Aβ peptides can enter the brain through a defective blood-brain barrier and bind selectively to neurons
We have investigated the possibility that soluble, blood-borne amyloid beta (Abeta) peptides can cross a defective blood-brain barrier (BBB) and interact with neurons in the brain. Immunohistochemical analyses revealed extravasated plasma components, including Abeta42 in 19 of 21 AD brains, but in only 3 of 13 age-matched control brains, suggesting that a defective BBB is common in AD. To more directly test whether blood-borne Abeta peptides can cross a defective BBB, we tracked the fate of fluorescein isothiocyanate (FITC)-labeled Abeta42 and Abeta40 introduced via tail vein injection into mice with a BBB rendered permeable by treatment with pertussis toxin. Both Abeta40 and Abeta42 readily crossed the permeabilized BBB and bound selectively to certain neuronal subtypes, but not glial cells. By 48 h post-injection, Abeta42-positive neurons were widespread in the brain. In the cerebral cortex, small fluorescent, Abeta42-positive granules were found in the perinuclear cytoplasm of pyramidal neurons, suggesting that these cells can internalize exogenous Abeta42. An intact BBB (saline-injected controls) blocked entry of blood-borne Abeta peptides into the brain. The neuronal subtype selectivity of Abeta42 and Abeta40 was most evident in mouse brains subjected to direct intracranial stereotaxic injection into the hippocampal region, thereby bypassing the BBB. Abeta40 was found to preferentially bind to a distinct subset of neurons positioned at the inner face of the dentate gyrus, whereas Abeta42 bound selectively to the population of large neurons in the hilus region of the dentate gyrus. Our results suggest that the blood may serve as a major, chronic source of soluble, exogenous Abeta peptides that can bind selectively to certain subtypes of neurons and accumulate within these cells.
Aβ peptides can enter the brain through a defective
blood–brain barrier and bind selectively to neurons
Peter M. Clifford
, Shabnam Zarrabi
, Gilbert Siu
, Kristin J. Kinsler
, Mary C. Kosciuk
, Michael R. D'Andrea
, Steven Dinsmore
, Robert G. Nagele
New Jersey Institute for Successful Aging, University of Medicine and Dentistry of New Jersey/SOM, 2 Medical Center Drive,
Stratford, NJ 08084, USA
Department of Cell Biology, University of Medicine and Dentistry of New Jersey, Stratford, NJ 08084, USA
Johnson and Johnson Pharmaceutical Research and Development, Spring House, PA 19477, USA
ARTICLE INFO ABSTRACT
Accepted 11 January 2007
Available online 27 January 2007
We have investigated the possibility that soluble, blood-borne amyloid beta (Aβ) peptides
can cross a defective blood–brain barrier (BBB) and interact with neurons in the brain.
Immunohistochemical analyses revealed extravasated plasma components, including
Aβ42 in 19 of 21 AD brains, but in only 3 of 13 age-matched control brains, suggesting that
a defective BBB is common in AD. To more directly test whether blood-borne Aβ peptides
can cross a defective BBB, we tracked the fate of fluorescein isothiocyanate (FITC)-labeled
Aβ42 and Aβ40 introduced via tail vein injection into mice with a BBB rendered
permeable by treatment with pertussis toxin. Both Aβ 40 and Aβ42 readily crossed the
permeabilized BBB and bound selectively to certain neuronal subtypes, but not glial cells.
By 48 h post-injection, Aβ42-positive neurons were widespread in the brain. In the
cerebral cortex, small fluorescent, Aβ42-positive granules were found in the perinuclear
cytoplasm of pyramidal neurons, suggesting that these cells can internalize exogenous
Aβ42. An intact BBB (saline-injected controls) blocked entry of blood-borne Aβ peptides
into the brain. The neuronal subtype selectivity of Aβ42 and Aβ40 was most evident in
mouse brains subjected to direct intracranial stereotaxic injection into the hippocampal
region, thereby bypassing the BBB. Aβ40 was found to preferentially bind to a distinct
subset of neurons positioned at the inner face of the dentate gyrus, whereas Aβ42 bound
selectively to the population of large neurons in the hilus region of the dentate gyrus. Our
results suggest that the blood may serve as a major, chronic source of soluble, exogenous
Aβ peptides that can bind selectively to certain subtypes of neurons and accumulate
within these cells.
© 2007 Elsevier B.V. All rights reserved.
Alzheimer's disease (AD) is a neurodegenerative disease of the
elderly that results in progressive and dramatic memory loss,
cognitive decline, changes in behavior, reactive gliosis,
infl ammation and extensive destruction of neu rons and
their synapses in the cerebral cortex, entorhinal area, hippo-
campus, ventral striatum and basal forebrain (Braak and
BRAIN RESEARCH 1142 (2007) 223– 236
⁎ Corresponding author. Fax: +1 419 791 3345.
E-mail address: email@example.com (R.G. Nagele).
0006-8993/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
available at www.sciencedirect.com
Braak, 1991; Dickson, 1997; Felician and Sandson, 1999;
Gomez-Isla et al., 1997; Hardy and Allsop, 1991; Kasa et al.,
1997; Mirra et al., 1993; Scott et al., 1991; Selkoe, 2002;
Wisniewski et al., 1997; Wisniewski and Wen, 1985). The
alarming rise in the incidence of this disease is coincident
with recent increases in the average lifespan, with a nearly
40% incidence in individuals over the age of 85 years (Hebert et
al., 2003 ). The deposition of amyloid beta (Aβ) peptides,
predominantly the 42 amino acid form (Aβ42), within neurons
and amyloid plaques in brain tissue appears early in the
course of the disease and is a well-known hallmark of AD
(D'Andrea et al., 2001; Masters et al., 1985; Nagele et al., 2002;
Selkoe, 2002). Aβ42 is generated by the sequential cleavage of
the amyloid precursor protein (APP) by beta- and gamma-
secretases, respectively, and is capable of self-assembling into
nondegradable fibrils that persist within the brain tissue (Koo
and Squazzo, 1994; Masters et al., 1985; Wilson et al., 1999).
An understanding of the factors and conditions that lead to
the deposition of Aβ42 in the brain is crucial to the develop-
ment of effective treatment methods. Some proposed ther-
apeutic strategies for AD are currently aimed at reducing or
eliminating the deposition of Aβ42 in the brain. The achieve-
ment of this will most likely require a reduction in the
generation of Aβ42 from APP and/or some means of lowering
existing Aβ42 levels from sources that directly contribute to
the deposition of this peptide in the brain (De Felice and
Ferreira, 2002). Surprisingly, the source(s) of the Aβ42 that
accumulates within neurons and plaques in AD brains has not
yet been clearly identified. One possibility is that this peptide
is generated endogenously by the same neurons that later
accumulate large quantities of Aβ42 within their perikarya
(Hardy and Allsop, 1991). If so, then these neurons should
express APP as well as the beta- and gamma-secretase
enzymes that cleave it into Aβ peptide fragments. Indeed,
several studies have indicated that all of these proteins are
expressed in neurons, suggesting that these cells may produce
Aβ peptides throughout life and carefully regulate the balance
between their production and clearance (Selkoe, 1996). Why
these Aβ peptides accumulate to a pathology-inducing level
within neurons and amyloid plaques in the elderly is a
question of central importance. A partial list of potential,
aging-associated causative factors in the development of
sporadic AD includes a shift in the balance between Aβ
peptide production and its clearance from neurons that favors
intracellular accumulation, increased secretion of Aβ peptides
by neurons into the surrounding extracellular space,
increased levels of oxidative damage to these cells and global
brain hypoperfusion and the associated compensatory meta-
bolic shifts in affected neurons (Cohen et al., 1988; Higgins
et al., 1990; Kalaria, 2000; Nalivaevaa et al., 2004; Teller et al.,
1996; Wen et al., 2004).
Alternatively, the Aβ42 that deposits within neurons and
plaques could also originate from outside of the neurons
(exogenous Aβ42) during AD pathogenesis. In normal healthy
brains, the soluble Aβ peptide levels within the interstitial
space of the brain tissue are usually extremely low (Andreasen
and Blennow, 2002; Seubert et al., 1992). This suggests that, in
the normal state, any Aβ peptides generated internally by
neurons either remain within these cells and are eventually
degraded or, if released, are rapidly cleared from the extra-
cellular matrix and possibly returned to the venous blood with
the CSF through the arachnoid villi. On the other hand, levels of
soluble Aβ peptides in the blood are known to be much higher
than in the interstitial space and CSF in the brains of healthy
individuals (Seubert et al., 1992), raising the possibility that the
blood could be a potential source of exogenous Aβ peptides
that eventually deposit in the AD brain (
Zlokovic et al., 1993).
