вход по аккаунту


Autoantibodies against -amyloid are common in Alzheimer's disease and help control plaque burden.

код для вставкиСкачать
Autoantibodies against ␤-Amyloid Are
Common in Alzheimer’s Disease and Help
Control Plaque Burden
Alexander Kellner, MSc,1 Jakob Matschke, MD,1 Christian Bernreuther, MD, PhD,1 Holger Moch, MD,2
Isidro Ferrer, MD,3 and Markus Glatzel, MD1
Objective: Active or passive immunization of Alzheimer’s disease (AD) patients leads to targeting of ␤-amyloid plaques by
immunoglobulins (IgG) and their subsequent removal by microglia. Here, we investigate whether naturally occurring autoantibodies to ␤-amyloid contribute to ␤-amyloid plaque removal in nonimmunized AD patients.
Methods: We generated an AD tissue microarray with 2,325 tissue specimens from 3 defined central nervous system regions of
48 AD patients and 48 age-matched control patients. Absolute quantification of ␤-amyloid, ␤-amyloid plaque-bound IgG, and
phagocytic, resting, and activated microglia and microhemorrhages was done using a standardized, highly reproducible scoring
Results: The majority of neuritic plaques are decorated by IgG. AD patients with prominently IgG-labeled neuritic plaques have
a significantly reduced plaque burden and an increase in phagocytic microglia, yet no increase in microhemorrhages.
Interpretation: Autoantibodies directed against ␤-amyloid are common in AD patients and may contribute in controlling
plaque burden.
Ann Neurol 2009;65:24 –31
Alzheimer’s disease (AD) is the leading cause of dementia in adults and affects approximately 5% of people older than 65 years and 20% of people older than
80 years.1 Deposition of ␤-amyloid has a central role
in the pathophysiology of AD and correlates with cognitive decline.2,3 Although AD is a degenerative disease, the importance of immunological reactions in its
pathophysiology is obvious. Many studies have shown
that microglia are attracted to ␤-amyloid and may help
to eliminate ␤-amyloid deposits.4,5 Therapeutic options in AD are limited, yet recent studies have shown
that both active and passive immunization against
␤-amyloid can effectively reduce plaque burden and
may slow cognitive decline in a subset of patients.4,6 – 8
In immunized AD patients, removal of ␤-amyloid is
thought to be mediated by phagocytic microglia.9 –11
Recent postmortem studies on ␤-amyloid–immunized
individuals have shown a dramatic reduction of the
plaque burden when compared with nonimmunized
individuals. Plaques of ␤-amyloid–immunized patients
are decorated by IgG.9,10,12,13 This was interpreted as a
sign of ␤-amyloid antibody-dependent activation of
immune response. However, the significance of this
finding remained unclear because similar responses
could be observed in control patients.12 Studies on immunized patients and AD mouse models raised concerns that ␤-amyloid immunization may lead to an increase in cerebral microhemorrhages.10 These effects
are thought to be caused either by a weakening of the
vessel wall through local inflammation or disturbances
in the blood–brain barrier.14
Naturally occurring antibodies against ␤-amyloid are
present in the sera of AD patients and in nondemented
individuals alike.15 Although these antibodies inhibit
␤-amyloid aggregation in vitro and decorate ␤-amyloid
plaques of AD patients,15 it is still unclear whether
these antibodies can effectively reduce amyloid plaques
in humans.16
In this study, we investigate the relation between
IgG decoration of ␤-amyloid plaques and plaque load
in AD patients and age-matched nondemented control
patients. We correlate these data to immune activation
From the 1Institute of Neuropathology, University Medical Center
Hamburg-Eppendorf, Hamburg, Germany; 2Department of Pathology, Institute of Surgical Pathology, University Hospital Zurich,
Zurich, Switzerland; and 3Institute of Neuropathology, Hospital
Universitari de Bellvitge, Hospitalet de Llobregat, Barcelona, Spain.
Potential conflict of interest: Nothing to report.
Address correspondence to Dr Glatzel, Institute of Neuropathology,
University Medical Center Hamburg-Eppendorf, Martinistrasse
52, D-20246 Hamburg, Germany.
Additional Supporting Information may be found in the online version of this article.
Received Dec 19, 2007, and in revised form Jun 17, 2008. Accepted for publication Jun 20, 2008.
