 1975;12:189-138
21. Khachaturian 2s. Diagnosis of Alzheimer’s disease. Arch Neurol 1985;42:1097-1105
22. Hyman BT, Tanzi RE, Marzloff KM, et al. Kunirz prorease
inhibitor containing amyloid precursor protein immunoreactivity in Alzheimer’s disease: a quantitative study. J Neuropathol
Exp Neurol 1992;51:76-83
23. Vincent I, Davies P. A protein kinase associated with paired
Gomez-Isla et al: ApoE
E4
in Alzheimer’s Disease
69
helical tilaments in Alzhci tiier’s disease. I’roc Natl Acad Sci
USA 109L;89:2X7X-7XH2
24. Yeterian E. Pandya D. l ‘ h e a w x i i t i o n cortices. In: Jones EG,
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
Peters A, eds. Ccrcbrd Cortex, vol. 5. New York: Plenum Press,
1989: 1-53
Arnold SE. Hynian RT. Flory 1. et al. l’he topographical and
neuroanaromical distrihurion of neurofihrillary tangles and
neuritic plaques i n ccrehr;il cortex o f patienrs with Alzheimer’s
disease. Cereb Cortex IC)O 1 ;1 : 10-3- 1 1 (I
Arriagada PV, G r o ~ ~ * d1H.
o n Hedlry-\)“hyte ET, Hynian RT.
Neurofihrillary rmgle.; hut not senile plaques pxallel duration
and severity of Alzlieimrr disease. Neurology l092;42:631-639
Hvrnan BT, hlarzloff K. Ai-ri‘ig.ida I’V. ‘The lack of accumulation o f w i d e plaquec OK .imyloid hurden i n Alzheimer’s discase
suggest5 a dynamic h.ilaiice between amyloid deposition a n d
resolution. 1 Neurop.ithol EYF Neurol 1993:52:594-600
Wesr M. New stcrcoIogiCnl methods for counting neurons.
Ncurobiol Aging 1093;14:275-285
Lucorrc G, Turpin 1-C, \.andais P. Apolipoprotein E €4 allele
doses in late onset Al~ihcimcr’sdise.ise. Ann Neurol 1994;36:
681-687
Saunders A M . Schmadrr K, Brritner JCS, e t al. Apolipoprotein
E E4 allele distributions in late-onset Alzheimer’s disease and
in orher amyloid-formiii~diseases. Lancet 1993;342:710-7 1 I
Roses AD. Apolipoprotein E affects t h e rate of Alrhcimer disease cxprrshion: I>-amyloid burden is a secondary coiisequcnce
dependent on APOE genotype a n d duration of disease. J Ncuropathol Exp NeLirol 1014:53:429-437
Mui S, R e h d GW. M c K r n n a - Y d i D, et al. Apolipoprotein
E E 4 allele is nix associared with earlier age of onset in sporadic
or Familial nmyotropliic later.11 sclerosis. Ann Neurol 1995;38:
460-463
Gearing M, Schneider I/\. Rebeck GW. et al. Alzheimer‘s disease with and wirhout coexisting Parkinson’s disease changes:
apolipoprorein E genotype md neurop.itliologic:11 correlatec.
Neurology 19% ( i n press)
Koller WC, Glart S1.. Hubble JI’. et al. Apolipoprotein E genoy p e s in Parkinwn’s disease with and without demenria. A n n
rrirtmatter W’, et al. Apolipoprotein
E. survival in Alzheimer’s di
e patients. nnci t h e competing
risks of death and hlzheimer’s di
e. Neurology 1995;45:
1323-1328
36. Frisoni G,Govoni S , Geroldi C , et aI. Gene dose of the epsilon
4 allele of q>olipoprotrin E and disease progression i n lateonser Alzheimer’s diseac. A n n Neurol 1995;37:596-604
37. Corder E H , Saunders AM, Strittniatter \VJ. et al. Gene dose
of apolipoprorein E type 4 allele and the risk of Alzheimer’s
disease in I.ite otiscr f h i l i e s . Science 1993;261:921-923
38. Corder E H , Saundrrs A M , Ilisch NJ, et al. Apolipoprorein E
type 2 allele decreases thr risk of latr onset Alzheimer disease.
40.
41.
42,
43.