However, except for trace amounts of Aβ that are actively
transported across endothelial cells, it is well-known that
access of blood-borne Aβ peptides to brain tissue in normal
healthy individuals is effectively blocked by the integrity of the
blood–brain barrier (BBB) (Kandimalla et al., 2005; Poduslo et al.,
1999). The BBB is a complex structure composed of cerebral
endothelial cells resting on a basal lamina that is further
supported by the foot processes of local astrocytes (Gloor et al.,
2001; Risau et al., 1998). It closely regulates the passage of blood
components into the brain tissue and is highly impermeable to
nearly all proteins and other macromolecules while, at the
same time, allowing the selective entry of essential molecules
(Mayhan, 2001). Much evidence has revealed that aging is
associated with degenerative changes to blood vessels that
may compromise the integrity of the BBB. For example, a
number of relatively common neurodegenerative diseases in
the elderly, inc luding stroke, vascular dementia and AD
originate, at least in part, from cerebrovascular pathologies
that develop within the microvasculature of the brain (Breteler,
2000; Buee et al., 1997; de la Torre, 1997; Esiri et al., 1999; Kalaria
et al., 1996). Although stroke is most often attributed to defects
within larger vessels, smaller vessels of the brain microvascu-
lature are also involved in these pathologies and are the
leading cause of lacunar stroke, vascular dementia and
intracerebral hemorrhage (Esiri et al., 1997; Greenberg, 2006).
Additional testimony for a link between the neurovasculature
and neurodegenerative disease is the well-known fact that
Alzheimer pathology, including amyloid plaques and neurofi-
brillary tangles, develops subsequently within the vicinity of
stroke lesions (Jellinger, 2002; Kalaria, 1996, 2002; Natte et al.,
1998). In view of the expected high incidence of cerebrovas-
cular pathology and BBB breakdown in AD patients, it is critical
to determine if the defective cere bral mic rovascu latu re
represents an important source of the Aβ peptides that
contribute to amyloid deposition in AD brains.
In the present study, we have investigated the possibility
that disruption of the integrity of the BBB can lead to a chronic
influx of plasma components, including soluble Aβ peptides,
into the brain. We have also examined the fate of key soluble
exogenous Aβ peptides, Aβ42 and Aβ40, within the brain
parenchyma. To accomplish this, we tracked the fate of
fluorescein isothiocyanate (FITC)-labeled Aβ peptides intro-
duced via tail vein injection into mice in which the BBB had
been compromised by prior exposure to pertussis toxin.
Results show that blood-borne FITC-labeled Aβ peptides can
indeed traverse a defective BBB, enter into the brain parench-
yma, bind selectively to the surfaces of certain neurons and, in
the case of Aβ42, accumulate selectively within the same
subtype of neurons that are known to exhibit prominent Aβ42-
immunopositive deposits in AD brains. In view of the high
incidence of BBB compromise in nearly all patients with AD, as
supported here by direct detection of extravasated plasma
components in AD brains, we propose that the blood repre-
224 BRAIN RESEARCH 1142 (2007) 223– 236
sents a major and chronic source of exogenous, soluble Aβ42
that eventually deposits within neurons in AD brains. These
findings provide a possible link between BBB compromise and
the development of AD pathology and highlight the potential
effectiveness of therapies aimed at maintaining an d/or
reinforcing the integrity of the BBB, lowering plasma levels
of Aβ peptides and blocking the interactions of soluble
exogenous Aβ peptides with neurons.
2.1. Aβ42 readily leaks through the defective blood–brain
barrier (BBB) of arterioles into the parenchyma of AD brains,
but only rarely in age-matched control brains
We have applied immunohi stochemistry to post-mortem
human brains to determine if Aβ42 leaks from blood vessels
through a defective BBB into the parenchyma of AD brains. If
so, it might be possible to detect significant amounts of Aβ42
as well as other extravasated plasma components in these
brains (Figs. 1A–D). Results showed that, in addition to the
well-documented localization of Aβ42 within neurons, amy-
loid plaques and the walls of blood vessels, Aβ42 was also
often localized to diffuse “perivascular leak clouds” occupying
the region immediately surrounding arterioles (Fig. 1A). The
gradual diminution in the intensity of Aβ42-positive immu-
nostaining with increasing distance from arterioles is con-
sistent with the notion that these blood vessels are the main
source of the Aβ42 within the leak cloud. In most AD brains,
leak clouds were not associated with capillaries or venules and
thus appeared to be selective for arterioles (Figs. 1A–D). An
exception to this selectivity was occasionally observed in
regions of AD brains exhibiting particularly severe pathology,
as noted by widespread neuronal loss, numerous amyloid
plaques and signs of local inflammation. In this case, smaller
and more numerous leak clouds were also found in associa-
tion with capillaries and venules.
To assess the prevalence of BBB breakdown in AD brains,
we used immunohistochemistry to detect the presence of
extravasated plasma components (e.g., Aβ42, IgG and comple-
ment C1q) in the hippocampal/entorhinal cortex region of 21
AD brains and 13 brains from non-demented age-matched
controls. Significant extravasation of plasma components (as
judged by the pr esence of perivas cular leak clouds and
positive immunostaining of the extracellular space) was
Fig. 1 – Leak of Aβ42 and other plasma components through a defective blood–brain barrier in the cerebral cortex of AD
brains. (A) Section of AD brain double-immunostained with antibodies to Aβ42 (red) and collagen IV (blue) showing an
Aβ42-positive leak cloud (red dotted line) associated with an arteriole, but not local capillaries. Scale bar equals 100 μm. (B)
Section of AD brain immunostained with antibodies to complement C1q (brown) showing an arteriole-associated C1q-
positive leak cloud (red dotted line). Scale bar equals 30 μm. (C) Low magnification view of AD brain section immunostained
to reveal the widespread presence of immunoglobulin G (Ig) within the brain parenchyma which appears to originate
from local arterioles. Red dotted line indicates leak cloud. Scale bar equals 60 μm. (D) Section through age-matched control
brain showing much less parenchymal Ig and a relatively rare arteriole exhibiting an Ig-positive perivascular leak cloud
(red dotted line). Scale bar equals 60 μ m.
225BRAIN RESEARCH 1142 (2007) 223– 236
detected in 19 of 21 AD brains, but in only 3 of 13 age-matched
control brains (Figs. 1A–D). In AD brains, arterioles with leak
clouds were numerous, providing a plausible explanation for
the much greater overall abundance of detectable plasma
components in the parenchyma of AD brains than in age-
matched control brains (compare Figs. 1C and D). Despite the
detection of periarteriolar leak clouds in three of the control
brains, affected blood vessels were quite rare in sections taken
through these brains (Fig. 1D) and the relative amount of
Aβ42-, IgG- and C1q-immunopositive material detected by
immunohistochemistry in the intercellular space was much
lower in controls than in AD brains.
2.2. Blood-borne Aβ peptides can enter into the brain of
adult mice after breakdown of the BBB induced by pertussis
Next, we sought to more directly test the idea that the major
soluble Aβ peptides in the blood, Aβ40 and Aβ42, can cross a
defective BBB and enter into the brain tissue where they can
interact with neurons. To accomplish this, we disrupted the
integrity of the BBB in adult mice using pertussis toxin and
tracked the fate of fluorescent (FITC-labeled) Aβ peptides that
were introduced directly into the venous blood via tail vein
2.2.1. Confirmation that pertussis toxin (PT) induces
blood–brain barrier breakdown in the adult mouse brain
PT was introduced into mice by tail vein injection, and its
effects on the integrity of the BBB were monitored for up to 2
weeks post-injection using Evans blue dye (Figs. 2A–D). This
dye forms a complex with serum albumin and generates an
intense red fluorescence when illuminated with ultraviolet
light and examined using darkfield optics. In animals with an
intact BBB (i.e., those treated with an equivalent volume of
saline in place of pertussis toxin), Evans blue-specific fluores-
cence was restricted to the lumen of all blood vessels that pass
through the brain tissue, reflecting the integrity of the BBB in
these vessels (Fig. 2B). However, in all animals treated with PT
for a minimum of 4 days prior to sacrifice, Evans blue-specific
fluorescence was clearly seen within vessels, but many of
these vessels also exhibited prominent, Evans blue-positive
perivascular leak clouds (Fig. 2A). As in the post-mortem
human brains subjected to immunohistochemistry described
above, leak clouds were associated primarily with small
arterioles, rather than with capillaries or venules, suggesting
that elevated hydrostatic pressure within these vessels may
facilitate the efflux of blood-borne materials. Comparison of
brains derived from PT-treated mice revealed that, although
BBB compromise was detected in all PT-treated brains, we
observed individual variations in the location and extent of
Fig. 2 – Disruption of the blood–brain barrier in pertussis toxin-treated mice and confirmation with Evans blue dye. (A) Section
of PT-treated mouse cerebral cortex showing a gradient of Evans blue-positive material (red) leaking from an arteriole and
apparently binding to neuronal surfaces (free-standing arrows). Scale bar equals 30 μm. (B) Section of the brain of a mouse
injected with saline in place of PT showing the lack of Evans blue-positive material (red) in the brain parenchyma. Scale bar
equals 30 μm. (C +D) Sections through the hippocampal regions of PT-treated mice showing regional variations in the intensity
of Evans blue-positive granule neurons that presumably reflect similar variations in the relative amount of extravasated
blood-borne Evans blue–albumin complex (red). Scale bar equals 60 μm.