Published online in Wiley InterScience (
DOI: 10.1002/ana.21475
© 2009 American Neurological Association
Published by Wiley-Liss, Inc., through Wiley Subscription Services
of the brain and the presence of microhemorrhages using a highly standardized scoring system on a tissue
microarray platform (TMA). Our results indicate that
IgG decoration of neuritic ␤-amyloid plaques is a common feature in AD patients and may contribute to reducing plaque burden in these individuals.
Patients and Methods
Alzheimer’s Disease Tissue Microarray
We selected anatomically defined areas (gyrus frontalis medius, temporal cortex at the level of the lateral corpus geniculatum, and entorhinal cortex) from a total of 192 formalinfixed (buffered neutral aqueous 4% solution), paraffinembedded tissue specimens for TMA generation. The use of
these specimens in research was in agreement with local ethical standards at the Zurich University Hospital and at the
University Medical Center Hamburg-Eppendorf. The TMAs
were constructed to test the diagnostic value of antibodies
mentioned later, as well as for quality assurance of these antibodies in brain tissue. All specimens were reviewed by at
least two neuropathologists (J.M., M.G.) and characterized
according to current AD classification schemes, and the exact
site of tissue removal was determined.17,18 The TMA composition is given in the Table. TMA construction was as described elsewhere.19 In brief, tissue cylinders with a diameter
of 0.6mm were punched from anatomically defined areas of
a “donor” tissue block using a semiautomatic robotic precision instrument and brought into 4 different recipient par-
Table. Demographic and Neuropathological Data on
Patients Included in Alzheimer’s Disease Tissue
Age at death, yr
Sex ratio, M:F
80.3 ⫾ 10.7
77.5 ⫾ 10.9
Braak and Braak staging
AD ⫽ Alzheimer’s disease; CERAD ⫽ Consortium to
Establish a Registry for Alzheimer’s Disease.
affin blocks, each containing 576 samples. Multiple 4␮m
sections of the resulting TMA block were cut, mounted to
an adhesive-coated slide system, and further processed for
histological staining according to standard protocols. For
analysis of antibody specificity in sera of AD patients and
studies on human sera/matched central nervous system tissue
(parietal cortex), we cut 3␮m paraffin sections from agematched, aged mice expressing human amyloid precursor
protein containing the Swedish mutation (APPsw), and control mice for histological processing (see later).20 Parietal cortex samples and serum (n ⫽ 8; definite AD according to
current AD classification schemes; 2 women, 6 men; average
age, 76.9 years; standard deviation, 5.6 years; use of these
tissues in research is in agreement with local ethical standards
at the Universitari de Bellvitge, Spain) were processed as described earlier for tissues used for TMA generation.
Histological Assessment
Prussian blue staining of ferric iron and Bielschowsky staining were performed according to standard protocols. Immunohistochemistry was conducted using appropriate antigen
retrieval methods with internal controls according to published protocols.21 The following primary antibodies were
used: ␤-amyloid (1:300; M0872; Dako, Glostrup, Denmark), human leukocyte antigen-D related (HLA-DR;
1:100; M0775; Dako), IgG (1:1,000; A0262; Dako), CD68
(1:50; 2164; Immunotech, Fullerton, CA). Primary antibodies were visualized using a standard diaminobenzidine
streptavidin-biotin horseradish peroxidase method (Histofine
simple Stain Max PO; Nichirei Bioscience, Nichirei, Japan)
(A␤, HLA-DR, IgG) or the alkaline phosphatase method
(Histofine Simple Stain AP; Nichirei Bioscience) (CD68).
For double-immunofluorescence staining, unfixed snapfrozen frontal cortices (whole brains for APPsw and wildtype mice) were cut at a thickness of 5␮m, fixed in acetone,
blocked with donkey serum (Sigma, Munich, Germany),
stained with 6E10 (1:20; Signet, Dedham, MA) and adequate Alexa 555–labeled secondary antibody and/or fluorescein isothiocyanate–labeled anti–human IgG (1:4; Sigma)
(to test specificity of IgG staining) and serum from a patient with a high or a low IgG plaque labeling index and
adequate fluorescein isothiocyanate–labeled secondary antibody (1:5 anti–human IG; Sigma) (to test for autoantibody
specificity). Nuclei were counterstained with 4⬘,6-diamidino2-phenylindole (DAPI; Sigma).