44.
et aI. Al/.heinier’s discase. apolipoprotein E .and gender. IAMA ~ ~ ) ‘ ) ~ : i 3 I 6 - 1 3(Lertcrj
i7
Corder E, Saunders A. Srrirtnimer W. et ;iI. T h e apolipoprorein E E4 allclc :ind scx-specific risk of Alzheimer‘s disease
JAMA 1995;27.3:373-374 (Lrtterj
Goarc A, C h x r i c r - H d i n M-C, hlullan M. Segreg.ation of a
missrnse m u t . ~ t i o ni n thc ;itmyloid precursor protein gene wirh
h n i l i a l Alzheimer’s dise.iw. Nature 199 1 ;349:704-707
Hyman BT. Stiidyins the Alzlieinier’s disease hrain: insights.
p i i u l e s . and opporttiniries. Neurohiol Aging 1004: 15:579-583
Berg L, McKrt.1 DW. Miller JP. e t al. Neuropathological indexes of A l z h r i t i i C r ’ c disc.ise i n denierired and nondeniented
persons aged 80 r?nd older. Arch Neurol 1993;50:349-358
Bouras C . H o f 1’. (;iantiaho~~oulosP. cr al. llegional distribu-
70 Annals of Neurology
Vol 39
No I
January 1996
tion of neurofihrillary tangles a n d senile plaques in the cerehral
corcex of clderly patients: a qiiantitative evaluation of a oncyear duropsy population in a geriatric hospital. Cereb Cortex
1994;4:138-150
45. Hynian BT, West HL. Rebeck GW. rt al. Neuropathological
changes i n Down’s syndrome hippocampal formation: effect of
age arid apolipoprotein E genotype. Arch Neurol 1995;52:
373-378
46. LaDu MJ, Falduro MT. hlanelli AM. et al. Isoform-specific
binding of apolipoprotein E to p-amyloid. 1 Biol Chem 1994;
269:23403-23406
47. Rebrck G W , H x r SD. Strickland DK. Hynian RT. h4ultiplr.
diverse senilc plaque-associated proteins arc ligandc of a n apolipoprotein E receptor. the a2-macroglobulin receprorllowdensiry-lipopi-otcin receptor-related protein. Ann Neurol 19%;
3721 1-2 17
48. Terry RD, hlasli‘ih E. Salmon DP. Physical basis o f cognitive
alterations i n Alzlieirner‘s disease: synapsc loss is the major
correlate of cognitive impairment. Ann Ncurol. 1091 ;41:
572-580
S T . Scheff SW. Synapse loss in frontal cortex biopsies
in Alzheimer‘s discase; correlation with cognitive severity. Ann
Neurol 1 9OO;?7:457-464
50. Hansen L, Galasko D, Samurl W. et al. Apolipoprotein E €4
is associared with increased neurofihrillary tangles in the Levq
body variant of i\l/.lieimer’s disease. Neurosci Lett 1394;182:
63-65
51. Ohm T, K m a h4. Bohl J, et al. Apolipoprotein E polymorphism influences nor only cerebral senile pl‘ique load but also
Alzheimer-type ncurofibrillary tangle formation. Neuroscience
1995;66:583-587
52. Schmechel DE, Saunders Ahl. Strittniatter WJ. et al. increased
amyloid P-pepride deposition as a consequence of apolipoprotein E genotype in late-onset Alzheimers disease. Proc Narl
Acad Sci USA 1993;~)0:‘~649-9653
53. Zubenko GS, Stiffler 5. Stabler S. et al. AssociJtion of t h e
apolipoprorein E epsilon 4 .illele with clinical subtypes of autopsy-confirmed Alzheinicr‘~disease. A m 1 Hum Genet 1994;
54:139-205
54. Harrington CR. L.ouwagie J. Rossau R. et 11. Influence of apolipoprorein E genotype on senile dementia of the Alzheimer
and I.ewy body types. Significance for etiological theories of
Alzheimer’s diseasc. Am 1 Pathol 1994:145:1472-1484
5 5 . Hynian BT. W’esr 1 IL, Kebeck GW. et al. Qirantitative analysis
of senile plaques in Alzheinier disease: ohsewation o f log-norma1 size distribution and irioleciilar epidemioloby o f differences
associated with apolipoprotein E genotype and trisomy 21
(Down syndrome). Proc Natl Acad Sci USA 1995;92:35863590
56. Greenberg S. Rehcrk GW, Vonsattel JPG, et al. Apolipoprorein E €4 and cerebral hemorrhagr associated with amyloid
angiopathy. Ann Neurol 1995;38:254-259
57. Bcrr C , H ~ L I JJ.
W Delaere 1’. et al. Apolipoprotein E allele
epsilon 4 is linked to increased deposition of the amyloid betapepride (A-heta) i n cases with or withour Alzheimer’s disease.
Ncurosci Lcrr 1004: 178:221 -??it
58. Cai X-D, Golde TE. Younkin SG. Release of excess amyloid
P protein from a niiitant aniyloid TJ protein precursnr. Science
199?;255):514-51 6
59. Citron M, OlrersdorC T. Haass C. et al. Aluration o f the panyloid prccursor protein in hmilial A17heimer’s disease increases 0-protcin production. Nature 1992;360:072-674
60. Suzuki N , Cheung’IT. Cai X-D, et al. ,411increased percentage
o f long aniyloid 0 protein secreted by hniilial aniyloid p protein precursor (PAI’P717) m u t a n t s . Science 1994;264:13361.340
Документ
Категория
Без категории
Просмотров
2
Размер файла
914 Кб
Теги
pathologic, clinical, correlates, apolipoprotein, disease, alzheimers
1/--страниц
Пожаловаться на содержимое документа