226 BRAIN RESEARCH 1142 (2007) 223– 236
BBB breakdown as judged by relative differences in the local
amount of extravasated Evans blue-positive material and the
number of blood vessels with associated leak clouds (Fig. 2C).
The hippocampus and medulla consistently were particularly
prone to exhibiting vascular leaks (Figs. 2C–D). To further
confirm the action of PT on inducing permeability of the BBB,
tail vein injections of heat-deactivated PT failed to render the
blood–brain barrier permeable to Evans blue dye or FITC-
labeled Aβ peptides (data not shown).
2.2.2. Blood-borne Aβ42 and Aβ40 cross the BBB of PT-treated
mice, enter into the brain tissue and bind to neurons
To track the fate of blood-borne A β peptides, we introduced
FITC-labeled Aβ42 and Aβ40 via tail vein injection into adult
mice that were first treated with PT at 4 and 2 days prior to Aβ
peptide injection. As early as 1 h post-injection, fluorescent
peptides were detectable within blood vessels traversing the
brain tissue as well as in the intercellular spaces of the brain
parenchyma (Figs. 3A–F). This demonstrates that both FITC-
labeled peptides can readily penetrate through the walls of
cerebral blood vessels made permeable by treatment with PT.
Consistent with the observations described above in post-
mortem human brain, blood vessels showing the most
pronounced Aβ peptide-positive leaks were arterioles, rather
than capillaries, venules or larger vessels (Fig. 3A). At 1 h post-
injection, the leak of FITC-labeled Aβ42 and Aβ40 in most
brain regions was usually confined to a small perivascular
region. Neurons positioned within the boundaries of these
small leak clouds often exhibited intense fluorescence,
suggesting that both Aβ40 and Aβ42 have affinity for neuronal
surfaces (Figs. 3A–I). Neurons positioned within the leak cloud,
but on the outer edge, were less fluorescent than those
situated closer to the blood vessel lumen (Fig. 3A). In addition,
neurons located outside of leak clouds, even those of the same
neuronal subtype, lacked Aβ-specific fluorescence. In some
brain regions at 1 h post-injection, especially the subcortex,
the amount of Aβ peptide in the brain parenchyma was more
substantial and focal clusters containing many Aβ peptide-
positive neurons were observed (Figs. 3B–E). At 48 h post-
injection, cells exhibiting Aβ peptide-specific fluorescence
were comparable to those observed at 1 h post-injection
(Fig. 3G), but were more widespread in the brain. Brains exa-
mined at 2 weeks and 1 month post-injection generally lacked
detectable fluorescent Aβ peptides (data not shown). Whether
this is due to degradation of the FITC-labeled peptide and/or
its clearance from these cells is not known. To further test the
specificity of the Aβ peptides for neuron, we used FITC-
conjugated reverse sequence Aβ peptides and administered
them via tail vein injection into mice with a pertussis toxin-
permeabilized blood–brain barrier. Results showed that these
blood-borne A β peptides entered into the brain tissue
(perivascular leak clouds were detected) but failed to bind to
neurons and glial cells (data not shown).
2.3. Blood-derived Aβ42 is internalized and can
accumulate in cortical pyramidal neurons
At 48 h post-injection, some of the larger cortical pyramidal
neurons exhibited tiny but intensely fluorescent, Aβ42-
specific spots in their perinuclear cytoplasm (Figs. 3H, I). The
extremely small but uniform size of these spots suggests that
they may be endocytic granules and that these cells may have
internalized surface-bound FITC-labeled Aβ42 via endocyto-
sis. In addition, the size and intracellular location of granules
in the basal cytoplasm of these cells closely resembles that of
intracellular clusters of Aβ 42-posit ive granules seen in
pyramidal neurons in AD brains. Comparable cytoplasmic
fluorescent spots were not observed in neurons in the brains
of animals injected with Aβ40 (data not shown).
2.4. A disrupted BBB is required for entry of blood-borne
Aβ peptides into the brain
To test the requirement for disruption of the BBB to allow the
influx of blood-borne Aβ40 and Aβ42 peptides into the brain
tissue, we subjected mice to saline injections in place of PT
prior to administration of labeled Aβ peptides by tail vein
injection. Results showed that FITC-labeled Aβ42 and Aβ40
were detected in the brain but were confined to the lumen of
blood vessels up to 48 h post-injection (Figs. 3J–L). In addition,
neurons throughout the brain were completely devoid of Aβ
peptide-specific fluorescence (Figs. 3J–L). These results suggest
that neither peptide can sufficiently permeate the intact BBB to
the point where it is detectable within the brain parenchyma.
2.5. Confirmation of the binding of blood-derived
FITC-labeled Aβ peptides to neurons using
To confirm the observed distribution of FITC-related fluores-
cence accurately reflecting the distribution of Aβ peptides, we
subjected brain sections of Aβ peptide-injected animals to
brightfield immunohistochemistry using antibodies specific
for Aβ42 (Figs. 4A–F). The observed staining patterns were
found to closely match those obtained using fluorescence
microscopy, including the apparent specificity for certain
neuronal subtypes in the cerebral cortex, subcorte x and
cerebellum (Figs. 4A–F). The selectivity of Aβ peptides for
neurons was readily apparent in these preparations as
astrocytes were unlabeled (Figs. 4C–F). In addition, regions of
brain tissue lacking Aβ-positive neurons were often juxta-
posed with areas containing labeled neurons, suggesting that
the unlabeled neurons were resident in regions where the BBB
remained intact (Fig. 4E). At 48 h post-injection with Aβ
peptides, some cortical pyramidal neurons showed increased
immunostaining intensity for Aβ42 in their basal cytoplasm
(Fig. 4F), a location which is coincident with the small,
intensely fluorescent, FITC-Aβ42-positive granules seen in
these cells in darkfield images (cf. Figs. 3H–I).
2.6. Bypass of the BBB via direct stereotaxic, intracranial
injection yields similar patterns of Aβ peptide influx and
It is well-known that Aβ peptides can associate with a number
of different plasma components during their transit in the
blood (Strittmatter et al., 1993). The effects of these associa-
tions and of crossing the BBB on the behavior of the peptides in
the brain, if any, are unknown. In an effort to circumvent these
potential limitations, we used direct stereotaxic injection to
227BRAIN RESEARCH 1142 (2007) 223– 236
Fig. 3 – Blood-borne FITC-labeled Aβ42 and Aβ40 cross the blood–brain barrier of PT-treated mice, enter into the brain
tissue and bind selectively to neurons (but not glial cells) in the cerebral cortex, subcortex, hippocampus and cerebellum.
In all images, free-standing arrows designate neurons. (A–C) Within 1 h post-injection, FITC-labeled Aβ40 leaks from local
vessels (red dotted line) and binds to the surfaces of neurons. Neurons positioned outside of the leak zone show little or
no labeling (yellow arrow). GC, granule cells. Scale bar equals 60 μm in A +C, 100 μminB.(D–F) Neurons with bound
FITC-labeled Aβ42 are abundant in the indicated brain regions, and comparison of FITC and DAPI image pairs of the same
section reveals a preferential binding of Aβ42 on the large neurons occupying the region enclosed by the dentate gyrus.
” added for orientation. Scale bar equals 100 μm in D and 150 μm in E and F. (G) At 48 h post-injection, labeled
neurons are still abundant. Scale bar equals 150 μm. (H–I) Some of the larger pyramidal neurons show small bright, FITC-
Aβ42-positive granules in the basal portion of the perinuclear cytoplasm. Scale bar equals 30 μm μm in H and 20 μminI.
(J–K) Basal level of autofluorescence in the cerebral cortex demonstrated in a mouse treated with only pertussis toxin.
Scale bar equals 60 μm. (L) Mouse given saline in lieu of pertussis toxin subsequently treated with Aβ42-FITC
demonstrates confinement of fluorescence in the context of an intact BBB. Scale bar equals 100 μm.