Quantification of diffuse plaques, neuritic plaques, ferric
iron (Fe) deposits, CD68-positive cells, and HLA-DR–positive cells was accomplished by counting positive signals for
IgG, ␤-amyloid, Fe (with an amyloidogenic core in the case
of neuritic plaques and IgG-decorated neuritic plaques), or
stained cells for CD68, HLA-DR on consecutive sections
(human samples), or double-immunofluorescence staining
(murine samples). The presence of neuritic plaques was confirmed using Bielschowsky staining. The average of positive
signals/cells was determined for each region of AD/control
samples or entire murine brain sections by calculating the
mean value of all assessed tissue specimens. For AD-TMA,
7.04 tissue specimens were analyzable per region on average.
For analysis of murine samples, we assessed only intact sections of entire brains. For TMA, a region was considered
Kellner et al: Autoantibodies in AD
analyzable if at least two tissue specimens could be assessed.
A total of 23 regions had to be excluded from TMA analysis
because of poor quality.
We used Excel (Microsoft 2003; Microsoft Corp, Redmond,
WA) and Origin 8 for statistical analysis. The Mann–Whitney U test was used to compare diffuse and neuritic plaques,
Fe, CD68, and HLA-DR–positive cells of frontal, temporal,
and entorhinal regions of AD and control patients. The limit
of significance for all analyses was defined as a p value less
than 0.01.
Alzheimer’s Disease Tissue Microarray Is Able to
Discriminate between Alzheimer’s Disease and
Nondemented Control Specimens
The Table provides a summary of the neuropathological and demographic characteristics of the AD and
control patients included in the study. One aim of the
study was to generate an AD-TMA, which would allow
for high-throughput assessment of plaque load, IgG
deposition, immune activation, and microhemorrhages
on consecutive sections of identical brain regions. Tissue cylinders were punched from anatomically defined
areas within paraffin blocks of gyrus frontalis medius,
temporal cortex at the level of the lateral corpus
geniculatum, and the entorhinal cortex. A total of eight
tissue cylinders were punched from each region, and
punches were randomly taken from the six layers of the
cortex. The design of the AD-TMA allowed for absolute quantification of respective immunohistochemical
signals in each tissue punch, minimizing variations
caused by antibody binding efficiencies or differences
in visualization reaction (Fig 1). For AD, the median
number of diffuse and neuritic plaques/neuritic plaques
alone per tissue punch was 16.06/0.58 (first quartile,
7.33/0.25; third quartile, 32.5/1.06) for frontal cortex,
10.35/0.5 (first quartile, 6.67/0.25; third quartile,
15.94/1.06) for temporal cortex, and 5/0.0 (first quartile, 2.29/0.0; third quartile, 10.87/0.25) for entorhinal
cortex. The respective values for control patients were
0.0/0.0 (first quartile, 0.0/0.0; third quartile, 0.0/0.0)
for frontal cortex, 0.0/0.0 (first quartile, 0.0/0.0; third
quartile, 2.75/0.0) for temporal cortex, and 0.0/0.0
(first quartile, 0.0/0.0; third quartile, 0.375/0.0) for
entorhinal cortex (see Figs 1A, 2A). All observed differences were statistically significant ( p ⬍ 0.01, Mann–
Whitney U test).
Prominent Activation of the Immune System in
Alzheimer’s Disease
Immune activation is a known feature of AD.22 The
standardized tissue size and the identical staining procedure within AD-TMA allowed for the absolute quantification of immune activation by counting phagocytic
microglia (immunoreactive against CD68) and resting/
activated microglia (immunoreactive against HLADR). This analysis confirmed prominent immune activation in AD. Cell counts for microglia/phagocytic
microglia per tissue punch of AD patients were 3.25/
0.75 (first quartile, 0.5/0.38; third quartile, 8.625/
1.875) for frontal cortex, 5.26/1.23 (first quartile,
1.14/0.4; third quartile, 13.62/3.34) for temporal cortex, and 4.88/0.88 (first quartile, 1.0/0.20; third quartile, 12.5/3.07) for entorhinal cortex (see Figs 1C, 2B).