228 BRAIN RESEARCH 1142 (2007) 223– 236
bypass the BBB, thus providing an alternative means for highly
purified, soluble, monomeric Aβ peptide to gain access to
brain neurons. Results showed that fluorescent Aβ40 and
Aβ42 exhibited exceptional affinity for the neurons, but glial
cells remained unlabeled (Figs. 5A–D). Furthermore, neurons
in the vicinity of the injections site were intensely fluorescent,
whereas more remote neurons, incl uding those on the
contralateral side of the brain, showed little or no fluorescence
(data not shown). Interestingly, although the FITC-labeled
peptides were released into the brain parenchyma, only
neurons demonstrated intense labeling, with little or no
residual background labeling (Figs. 5A, C). This suggests that
Fig. 4 – Brightfield immunohistochemical detection of blood-borne FITC-labeled Aβ40 and Aβ42 peptides in the brains of
PT-treated mice. (A) In the cerebellum, Purkinje neurons, but not stellate, basket granule neurons, are immunopositive for Aβ40
at 1 h post-injection. Scale bar equals 150 μm. (B + C) Neurons in the subcortex are also Aβ40-immunopositive at 1 h
post-injection, but white matter and astrocytes are unlabeled. Scale bar equals 150 μm in B and 60 μm in C. (D) Section through
cerebral cortex of mouse at 48 h post-injection with Aβ42 showing variations in the relative intensity of immunolabeling
among pyramidal neurons and a lack of astrocytic labeling in both the cortical molecular layer (ML) and underlying pyramidal
cell layers (PCL). Scale bar equals 150 μm. (E) Section through cerebral cortex of mouse at 48 h post-injection with Aβ42 showing
both labeled and unlabeled neurons in positive and negative zones, respectively, that presumably reflect the presence and
absence of leaks in these zones. Scale bar equals 150 μm. (F) Higher magnification image of several Aβ42-immunopositive
cortical pyramidal neurons, one of which shows particularly intense labeling (isolated arrow) in the basal perinuclear
cytoplasm that may correspond to accumulated Aβ42. Scale bar equals 50 μm.
229BRAIN RESEARCH 1142 (2007) 223– 236
unbound Aβ peptide is rapidly cleared from the intercellular
space in the region of the injection site in these otherwise
2.7. Blood-derived Aβ42 and Aβ40 bind selectively to
certain neuronal subtypes after entering the brain
Int erestingly, in mice injected intracranially, Aβ peptide
fluorescence appeared to be specific for certain neuronal
subtypes, with neurons nearly devoid of surface-bound Aβ
peptide interspersed among those with relatively intense
labeling (Figs. 5A, C). In addition, there was a marked tendency
for Aβ40 and Aβ42 to each bind selectively to different
neuronal subtypes. For example, Fi gs. 5A, B show th e
hippocampal region of a mouse that was injected intracra-
nially with FITC-labeled Aβ40. Here, the Aβ40 peptide is clearly
bound to a distinct subset of neurons positioned at the inner
face of the dentate gyrus. The Aβ40-binding neurons were also
morphologically distinguishable from non-binding cells, with
Fig. 5 – Aβ peptides administered into the hippocampal region by intracranial stereotaxic injection bind selectively to
certain neuronal subtypes. (A+B) Image pair showing neurons lining the inner aspect of the dentate gyrus are selectively
labeled with FITC-Aβ40. Scale bar equals 150 μm. (C+D) Image pair showing that neurons positioned in the region
enclosed by the dentate gyrus are selectively labeled with FITC-Aβ42, whereas granule neurons are unlabeled, including
those that were found to selectively bind Aβ40. Scale bar equals 150 μm. (E+F) Low and higher magnification images of
a section of the mouse hippocampal region immunostained with antibodies to the α7nAChR. Some immunopositive
neurons designated with arrows. Neurons in this region that bind FITC-Aβ42 also express elevated levels of α7nAChR.
Scale bar equals 60 μm in E and 40 μminF.
230 BRAIN RESEARCH 1142 (2007) 223– 236
the former cells being somewhat larger with more rounded
and lightly DAPI-stained nuclei (Figs. 5A–B). By contrast,
intracranially injected, FITC-labeled Aβ42 showed preferential
binding to the very large neurons found in the hilus of the
dentate gyrus (the region partially enclosed by the granule
cells forming the dentate gyrus) (Figs. 5C–D). This same pool of
neurons selectively accumu lated Aβ42-FITC when it was
applied in the peripheral venous blood following pertussis
toxin treatment (cf. Fig. 3F). It is interesting to note that
comparable neurons are prone to accumulate massive
amounts of intracellular Aβ42-positive material in AD brains.
Interestingly, these neurons also show relatively high levels of
expression of the α7 nicotinic acetylcholine receptor
(α7nAChR) (Figs. 5E, F), as do pyramidal neurons in the
cerebral cortex and Purkinje cells in the cerebellum that are
also prone to accumulate intracellular Aβ42 (data not shown)
(Nagele et al., 2002; Wang et al., 2000).
3.1. Evidence that the blood is a major source of Aβ
peptides in the brains of AD patients
Results of the present study strengthen the possibility that the
blood serves as a major, chronic source of soluble Aβ peptides
that are gradually deposited within neurons and amyloid
plaques in AD brains. A number of studies have suggested that
the blood can contribute Aβ peptides to the brain (Deane et al.,
2004; Mackic et al., 2002; Zlokovic, 2004, 2005). Aβ40 and Aβ42
are well-known to be present in the blood, and levels in the
plasma are generally ten-fold greater than in the interstitial
fluid of the brain and in the CSF of healthy individuals (Citron
et al., 1994; Ghiso et al., 1997). In addition, plasma Aβ40 and
Aβ42 levels generally increase with age and are elevated in
some patients during the early stages of AD, but usually
decline thereafter (Mayeux et al., 2003). For a chronic influx of
blood-borne Aβ peptides to contribute significantly to the
observed amyloid deposition, soluble Aβ peptides must be
able to first cross the defective BBB before entering into the
brain tissue. In support of this possibility, our immunohisto-
chemical analyses of post-mortem human brain tissues show
that extravasation of plasma components, including Aβ42, IgG
and complement C1q, from the microvasculature is common
in AD brains, but relatively rare in age-matched controls.
The most consistent risk factor associated with AD is aging.
One reason for this may be that aging is well-known to be
accompanied by deleterious structural and functional changes
in the walls of blood vessels that may compromise their
function. Consequently, it has been proposed that at least
some forms of AD may develop primarily from cerebrovas-
cular changes (de la Torre, 2004a ). In fact, a number of
relatively common aging-associated neurodegenerative dis-
eases, including stroke, vascular dementia and AD originate,
at least in part, from pathological changes that evolve within
the microvasculature of the brain ( Bailey et al., 2004; Breteler,
2000; Buee et al., 1997; de la Torre, 1997, 2004b; Esiri et al., 1997;
Farkas and Luiten, 2001; Jellinger, 2002; Zlokovic, 2005). One
expected consequence of these changes is a compromise of
BBB integrity, which would then allow plasma components
that are normally excluded from the brain to enter into the
interstitial spaces of the brain and come into direct contact
with local neurons. In the present study, immunohistochem-
ical analyses of post-mortem human brains show that
extravasation of plasma components, as revealed by the
presence of peri vascular leak clouds containing plasma
proteins, was both common in AD brains and primarily
associated with arterioles. Two features of arterioles may
make them particularly vulnerable to BBB compromise in AD
brains; the higher hydrostatic pressure within these vessels as
compared to capillaries and venules, and the presence of
smooth muscle cells that are often burdened with intracellular
accumulated amyloid, one of the diagnostic features of AD.
These changes may adversely alter the functioning of vascular
smooth muscle cells or could have a toxic effect upon the cells,
both of which could jeopardize the integrity of the BBB. Most
methods for experimentally assessing BBB integrity are based
on measuring the exclusion of defined molecules that do not
normally gain access to the brain. Evidence that BBB break-
down is common in the brains of living AD patients comes
from in vivo studies that have monitored the ratio of CSF
albumin to serum albumin and demonstrated a significant
perturbation in this ratio correlating to the severity of the
dementia (Skoog et al., 1998; Wada, 1998). BBB abnormalities
have been reported to occur early in the course of the disease
in humans, prior to the onset of clinical dementia, and BBB
permeability has also been shown to precede amyloid plaque
formation in Tg2576 mice (Skoog et al., 1998; Ujiie et al., 2003).
Although, taken together, these findings suggest that breach
of the BBB is temporally associated with AD, the question of
whether it triggers the onset of AD and/or is involved in its
progression remains to be resolved. However, in either case,
the potential merits of therapeutic strategies aimed at low-
ering plasma levels of Aβ peptides and preventing their transit
across the BBB are obvious.
3.2. Blood-borne Aβ42 and Aβ40 can cross a defective BBB
and bind selectively to neurons in the mouse brain
Although our immunohistochemical studies on post-mortem
human brain described above provide strong evidence that
plasma Aβ peptides can penetrate a defective BBB in AD
brains, it is essential to be able to demonstrate this more
directly and determine if these peptides will behave similarly
in the living brain. To test this, we have tracked the fate of
fluorescein isothiocyanate (FITC)-labeled Aβ 42 and Aβ40
peptides that were introduced by tail vein injection into
mice in which the BBB was disrupted by prior treatment with
pertussis toxin. This toxin, derived from Bordetella pertussis,is
known to effectively increase the permeability of the BBB in
mice as early as 24 h after injection and for a period of up to
several weeks (Amiel, 1976; Bruckener et al., 2003). This agent
is suspected of being involved in the development of
neurological sequelae associated with whooping cough that
are linked to disruption of BBB integrity in the epithelium of
the choroid plexus and/or cerebral capillary endothelium. It
has also been used to enhance the development of experi-
mental autoimmune encephalomyelitis (EAE), a widely used
mouse model for multiple sclerosis (Ben-Nun et al., 1997;
Linthicum et al., 1982; Yong et al., 1993).
231BRAIN RESEARCH 1142 (2007) 223– 236
Our results using the pertussis toxin-treated mouse model
show that blood-borne, FITC-labeled Aβ peptides can readily
penetrate the defective BBB, enter into the brain parenchyma
and bind selectively to the surfaces of certain neurons in the
mouse brain. For example, as early as 1 h post-injection, we
found that fluorescent Aβ40 and Aβ42 were both detectable in
blood vessels traversing the brain tissue and as well as in
numerous small perivascular leak clouds that emanated from
the microvasculature. Neurons positioned within or near
these leak clouds were fluorescent, whereas others located
outside the boundaries of leak clouds lacked Aβ peptide-
specific fluorescence, even neurons of the same subtype.
These observations confirm that the local blood vessels are
the source of the neuron-bound, fluorescent Aβ peptide. In
addition, the f act that neurons within the leak cloud
consistently showed greater fluorescence intensity than the
leak cloud itself suggests that the labeled peptides can
accumulate on neuronal surfaces. By 24 and 48 h post-
injection, neurons exhibiting Aβ peptide-specific fluorescence
were much more widespread in the brain and were found in
the cerebral cortex, hippocampal region, cerebellum and
brainstem. On the other hand, glial cells throughout the
brain were consistently unlabeled, suggesting that soluble,
exogenous Aβ peptides in the interstitial space of the brain are
able to recognize and bind to some factor(s) on neuronal
surfaces which is apparently lacking on the surfaces of glial
cells. The identity of this factor is unknown. We also found
that, once in the interstitial space of the brain, the fate of Aβ
peptides was similar, regardless of whether they gained entry
into the brain tissue by crossing the defective BBB or bypassing
the BBB by direct stereotaxic intracranial injection. Disruption
of the BBB in mice was required for entry of fluorescent Aβ
peptides into the brain since mice injected with saline instead
of pertussis toxin failed to show Aβ peptide-specific fluores-
cence outside of blood vessels in the brain parenchyma. This
finding is consistent with other studies that have used bolus
injections of soluble Aβ in place of brain perfusion, where very
little uptake of Aβ peptide was detected, thus confirming that,
in normal healthy animals, the integrity of the BBB protects
the brain from the entry of soluble, blood-borne Aβ peptides
(Maness et al., 1994; Poduslo et al., 1997; Shayo et al., 1997).
3.3. Blood-derived Aβ42 and Aβ40 bind selectively and
differentially to certa in neuronal subtypes
The present study also shows that, after gaining access to the
brain tissue, Aβ42 and Aβ40 not only bind selectively to
neurons, and not glial cells, but also that each peptide showed
preferential binding to different neuronal subtypes. For
example, in the hippocampal regions, Aβ40 targeted a distinct,
morphologically distinguishable subset of neurons positioned
at the inner face of the granule cell layer of the dentate gyrus.
By contrast, fluorescent Aβ42 preferentially labeled the much
larger neurons that populate the adjacent region that is
circumscribed by the granule cells of the dentate gyrus.
Factors governing the observed cell type specificity of these
Aβ peptides remain to be determined. However, it is interest-
ing to note that the Aβ42-positive neurons in the mouse
hippocampal region are comparable to hippocampal neurons
that are known to accumulate massive amounts of intracel-
lular Aβ42-positive material in human AD brains (data not
shown). Here, we have used immunohistochemistry to show
that the large, Aβ42-positive neurons in the hippocampal
formation of the mouse brain also express relatively high
levels of the alpha7 nicotinic acetylcholine receptor
(α7nAChR). The α7nAChR is a neuronal homopentameric
cation channel that is highly permeable to Ca
distributed among neurons in the brain, expressed in basal
forebrain cholinergic neurons that project to the hippocampus
and cortex and involved in cognition and memory (Breese et
al., 1997; Chang and Berg, 1999; Hellstrom-Lindahl et al., 1999;
Nordberg, 1994; Paterson and Nordberg, 2000; Seguela et al.,
1993; Ullian et al., 1997). Decreased levels of nAChRs, including
the α7nAChR, have been reported in specific regions of AD
brains early in the course of the disease and correlate well
with evolving cognitive dysfunctions (Banerjee et al., 2000;
Burghaus et al., 2000; Guan et al., 2000; Lee et al., 2000;
Nordberg, 1994; Wevers et al., 2000; Wevers and Schroder,
1999; Whitehouse and Kalaria, 1995). Liu et al. (2001) has
shown that Aβ42 can block the response of α7nAChRs on
hippocampal neurons, suggesting that binding of Aβ42 to
neuronal cell surfaces may alter the normal pattern of
neuronal signaling. Our previous studies have suggested that
α7nAChR may play a role in the intraneuronal deposition of
Aβ42 in AD brains ( D'Andrea and Nagele, 2006; Nagele et al.,
2002, 2003). Here, we propose that the high affinity interaction
of Aβ 42 and α7nAChR may act to concentrate sol uble,
exogenous Aβ42 in the brain parenchyma via interactions on
the surfaces of neurons that are abundantly endowed with
this receptor (Liu et al., 2001; Pettit et al., 2001).
3.4. Evidence that blood-derived Aβ42, but not Aβ40, can
accumulate within neurons
Lastly, we have observed that some of the largest cortical
pyramidal neurons in the brains of mice at 48 h post-injection
displayed tiny fluorescent, Aβ42-positive granules in their
cytoplasm that resemble endocytic granules. The location of
these granules in the basal cytoplasm of the highly polarized
pyramidal cells is similar to that seen in Aβ42-burdened
pyramidal neurons in AD brains. Interestingly, we failed to
detect comparable intracellular granules in mice injected with
Aβ40 (data not shown). Although it is possible that the
observed Aβ42-positive granules may represent an early
stage of intracellular Aβ42 accumulation, the number of
granules per cell and the number of cells possessing granules
were both too few to allow us to come to this conclusion with a
reasonable level of confidence. However, brightfield immuno-
histochemistry also showed evidence of intracellular Aβ42
accumulation in larger pyramidal neurons (cf. Fig. 4F). To
further test this possibility, we also examined the brains of
mice at longer post-injection times (e.g., 2 and 4 weeks), with
the idea that these accumulations may increase over time.
However, with the exception of what appeared to be an
increased number of small lipofuscin deposits, no significant
residual Aβ pepti de-specific neuronal fluorescence was
detected in these specimens. Several explanations for this
are possible. First, the neuron-bound Aβ42 peptide may now
be contained within the lipofuscin deposits which, unfortu-
nately, exhibit a level of autofluorescence that precludes our
232 BRAIN RESEARCH 1142 (2007) 223– 236
ability to discern any Aβ peptide-specific fluorescence.
Second, since these otherwise healthy animals received only
a single injection of FITC-labeled Aβ peptide, it is possible that
the cells were able to eventually clear both surface-bound and
accumulated Aβ peptide over these time intervals. Lastly,
degradation of the labeled peptides and/or FITC moiety may
have occurred. Additional long-term, time-course studies with
either constant infusions or repeated bolus injections of Aβ
peptides in BBB-compromised animals will be necessary to
determine if the chronic influx of soluble, blood-derived Aβ
peptides can eventually lead to the generation of more
substantial i ntrane uronal amyloid deposits and po ssibly
amyloid plaques in the brains of these mice.