The respective values for control samples were 0.6/0.0
(first quartile, 0.0/0.0; third quartile, 3.25/0.25) for
frontal cortex, 0.63/0.13 (first quartile, 0.0/0.0; third
quartile, 4.0/0.29) for temporal cortex, and 0.63/0.0
(first quartile 0.0/0.0; 3rd quartile, 3.625/0.286) for
entorhinal cortex (see Figs 1B, 2B). All observed dif-
Fig 1. Histopathological findings of Alzheimer’s disease (AD) tissue punches. Shown are representative tissue punches from AD tissue
microarrays (TMAs) (top) and greater magnifications of areas of interest (bottom). Respective immunohistochemical or histological
stainings are indicated above tissue punches (each tissue punch has a diameter of 0.6 mm). (A) Filled arrows indicate neuritic
plaques, whereas open arrows indicate diffuse plaques; (inset) a neuritic plaque. (B) Arrowheads label phagocytic microglia; (inset)
fast red–labeled phagocytic microglia. (C) Open arrowheads point to microglia; (inset) ramified microglia. (D) Asterisk demonstrates blue-stained hemosiderin; (inset) perivascular localization of blue color reaction. (E, F) Arrows point to IgG-decorated (E)
and non-IgG–decorated (F) neuritic plaques; (insets) close-up of labeled (E) and nonlabeled (F) neuritic plaques.
Annals of Neurology
Vol 65
No 1
January 2009
Fig 2. High plaque load and prominent immune involvement
in Alzheimer’s disease (AD) patients. (A) Shown are box plots
of quantifications of diffuse and neuritic plaques (all plaques)
and neuritic plaques in frontal, temporal, and entorhinal cortices of AD and control patients, and (B) CD68 and human
leukocyte antigen-D related (HLADR)–positive microglia in
frontal, temporal, and entorhinal cortices of AD and control
patients (Cont.). Boxes encompass first and third quartile;
median is indicated as a black bar in boxes.
ferences were statistically significant ( p ⬍ 0.01, Mann–
Whitney U test).
The Majority of Neuritic Plaques Are Decorated by
IgG in Alzheimer’s Disease Patients
We counted IgG deposits localized in the direct vicinity of neuritic plaques in frontal, temporal, and entorhinal cortices (see Figs 1E, F). For AD, the median
amount of plaque-bound IgG deposits per tissue
punch was 0.27 (first quartile, 0.0; third quartile,
0.63) for frontal cortex, 0.23 (first quartile, 0.0; third
quartile, 0.69) for temporal cortex, and 0.0 (first
quartile, 0.0; third quartile, 0.31) for entorhinal cortex. The respective values for control samples were
0.0 (first quartile, 0.0; third quartile, 0.0) for frontal
cortex, 0.0 (first quartile, 0.0; third quartile, 0.0) for
temporal cortex, and 0.0 (first quartile, 0.0; third
quartile, 0.0) for entorhinal cortex (Fig 3A). Because
we used consecutive sections stained for ␤-amyloid
and human IgG, we were able to correlate the average
of IgG signals present in the direct vicinity of neuritic
plaques to the average of neuritic plaques in identical
regions, creating a value we have designated IgG
plaque labeling index. This analysis demonstrated that
the majority of plaques are decorated by IgG in AD
and control patients (0.57 for AD compared with
0.78 for control patients in frontal cortex; 0.80 for
AD compared with 0.93 for control patients in temporal cortex). Significant differences could be observed in entorhinal cortex where labeling indices for
AD were ⱖ1 for AD compared with 0.56 for control
subjects ( p ⬍ 0.01, Mann–Whitney U test) (see
Fig 3A).
To assess the specificity of IgG staining, we performed double-immunofluorescence immunohistochemistry on unfixed brain sections of AD (n ⫽ 3) and
control patients (n ⫽ 3). In both instances, we could
observe intense IgG deposition in a subset of neuritic
plaques colocalizing with ␤-amyloid, whereas another
subset of neuritic plaques or diffuse deposits of
␤-amyloid did not show IgG decoration (Fig 4). The
specificity of IgG in sera of AD patients was further
evaluated by incubating sections of a murine AD
model with serum of an individual with low and an
individual with high IgG plaque labeling index.
Whereas autoantibodies in the sera from the high IgG
patients strongly labeled the majority of ␤-amyloid
plaques, we could observe only weak labeling of a small
subset of ␤-amyloid plaques with sera from the low
IgG patient (Fig 5).