3.5. Clinical and therapeutic perspectives
The results of the present study demonstrate that plasma-
derived Aβ peptides can penetrate a defective BBB, bind
selectively to the surfaces of certain types of neurons and
eventually accumulate within these cells. These findings lead
us to propose that, in AD brains, not only does the blood serve
as a chronic source of Aβ peptides that gain access to the brain
tissue through a defective BBB, but also that the influx of
soluble, exogenous Aβ42 into the brain may be the origin of
the intracellular Aβ42 that accumulates within neurons. If so,
then the potential benefits of therapeutic strategies aimed at
reducing levels of exogenous, peripheral Aβ peptides, espe-
cially Aβ42, in the plasma and maintaining or fortifying the
integrity of the BBB are evident (D'Andrea and Nagele, 2006).
From the clinical perspective, if Aβ peptides chronically enter
into the brain through a defective BBB, then the exact brain
location of the BBB compromise, the magnitude of the breach,
whether it is focal or global, and the rate of its progression will
continue to change as the patient ages and the disease
evolves. Thus, it is conceivable that some combination of
these variables in each patient contributes to the observed
heterogeneity in the nature and severity of the dementia from
one individual to the next, including the age of disease onset,
the specific site(s) of its developing pathology, the rate of
disease progression and the changing pattern of symptoms
that arise during the course of this disease.
4. Experimental procedures
4.1. Aβ peptides
Fluorescein isothiocyanate (FITC)-labele d Aβ40 and Aβ42
peptides were purchased from Anaspec (San Jose, CA).
Unlabeled Aβ peptides were obtained from Biosource (Camar-
illo, CA). Peptides were solubilized to the monomeric form
according to the method described by Zagorski et al. (1999).
Briefly, Aβ peptide was solubilized (1 mg/ml) in trifluoroacetic
acid (TFA), followed by removal of TFA under a slow, steady
stream of N
. Thi s was followed by three sequ ential
solubilizations in 1,1,1,3,3,3-hexafluoroisopropanol (HFIP) fol-
lowed by remov al of HFIP under N
each time. Stock
solutions (50.0 μM) were prepared in 0.5× PBS, 250.0 mM
HEPES buffer, pH 8.5. Working solutions were diluted in 0.9%
NaCl to 6. 9 μM and use d for tail vein injecti on. Protein
concentrations were confirmed using the Pierce Micro BCA
Assay (Rockford, IL).
Immunohistochemical analyses were performed using the
following Aβ peptide-specific antibodies: two anti-Aβ42 anti-
bodies (polyclonal, AB5078P Chemicon International, Teme-
cula, CA; polyclonal 556501 Pharmingen, San Diego, CA), anti-
Aβ40 antibody (polyclonal, AB5074P Chemicon International);
anti-α7 nicotinic acetylcholine receptor (α7nAChR, monoclo-
nal M-220 Sigma, St. Louis, MO). The specificity of each of
these antibodies was confirmed by Western blotting or ELISA.
For the anti-α7nAChR antibodies, positive controls included a
check for specific immunolabeling in known positive cells and
tissues in human multi-tissue checkerboards (Novagen, La
Jolla, CA) and Chemicon International.
All procedures involving animals were performed according to
the guide lines of t he institution al animal care and use
committee (IACUC) in accordance with NIH guidelines under
protocol 0762. Normal Swiss Webster mice (3–6 months) were
obtained from Taconic Farms (Hudson, NY), housed 3–5ina
cage, maintained on ad libitum food and water with a 12-hour
light/dark cycle in a fully equipped, AAALAC accredited
4.4. Tail vein injection
For administration of pertussis toxin (PT) (Sigma-Aldrich, St.
Louis, MO), animals (n =33) were injected with 100 μl of 0.9%
saline containing 3.0×10
μg/μl of PT at 4 and 2 days prior to
injection with Aβ peptides. Compromise of blood–brain barrier
was confirmed in a cohort of mice by Evans blue dye injection,
where some mice on the day slated for injection with Aβ
peptides were also given i.p. injections of 1 ml of 4% Evans blue
in PBS and then sacrificed at 3 h post-injection. The Evans blue
dye binds with high affinity to serum albumin and exhibits an
intense red fluorescence when viewed u nder ultraviolet
illumination. FITC-labeled peptides (100 μl of 6.9 μM in 0.9%
saline) were injected into the tail vein. Controls for Aβ
peptides and for pertussis toxin (n=10) were injected with
similar volume s of salin e with or wit hout Aβ peptides.
Animals were euthanized at 30 min, 1 h, 3 h, 6 h, 1 day,
2 days, 1 week, 2 weeks and 4 weeks post-injection using
pentobarbital at 200 mg/kg. Brains were removed rapidly and
fixed in 4% paraformaldehyde in PBS at 4 °C.
4.5. Intracranial stereotaxic injections
On the day of the surgery, adult Swiss Webster mice (30–34 g)
pentobarbital and placed in a Kopf stereotaxic device (Kopf
Instruments, Tujunga, CA). A midsagittal incision was made to
expose the cranium, and a hole was drilled to access the
hippocampus at the following coordinates relative to a
Bregma reference point: anterior–posterior, − 2.0 mm; med-
ial–lateral, 2.5 mm. A 29-gauge needle attached to a 5.0-μl
233BRAIN RESEARCH 1142 (2007) 223– 236
syringe was lowered 4.0 mm dorsoventral into the hippocam-
pal area, and a 5.0 μl injection was made gradually over a
40 min period. Experimental mice received a unilateral cortical
injection of 50.0 μM F ITC-labeled Aβ peptides and were
sacrificed at 30 min and 3 h after completion of the injection.
Five control mice were injected with comparable volumes of
0.9% saline and sacrificed in tandem with the experimental
groups. In all control mice, damage was limited to injection
trauma. Confirmation of successful injection was determined
by the detection of FITC fluorescence in the brain tissue.
4.6. Post-mortem human brain tissue
The hippocampus and entorhinal cortex from patients with
clinically diagnosed, sporadic AD (n =21, age range=72–84) and
control tissues from normal, age-matched, neurologically
normal individuals (n =13, age range=69–81) were obtained
from the Harvard Brain Tissue Resource Center (Belmont, MA)
and the Cooperative Human Tissue Network (Philadelphia,
PA). Post-mortem intervals for these brains were
24 h and
pathological confirmation of AD for each brain specimen was
carried out according to the criteria defined by The National
Institute on Aging and the Reagan Institute Working Group on
Diagnostic Criteria for the Neuropathological Assessment of
AD (1997). Brain specimens from age-matched controls were
screened using the same criteria. Tissues were characterized
immunohistochemically for the presence of neuritic (amyloid)
plaques and neurofibrillary tangles as described previously
(D'Andrea et al., 2001; Nagele et al., 2002). Control tissues
exhibited no gross pathology and minimal localized micro-
scopic AD-like neuropathology. Tissues were processed for
routine paraffin embedding and sectioning according to
established protocols. Five-micrometer sections were serially
cut, mounted onto SuperFrost Plus+ microslides ( Fisher
Scientific, Pittsburgh, PA) and dried overnight.
4.7. Mouse brain tissue
Mouse brains were quickly isolated, fixed with 4% parafor-
maldehyde in phosphate-buffered saline (PBS) and cut with a
tissue slicer into roughly 1 mm slabs. After continued fixation
overnight, each slab was infiltrated with 10% and then 30%
sucrose at 4 °C under constant gentle agitation. Specimens
were then snap frozen in liquid nitrogen and stored at − 80 °C
until used. Sections (10–12 μm thick) were cut using a Leica
cryostat, mounted onto Fisher SuperFrost Plus+ slides and air-
Immunohistochemistry was carried out on paraffin-embed-
ded tissues as described previously (D'Andrea et al., 2001;
Nagele et al., 2002). Briefly, after removal of paraffin with
xylene and rehydration through a graded series of decreasing
concentrations of ethanol, protein antigenicity was enhanced
by microwaving sections in Target Buffer (Dako, Carpenturia,
CA) for 2 min. Alternatively, frozen sections were treated for
2 min in acetone and again air-dried. Following a 30 min
incubation in 0.3% H
, sections were treated for 30 min in
normal blocking serum and then incubated with primary
antibodies at appropriate dilutions for 1 h at room tempera-
ture. Following a thorough rinse in PBS, a secondary biotin-
labeled antibody was applied for 30 min. Immunoreactions
were treated with the avidin–peroxidase-labeled biotin com-
plex (ABC, Vector Labs, Foster City, CA) and visualized by
treatment of sections with 3-3-diaminobenzidine-4 HCl (DAB)/
(Biomeda, Foster City, CA). Sections were lightly counter-
stained with hematoxylin, dehydrated through a graded series
of increasing concentrations of ethanols, cleared in xylene and
mounted in Permount. Controls consisted of brain sections
treated with either nonimmune serum, pre-absorbed antibody
or omission of the primary antibody. Specimens were
examined and photographed with a Nikon FXA microscope,
and digital images were recorded using a Nikon DXM1200F
digital camera and processed using Image Pro Plus (Phase 3
Imaging, Glen Mills, PA) imaging software.