Reduced Plaque Burden in Alzheimer’s Disease
Patients with High IgG Plaque Labeling Index
Because the average IgG plaque labeling index for control patients is 0.75, we decided to group AD patients
in two cohorts: one cohort with a IgG plaque labeling
index greater than 0.75 (high IgG) and one cohort
with a IgG plaque labeling index lower than 0.75 (low
IgG). We analyzed plaque load and presence of phagocytic microglia in these two cohorts. Plaque loads were
significantly lower in the cohort with high IgG plaque
labeling index (high IgG) when compared with the cohort with low IgG plaque labeling index (low IgG; see
Fig 3B). Differences were greater for diffuse plaques
(median of 8.6 diffuse plaques per tissue punch, first
quartile 6.78; third quartile 11.33 for the high IgG
group and 14.64 diffuse plaques per tissue punch, first
quartile 9.08; third quartile 23.01 for the low IgG
group; p ⬍ 0.0349, Mann–Whitney U test) than for
neuritic plaques (median of 0.28 neuritic plaque per
Kellner et al: Autoantibodies in AD
Fig 4. Colocalization of ␤-amyloid and IgG in a subset of
neuritic plaques. Double-immunofluorescence stains illustrate
presence (E–H) or absence (A–D) of IgG decoration of
␤-amyloid–containing neuritic plaques in frontal cortices of
Alzheimer’s disease patients belonging to the low (A–D) and
high (E–H) groups. No staining for IgG or ␤-amyloid in control patient (I–L). Immunostainings are indicated above respective photographs. Bars ⫽ 10␮m. DAPI ⫽
tissue punch, first quartile 0.14; third quartile 0.5 for
the high IgG group and 0.5 neuritic plaque per tissue
punch, first quartile 0.24; third quartile 0.96 for low
the IgG group; p ⬍ 0.057, Mann–Whitney U test).
Because IgG-decorated ␤-amyloid may be subject to
elimination by phagocytic microglia, we assessed presence of phagocytic microglia by counting CD68labeled cells. Whereas the high IgG group had a median of 1.76 macrophages per tissue punch (first
quartile, 0.75; third quartile, 5.41), tissue punches of
Fig 3. Targeting of ␤-amyloid plaques by IgG. (A) Shown are
box plots of quantifications of neuritic plaques and neuriticplaque bound IgG in frontal, temporal, and entorhinal cortices of Alzheimer’s disease (AD) and control patients (Cont.),
and (B) quantification of diffuse, neuritic plaques, CD68positive microglia, vascular ␤-amyloid, and ferric iron deposits
in frontal, temporal, and entorhinal cortices of AD patients
with high and low IgG labeling indices. Boxes encompass first
and third quartiles; median is indicated as a black bar in
boxes. (C) Bar graphs of quantifications of diffuse and neuritic plaques in parietal cortex of patients who were assessed by
murine plaque–immunoreactive assays. Patients with low
plaque scores (low IgG) harbor more diffuse plaques than patients with high plaque scores (high IgG).
Annals of Neurology
Vol 65
No 1
January 2009
Fig 5. Sera of Alzheimer’s disease patient of the high IgG
group specifically reacts with ␤-amyloid in a murine AD
model. Double-immunofluorescence stains illustrate specific
binding of IgGs to transgenically expressed human ␤-amyloid
(murine AD model) present in the sera from a patient in the
high IgG group (E–H). No significant binding of IgGs to
␤-amyloid could be observed in the sera of a patient in the
low IgG group (A–D). No staining is shown in control sections (I–L, sera from patient of the high IgG group incubated
with wild-type mouse brain section). Immunostainings are
indicated above respective photographs. Bars ⫽ 10␮m.
DAPI ⫽ 4⬘,6-diamidino-2-phenylindole.
the low IgG group had a median of only 0.92 macrophage per tissue punch (1st quartile, 0.26; 3rd quartile,
2.1) ( p ⫽ 0.101, Mann–Whitney U test) (see Fig 3B).
No Evidence for Increased Microhemorrhages in
Alzheimer’s Disease Patients with High IgG Plaque
Labeling Index
Antibody-mediated clearance of ␤-amyloid may lead to
weakening of vessel walls resulting in microhemorrhages.14 We assessed the presence of microhemorrhages in AD cohorts mentioned earlier by assessing
Prussian blue stainings for ferric iron (see Fig 1D). As
expected, there was a remarkable difference in microhemorrhages between AD patients (median, 0.21; first
quartile, 0.13; 3rd quartile, 0.38) and nondemented
control patients (median, 0.13; first quartile, 0.0; third
quartile, 0.25) ( p ⫽ 0.015, Mann–Whitney U test).
For the high IgG cohort, the median amount of positive signals indicative of microhemorrhages per tissue
punch was 0.8 (first quartile, 0.04; third quartile,
0.125), whereas patients of the low IgG cohort had a
median of 0.05 microhemorrhage per tissue punch
(first quartile, 0.04; third quartile, 0.125) (see Fig 3B)
(statistically not significant). Furthermore, there was no
significant difference in the presence of vascular
␤-amyloid between patients of the low IgG (median,
0.13; first quartile, 0.0; third quartile, 0.63) and the
high IgG groups (median, 0.13; first quartile, 0.0;
third quartile, 0.14).
Murine Plaque Immunoreactivity Assays as a Tool to
Determine Plaque Load in Alzheimer’s
Disease Patients
To investigate whether assessment of murine IgG
plaque labeling indices could be used as a tool to predict plaque load of AD patients, we incubated sera
from AD patients (n ⫽ 8) with sections derived from a
murine AD model, calculated plaque labeling indices,
grouped patients into cohorts with low and high
plaque labeling indices (⬎0.75 ⫽ high; ⬍0.75 ⫽ low),
and compared ␤-amyloid plaque loads in parietal cortices (see Supplemental Fig 2). Plaque loads (diffuse
plaques) were lower in the cohort with high plaque labeling index when compared with the cohort with low
plaque labeling index as determined by murine plaque
immunoreactivity assay (high group: average, 13.22/2.62
diffuse/neuritic plaques per tissue core; standard deviation, 4.77/1.99; n ⫽ 4; low group: average, 17.24/2.94
diffuse/neuritic plaques per tissue core; standard deviation, 5.29/1.43; n ⫽ 4; see Fig 3C) (nonsignificant tendency, Mann–Whitney U test not performed because of
small sample size).
The use of TMAs in research has led to significant improvements in the standardization of immunohisto-
chemical stainings, has enabled high-throughput analysis of genes of interest, and has thus accelerated the
transition of basic research into clinical applications.23
Here, we have used this technique to assess involvement of naturally occurring ␤-amyloid antibodies in
AD. Although a TMA of AD patients has been used in
a previous study to compare histological protocols between laboratories, this is the first study in which largescale AD-TMAs have been used to address biological
questions related to AD.24 To accommodate the variability of neuropathological alterations found in AD,
we have decided to include three anatomically defined
brain regions per AD/control patient. Sections of donor blocks were examined carefully to select areas for
tissue removal. By removing a total of eight representative tissue punches per region, variations caused by
irregular distribution of ␤-amyloid were minimized.
We are aware that heterogeneity of plaque distribution
may have introduced an error. Because the likelihood
of this error is identical in all sampled individuals, this
should not influence the outcome of studies comparing
in situ analysis. The issue of tissue heterogeneity in tumor samples has been investigated in great detail, and
the majority of studies show that TMAs fully reproduce findings on large sections.19 The fact that our
AD-TMAs distinguish between AD and control patients by a quantitative assessment of diffuse and neuritic plaques or neuritic plaques alone ( p ⬍ 0.001) in
all three brain regions reinforces the validity and representativity of this approach.
AD-TMAs allow for the direct comparison of morphological parameters excluding confounding factors
such as section thickness, antibody binding efficiencies,
or differences in visualization reaction. Furthermore,
we were able to quantify morphological parameters by
counting positive-stained structures in tissue punches
of identical size. Because we used consecutive sections,
these values could be directly correlated.
In AD, involvement of the immune system is believed to initiate as a secondary event to ␤-amyloid
deposition.4,25,26 Using our AD-TMAs, we are able to
confirm prominent immune activation in AD patients
presenting as dramatic increase in the amount of
phagocytic and activated microglia. Recently, it could
be shown that ␤-amyloid–specific antibodies both reduce de novo aggregation of ␤-amyloid and increase
resolubilization of existing ␤-amyloid.4,27 Neuropathological studies have shown binding of IgG to
␤-amyloid in immunized patients.12,28 We have investigated this phenomenon in a large cohort of AD and
control patients, and our analysis indicates that neuritic
plaques found in AD and control patients are decorated by IgGs. The unique setup of AD-TMAs allowed
us to calculate IgG plaque labeling indices by calculating the ratio of plaque-bound IgG to neuritic plaques.
By comparing IgG plaque labeling indices of AD and
Kellner et al: Autoantibodies in AD
control patients, we show that the proportion of IgGlabeled neuritic plaques is significantly greater in entorhinal cortex of AD patients than in control patients.