The authors wish to acknowledge the generous support of the
National Institute on Aging (AG00925), Alzheimer's Associa-
tion, the New Jersey Governor's Council on Autism and the
New Jersey Gerontological Institute. Some tissue samples
were provided by the Cooperative Human Tissue Network
which is funded by the National Cancer Institute.
Amiel, S.A., 1976. The effects of Bordetella pertussis vaccine on
cerebral vascular permeability. Br. J. Exp. Pathol. 57, 653–662.
Andreasen, N., Blennow, K., 2002. Beta-amyloid (Abeta) protein in
cerebrospinal fluid as a biomarker for Alzheimer's disease.
Peptides 23, 1205–1214.
Bailey, T.L., et al., 2004. The nature and effects of cortical
microvascular pathology in aging and Alzheimer's disease.
Neurol. Res. 26, 573–578.
Banerjee, C., et al., 2000. Cellular expression of alpha7 nicotinic
acetylcholine receptor protein in the temporal cortex in
Alzheimer's and Parkinson's disease—A stereological
approach. Neurobiol. Dis. 7, 666–672.
Ben-Nun, A., et al., 1997. Immunomodulation of murine
experimental autoimmune encephalomyelitis by pertussis
toxin: the protective activity, but not the disease-enhancing
activity, can be attributed to the nontoxic B-oligomer. Proc.
Assoc. Am. Physicians 109, 120–125.
Braak, H., Braak, E., 1991. Demonstration of amyloid deposits and
neurofibrillary changes in whole brain sections. Brain Pathol. 1,
Breese, C.R., et al., 1997. Comparison of the regional expression of
nicotinic acetylcholine receptor alpha7 mRNA and
[125I]-alpha-bungarotoxin binding in human postmortem
brain. J. Comp. Neurol. 387, 385–398.
Breteler, M.M., 2000. Vascular risk factors for Alzheimer's disease:
an epidemiologic perspective. Neurobiol. Aging 21, 153–160.
Bruckener, K.E., et al., 2003. Permeabilization in a cerebral
endothelial barrier model by pertussis toxin involves the PKC
effector pathway and is abolished by elevated levels of cAMP.
J. Cell Sci. 116, 1837–1846.
Buee, L., et al., 1997. Brain microvascular changes in Alzheimer's
disease and other dementias. Ann. N. Y. Acad. Sci. 826, 7– 24.
234 BRAIN RESEARCH 1142 (2007) 223– 236
Burghaus, L., et al., 2000. Quantitative assessment of nicotinic
acetylcholine receptor proteins in the cerebral cortex of
Alzheimer patients. Brain Res. Mol. Brain Res. 76, 385–388.
Chang, K.T., Berg, D.K., 1999. Nicotinic acetylcholine receptors
containing alpha7 subunits are required for reliable synaptic
transmission in situ. J. Neurosci. 19, 3701–3710.
Citron, M., et al., 1994. Excessive production of amyloid
beta-protein by peripheral cells of symptoma tic and
presymptomatic patients carrying the Swedish familial
Alzheimer disease mutation. Proc. Natl. Acad. Sci. U. S. A. 91,
Cohen, M.L., et al., 1988. In situ hybridization of nucleus basalis
neurons shows increased beta-amyloid mRNA in Alzheimer
disease. Proc. Natl. Acad. Sci. U. S. A. 85, 1227–1231.
D'Andrea, M.R., Nagele, R.G., 2006. Targeting the alpha 7 nicotinic
acetylcholine receptor to reduce amyloid accumulation in
Alzheimer's disease pyramidal neurons. Curr. Pharm. Des. 12,
D'Andrea, M.R., et al., 2001. Evidence that neurones accumulating
amyloid can undergo lysis to form amyloid plaques in
Alzheimer's disease. Histopathology 38, 120–134.
Deane, R., et al., 2004. RAGE (yin) versus LRP (yang) balance
regulates Alzheimer amyloid beta-peptide clearance through
transport across the blood–brain barrier. Stroke 35, 2628–2631.
De Felice, F.G., Ferreira, S.T., 2002. Beta-amyloid production,
aggregation, and clearance as targets for therapy in
Alzheimer's disease. Cell. Mol. Neurobiol. 22, 545–563.
de la Torre, J.C., 1997. Hemodynamic consequences of deformed
microvessels in the brain in Alzheimer's disease. Ann. N. Y.
Acad. Sci. 826, 75–91.
de la Torre, J.C., 2004a. Alzheimer's disease is a vasocognopathy:
a new term to describe its nature. Neurol. Res. 26, 517–524.
de la Torre, J.C., 2004. Is Alzheimer's disease a neuro degenerative
or a vascular disorder? Data, dogma, and dialectics. Lancet
Neurol. 3, 184–190.
Dickson, D.W., 1997. Neuropathological diagnosis of Alzheimer's
disease: a perspective from longitudinal clinicopathological
studies. Neurobiol. Aging 18, S21–S26.
Esiri, M.M., et al., 1997. Neuropathological assessment of the
lesions of significance in vascular dementia. J. Neurol.,
Neurosurg. Psychiatry 63, 749–753.
Esiri, M.M., et al., 1999. Cerebrovascular disease and threshold for
dementia in the early stages of Alzheimer's disease. Lancet
Farkas, E., Luiten, P.G., 2001. Cerebral microvascular pathology in
aging and Alzheimer's disease. Prog. Neurobiol. 64, 575–611.
Felician, O., Sandson, T.A., 1999. The neurobiology and
pharmacotherapy of Alzheimer's disease. J. Neuropsychiatry
Clin. Neurosci. 11, 19–31.
Ghiso, J., et al., 1997. Alzheimer's soluble amyloid beta is a normal
component of human urine. FEBS Lett. 408, 105–108.
Gloor, S.M., et al., 2001. Molecular and cellular permeability control
at the blood–brain barrier. Brain Res. Brain Res. Rev. 36,
Gomez-Isla, T., et al., 1997. Neuronal loss correlates with but
exceeds neurofibrillary tangles in Alzheimer's disease. Ann.
Neurol. 41, 17–24.
Greenberg, S.M., 2006. Small vessels, big problems. N. Engl. J. Med.
Guan, Z.Z., et al., 2000. Decreased protein levels of nicotinic
receptor subunits in the hippocampus and temporal cortex of
patients with Alzheimer's disease. J. Neurochem. 74, 237–243.
Hardy, J., Allsop, D., 1991. Amyloid deposition as the central event
in the aetiology of Alzheimer's disease. Trends Pharmacol. Sci.
Hebert, L.E ., et al., 2003. Alzheimer disease in the US population:
prevalence estimates using the 2000 census. Arch. Neurol. 60,
Hellstrom-Lindahl, E., et al., 1999. Regional distribution of nicotinic
receptor subunit mRNAs in human brain: comparison between
Alzheimer and normal brain. Brain Res. Mol. Brain Res. 66,
Higgins, G.A., et al., 1990. Altered levels of amyloid protein
precursor transcripts in the basal forebrain of behaviorally
impaired aged rats. Proc. Natl. Acad. Sci. U. S. A. 87, 3032–3036.
Jellinger, K.A., 2002. Alzheimer disease and cerebrovascular
pathology: an update. J. Neural Transm. 109, 813–836.
Kalaria, R.N., 1996. Cerebral vessels in ageing and Alzheimer's
disease. Pharmacol. Ther. 72, 193–214.
Kalaria, R.N., 2000. The role of cerebral ischemia in Alzheimer's
disease. Neurobiol. Aging 21, 321 – 330.
Kalaria, R.N., 2002. Small vessel disease and Alzheimer's
dementia: pathological considerations. Cerebrovasc. Dis. 13
(Suppl. 2), 48–52.
Kalaria, R.N., et al., 1996. Production and increased detection of
amyloid beta protein and amyloidogenic fragments in brain
microvessels, meningeal vessels and choroid plexus in
Alzheimer's disease. Brain Res. Mol. Brain Res. 35, 58–68.
Kandimalla, K.K., et al., 2005. Pharmacokinetic analysis of the
blood–brain barrier transport of 125I-amyloid beta protein 40 in
wild-type and Alzheimer's disease transgenic mice (APP,PS1)
and its implications for amyloid plaque formation.
J. Pharmacol. Exp. Ther. 313, 1370–1378.
Kasa, P., et al., 1997. The cholinergic system in Alzheimer's
disease. Prog. Neurobiol. 52, 511–535.