The clinical relevance of these data became obvious
when we separated AD patients into two cohorts with
either high or low IgG plaque labeling indices. Patients
with high IgG plaque labeling indices have a drastically
reduced ␤-amyloid plaque burden. The biological relevance of this finding is further substantiated by the
negative correlation between total plaque load and IgG
deposits (see Supplemental Fig 1). In immunized patients, ␤-amyloid is mobilized by phagocytic microglia.9,12 In agreement with these data, we find more
phagocytic microglia in the patient cohort with a high
IgG plaque labeling index. Whether ␤-amyloid mobilized by active or passive immunization leads to the increased occurrence of microhemorrhages is unresolved to
date.10,14,29,30 In our analysis, we did not detect a correlation between the proportion of IgG-labeled neuritic
plaques and microhemorrhages. Because our studies are
on non-␤-amyloid–immunized patients, these data suggest involvement of ␤-amyloid–specific antibodies not
only in actively immunized AD patients. This is in
agreement with a number of studies that have shown
␤-amyloid–specific antibodies in sera of AD and control
patients.15,16,31 Although the relevance of ␤-amyloid–
specific antibodies found in non-␤-amyloid–immunized
patients remains to be determined, our data underscore
the physiological relevance of this phenomena.16 Possibly, naturally occurring ␤-amyloid–specific antibodies
help to maintain ␤-amyloid homeostasis. Although the
affinity of these antibodies appears to be high, their impact on ␤-amyloid clearance is limited compared with
antibodies generated in the framework of ␤-amyloid in
immunization.9,15 We have utilized our findings in the
framework of potential diagnostic tests. To this respect,
we have scored immunoreactivity to ␤-amyloid after incubation of sera from AD patients with histological sections from murine AD models to predict cerebral
␤-amyloid loads. Indeed, patients whose sera showed
high ␤-amyloid immunoreactivity on mouse sections
had reduced ␤-amyloid loads in parietal cortices, a finding that is in agreement with similar studies on ␤-amyloid–immunized patients.7 Further studies are needed to
identify the contribution of naturally occurring ␤-amyloid–specific antibodies to the pathophysiology of AD. It
will be interesting to see whether naturally occurring
␤-amyloid–specific antibodies are able to clear
␤-amyloid in murine models of AD.
This study was supported by the Thyssen Foundation (
and the Hans und Ilse Breuer Foundation to S. Lange.
We thank S. Lange, U. Rumpf, M. Haberkorn, and
M. Storz for support. Transgenic APPsw mice were
kindly provided by M. Staufenbiel.
Annals of Neurology
Vol 65
No 1
January 2009
1. Ferri CP, Prince M, Brayne C, et al. Global prevalence of
dementia: a Delphi consensus study. Lancet 2005;366:
2. Gandy S. The role of cerebral amyloid beta accumulation in
common forms of Alzheimer disease. J Clin Invest 2005;115:
3. Goedert M, Spillantini MG. A century of Alzheimer’s disease.
Science 2006;314:777–781.
4. Weiner HL, Frenkel D. Immunology and immunotherapy of
Alzheimer’s disease. Nat Rev Immunol 2006;6:404 – 416.
5. Simard AR, Soulet D, Gowing G, et al. Bone marrow-derived
microglia play a critical role in restricting senile plaque formation in Alzheimer’s disease. Neuron 2006;49:489 –502.
6. Schenk D, Barbour R, Dunn W, et al. Immunization with
amyloid-beta attenuates Alzheimer-disease-like pathology in the
PDAPP mouse. Nature 1999;400:173–177.
7. Hock C, Konietzko U, Streffer JR, et al. Antibodies against
beta-amyloid slow cognitive decline in Alzheimer’s disease.
Neuron 2003;38:547–554.
8. Gilman S, Koller M, Black RS, et al. Clinical effects of Abeta
immunization (AN1792) in patients with AD in an interrupted
trial. Neurology 2005;64:1553–1562.
9. Nicoll JA, Barton E, Boche D, et al. Abeta species removal after
abeta42 immunization. J Neuropathol Exp Neurol 2006;65:
1040 –1048.
10. Ferrer I, Boada Rovira M, Sanchez Guerra ML, et al. Neuropathology and pathogenesis of encephalitis following amyloidbeta immunization in Alzheimer’s disease. Brain Pathol 2004;
11. Wilcock DM, Rojiani A, Rosenthal A, et al. Passive amyloid
immunotherapy clears amyloid and transiently activates microglia in a transgenic mouse model of amyloid deposition. J Neurosci 2004;24:6144 – 6151.
12. Nicoll JA, Wilkinson D, Holmes C, et al. Neuropathology of
human Alzheimer disease after immunization with amyloid-beta
peptide: a case report. Nat Med 2003;9:448 – 452.
13. Orgogozo JM, Gilman S, Dartigues JF, et al. Subacute meningoencephalitis in a subset of patients with AD after Abeta42
immunization. Neurology 2003;61:46 –54.
14. Pfeifer M, Boncristiano S, Bondolfi L, et al. Cerebral hemorrhage after passive anti-Abeta immunotherapy. Science 2002;
15. Geylis V, Kourilov V, Meiner Z, et al. Human monoclonal
antibodies against amyloid-beta from healthy adults. Neurobiol
Aging 2005;26:597– 606.
16. Du Y, Dodel R, Hampel H, et al. Reduced levels of amyloid
beta-peptide antibody in Alzheimer disease. Neurology 2001;
57:801– 805.
17. Mirra SS, Heyman A, McKeel D, et al. The Consortium to
Establish a Registry for Alzheimer’s Disease (CERAD). Part II.
Standardization of the neuropathologic assessment of Alzheimer’s disease. Neurology 1991;41:479 – 486.
18. Braak H, Braak E. Neuropathological stageing of Alzheimerrelated changes. Acta Neuropathol (Berl) 1991;82:239 –259.
19. Bubendorf L, Nocito A, Moch H, Sauter G. Tissue microarray
(TMA) technology: miniaturized pathology archives for highthroughput in situ studies. J Pathol 2001;195:72–79.
20. Sturchler-Pierrat C, Abramowski D, Duke M, et al. Two amyloid precursor protein transgenic mouse models with Alzheimer
disease-like pathology. Proc Natl Acad Sci U S A 1997;94:
21. Glatzel M, Rogivue C, Ghani A, et al. Incidence of CreutzfeldtJakob disease in Switzerland. Lancet 2002;360:139 –141.
22. Sastre M, Klockgether T, Heneka MT. Contribution of inflammatory processes to Alzheimer’s disease: molecular mechanisms.
Int J Dev Neurosci 2006;24:167–176.
23. Holst F, Stahl PR, Ruiz C, et al. Estrogen receptor alpha
(ESR1) gene amplification is frequent in breast cancer. Nat
Genet 2007;39:655– 660.
24. Alafuzoff I, Pikkarainen M, Al-Sarraj S, et al. Interlaboratory
comparison of assessments of Alzheimer disease-related lesions:
a study of the BrainNet Europe Consortium. J Neuropathol
Exp Neurol 2006;65:740 –757.
25. Takata K, Kitamura Y, Yanagisawa D, et al. Microglial transplantation increases amyloid-beta clearance in Alzheimer model
rats. FEBS Lett 2007;581:475– 478.
26. Yamamoto M, Kiyota T, Horiba M, et al. Interferon-gamma
and tumor necrosis factor-alpha regulate amyloid-beta plaque
deposition and beta-secretase expression in Swedish mutant
APP transgenic mice. Am J Pathol 2007;170:680 – 692.
27. Solomon B, Koppel R, Frankel D, Hanan-Aharon E. Disaggregation of Alzheimer beta-amyloid by site-directed mAb. Proc
Natl Acad Sci U S A 1997;94:4109 – 4112.
28. Lee M, Bard F, Johnson-Wood K, et al. Abeta42 immunization
in Alzheimer’s disease generates Abeta N-terminal antibodies.
Ann Neurol 2005;58:430 – 435.
29. Gandy S, Walker L. Toward modeling hemorrhagic and encephalitic complications of Alzheimer amyloid-beta vaccination
in nonhuman primates. Curr Opin Immunol 2004;16:
607– 615.
30. Asuni AA, Boutajangout A, Scholtzova H, et al. Vaccination of
Alzheimer’s model mice with Abeta derivative in alum adjuvant
reduces Abeta burden without microhemorrhages. Eur J Neurosci 2006;24:2530 –2542.
31. Lemke MR, Glatzel M, Henneberg AE. Antimicroglia antibodies in sera of Alzheimer’s disease patients. Biol Psychiatry 1999;
45:508 –511.
Kellner et al: Autoantibodies in AD
Без категории
Размер файла
452 Кб
burden, plaques, autoantibodies, common, amyloid, disease, alzheimers, help, control
Пожаловаться на содержимое документа