Koo, E.H., Squazzo, S.L., 1994. Evidence that production and release
of amyloid beta-protein involves the endocytic pathway. J. Biol.
Chem. 269, 17386–17389.
Lee, D.H.-S., et al., 2000. Decreased alpha-7 nicotinic acetylcholine
receptor protein levels in sporadic Alzheimer's disease
hippocampus. Alzheimer's Rep. 3, 215–218.
Linthicum, D.S., et al., 1982. Acute experimental autoimmune
encephalomyelitis in mice. I. Adjuvant action of Bordetella
pertussis is due to vasoactive amine sensitization and increased
vascular permeability of the central nervous system. Cell.
Immunol. 73, 299–310.
Liu, Q., et al., 2001. beta-Amyloid peptide blocks the response of
alpha 7-containing nicotinic receptors on hippocampal
neurons. Proc. Natl. Acad. Sci. U. S. A. 98, 4734–4739.
Mackic, J.B., et al., 2002. Circulating amyloid-beta peptide crosses
the blood–brain barrier in aged monkeys and contributes to
Alzheimer's disease lesions. Vascul. Pharmacol. 38, 303–313.
Maness, L.M., et al., 1994. Passage of human amyloid beta-protein
1–40 across the murine blood–brain barrier. Life Sci. 55,
Masters, C.L., et al., 1985. Neuronal origin of a cerebral amyloid:
neurofibrillary tangles of Alzheimer's disease contain the same
protein as the amyloid of plaque cores and blood vessels.
EMBO J. 4, 2757– 2763.
Mayeux, R., et al., 2003. Plasma A[beta]40 and A[beta] 42 and
Alzheimer's disease: relation to age, mortality, and risk.
Neurology 61, 1185–1190.
Mayhan, W.G., 2001. Regulation of blood–
permeability. Microcirculation 8, 89–104.
Mirra, S.S., et al., 1993. Making the diagnosis of Alzheimer's
disease. A primer for practicing pathologists. Arch. Pathol. Lab.
Med. 117, 132–144.
Nagele, R.G., et al., 2002. Intracellular accumulation of
beta-amyloid(1–42) in neurons is facilitated by the alpha 7
nicotinic acetylcholine receptor in Alzheimer's disease.
Neuroscience 110, 199–211.
Nagele, R.G., et al., 2003. Astrocytes accumulate A beta 42 and give
rise to astrocytic amyloid plaques in Alzheimer disease brains.
Brain Res. 971, 197–209.
Nalivaevaa, N.N., et al., 2004. Effect of hypoxia/ischemia and
hypoxic preconditioning/reperfusion on expression of some
amyloid-degrading enzymes. Ann. N. Y. Acad. Sci. 1035, 21–33.
Natte, R., et al., 1998. Microvasculopathy is associated with the
235BRAIN RESEARCH 1142 (2007) 223– 236
number of cerebrovascular lesions in hereditary cerebral
hemorrhage with amyloidosis, Dutch type. Stroke 29, 1588–1594.
Nordberg, A., 1994. Human nicotinic receptors—Their role in aging
and dementia. Neurochem. Int. 25, 93–97.
Paterson, D., Nordberg, A., 2000. Neurona l nicotinic receptors in
the human brain. Prog. Neurobiol. 61, 75–111.
Pettit, D.L., et al., 2001. beta-Amyloid(1–42) peptide directly
modulates nicotinic receptors in the rat hippocampal slice.
J. Neurosci. 21, RC120.
Poduslo, J.F., et al., 1997. Permeability and residual plasma volume
of human, Dutch variant, and rat amyloid beta-protein 1–40 at
the blood–brain barrier. Neurobiol. Dis. 4, 27–34.
Poduslo, J.F., et al., 1999. Receptor-mediated transport of human
amyloid beta-protein 1–40 and 1–42 at the blood–brain barrier.
Neurobiol. Dis. 6, 190–199.
Risau, W., et al., 1998. Differentiation of blood–brain barrier
endothelial cells. Pathol. Biol. (Paris) 46, 171–175.
Scott, S.A., et al., 1991. Volumetric atrophy of the amygdala in
Alzheimer's disease: quantitative serial reconstruction.
Neurology 41, 351–356.
Seguela, P., et al., 1993. Molecular cloning, functional properties,
and distribution of rat brain alpha 7: a nicotinic cation channel
highly permeab le to calcium. J. Neurosci. 13, 596–604.
Selkoe, D.J., 1996. Amyloid beta-protein and the genetics of
Alzheimer's disease. J. Biol. Chem. 271, 18295–18298.
Selkoe, D.J., 2002. Alzheimer's disease is a synaptic failure. Science
Seubert, P., et al., 1992. Isolation and quantification of soluble
Alzheimer's beta-peptide from biological fluids. Nature 359,
Shayo, M., et al., 1997. The putative blood–brain barrier transporter
for the beta-amyloid binding protein apolipoprotein j is
saturated at physiologicalconcentrations. Life Sci. 60, L115–L118.
Skoog, I., et al., 1998. A population study on blood–brain barrier
function in 85-year-olds: relation to Alzheimer's disease and
vascular dementia. Neurology 50, 966–971.
Strittmatter, W.J., et al., 1993. Binding of human apolip oprotein E
to synthetic amyloid beta peptide: isoform-specific effects and
implications for late-onset Alzheimer disease. Proc. Natl. Acad.
Sci. U. S. A. 90, 8098–8102.
Teller, J.K., et al., 1996. Presence of soluble amyloid beta-peptide
precedes amyloid plaque formation in Down's syndrome. Nat.
Med. 2, 93–95.
The National Institute on Aging and Reagan Institute Working
Group on Diagnostic Criteria for the Neuropathological
Assessment of Alzheimer's Disease, 1997. Consen sus
recommendations for the postmortem diagnosis of
Alzheimer's disease. Neurobiol. Aging 18, S1–S2.
Ujiie, M., et al., 2003. Blood–brain barrier permeability precedes
senile plaque formation in an Alzheimer disease model.
Microcirculation 10, 463–470.
Ullian, E.M., et al., 1997. Rapid synaptic transmission in the avian
ciliary ganglion is mediated by two distinct classes of nicotinic
receptors. J. Neurosci. 17, 7210 – 7219.
Wada, H., 1998. Blood–brain barrier permeability of the demented
elderly as studied by cerebrospinal fluid–serum albumin ratio.
Intern. Med. 37, 509–513.
Wang, H.Y., et al., 2000. beta-Amyloid(1–42) binds to alpha7
nicotinic acetylcholine receptor with high affinity.
Implications for Alzheimer's disease pathology. J. Biol. Chem.
Wen, Y., et al., 2004. Increased beta-secretase activity and
expression in rats following transient cerebral ischemia. Brain
Res. 1009, 1–8.
Wevers, A., Schroder, H., 1999. Nicotinic acetylcholine receptors in
Alzheimer's disease. J. Alzheimer's Dis. 1, 207–219.
Wevers, A., et al., 2000. Expression of nicotinic acetylcholine
receptors in Alzheimer's disease: postmortem investigations
and experimental approaches. Behav. Brain Res. 113,
Whitehouse, P.J., Kalaria, R.N., 1995. Nicotinic receptors and
neurodegenerative dementing diseases: basic research and
clinical implications. Alzheimer Dis. Assoc. Disord. 9 (Suppl. 2),
Wilson, C.A., et al., 1999. Intracellular APP processing and A beta
production in Alzheimer disease. J. Neuropathol. Exp. Neurol.
Wisniewski, H.M., Wen, G.Y., 1985. Substructures of paired helical
filaments from Alzheimer's disease neurofibrillary tangles.
Acta Neuropathol. (Berl.) 66, 173–176.
Wisniewski, H.M. , et al., 1997. Amyloid angiopathy and
blood–brain barrier changes in Alzheimer's disease. Ann. N. Y.
Acad. Sci. 826, 161–172.
Yong, T., et al., 1993. Enhancement of histamine-induced vascular
leakage by pertussis toxin in SJL/J mice but not BALB/c mice.
J. Neuroimmunol. 45, 47–52.
Zagorski, M.G., et al., 1999. Methodological and chemical factors
affecting amyloid beta peptide amyloidogenicity. Methods
Enzymol. 309, 189–204.
Zlokovic, B.V., 2004. Clearing amyloid through the blood–brain
barrier. J. Neurochem. 89, 807–811.
Zlokovic, B.V., 2005. Neurovascular mechanisms of Alzheimer's
neurodegeneration. Trends Neurosci. 28, 202–208.
Zlokovic, B.V., et al., 1993. Blood–brain barrier transport of
circulating Alzheimer's amyloid beta. Biochem. Biophys. Res.
Commun. 197, 1034–1040.
236 BRAIN RESEARCH 1142 (2007) 223– 236