Brain Research 888 2001 107–116 www.elsevier.com locate bres
Research report
Hyperglycemic but not normoglycemic global ischemia induces marked early intraneuronal expression of b-amyloid precursor protein
Baowan Lin, Myron D. Ginsberg , Raul Busto
Cerebral Vascular Disease Research Center , Department of Neurology, University of Miami School of Medicine, Miami, FL 33101, USA
Accepted 19 September 2000
Abstract
Preischemic hyperglycemia is known to accentuate acute ischemic injury to neurons, microglia, and endothelia. In the present study, we used a monoclonal antibody to the N-terminal portion of b-APP to examine how the immunoreactivity of this normal membrane
glycoprotein is differentially influenced by transient cerebral ischemia when carried out under normoglycemic vs. hyperglycemic conditions. Anesthetized, physiologically regulated rats received 12.5 min of global forebrain ischemia by bilateral carotid artery
occlusions plus systemic hypotension. Hyperglycemia was induced by intraperitoneal dextrose administration prior to ischemia. One or three days later, brains were examined by b-APP immunohistochemistry. Ischemia under hyperglycemic conditions led to the robust,
widespread intraneuronal expression of b-APP immunoreactivity in neocortex, hippocampus, thalamus, and striatum of all 11 rats; this was most prominent at 24 h postischemia. Compared to rats with normoglycemic ischemia, numbers of b-APP-immunopositive neurons
in the parietal cortex of hyperglycemic rats were increased by 5.9 fold at 24 h, and by 10.6 fold at 3 days postischemia. b-APP-immunopositive neurons in hyperglycemic rats often exhibited striking morphological alterations typical of ischemic necrosis;
however, no b-APP immunoreaction was observed in zones of frank infarction. Brains of normoglycemic rats n511, by contrast, showed only weak b-APP immunostaining in occasional non-necrotic pyramidal neurons of parietal neocortex; no necrosis was present in
thalamus. In sham-operated hyperglycemic rats, b-APP immunostaining of thalamic neurons was somewhat increased at 24 h. Western analysis revealed that the hyperglycemia-induced intraneuronal overexpression of b-APP was not associated with an overall increase in
tissue levels. The results of this study demonstrate that transient forebrain ischemia under hyperglycemic conditions leads to the early intraneuronal expression of b-APP within neuronal populations showing a heightened susceptibility to hyperglycemia-induced
accentuation of ischemic injury. Our data suggest that b-APP or its metabolites may be involved in the injury process.
2001 Elsevier
Science B.V. All rights reserved.
Keywords : Hyperglycemia; Amyloid; Selective vulnerability; Immunochemistry; Rat
1. Introduction determine its physiological or pathological role. Increased
b-APP levels have been noted in acute cerebral ischemia Acute hyperglycemia markedly accentuates the neuro-
[28,29,33,39,41], and we have shown that b-APP accumu- pathological alterations resulting from cerebral ischemia
lates, as well, in the subacute and chronic stages after a [7,10,16,24,31], but the mechanisms underlying this effect
brief episode of global forebrain ischemia in nor- are not fully understood. b-amyloid precursor protein b-
moglycemic rats [18]. However, the influence of hy- APP comprises a group of highly conserved 100–140 kDa
perglycemia on ischemia-induced changes of b-APP has integral membrane glycoproteins expressed in almost all
not been studied. In this immunohistochemical study, we mammalian cells and present in normal neuronal perikarya,
employed a monoclonal antibody to the N-terminal region proximal dendrites and axons [20,22,37] as well as in
of b-APP [18] to explore whether hyperglycemia influ- blood vessels, meningeal membranes, and ependyma [6].
ences the postischemic expression of this important pro- b-APP is metabolized via several alternative pathways that
tein.
Corresponding author. Tel.: 11-305-243-6449; fax: 11-305-243-
2. Materials and methods
5830. E-mail address
: mdginsbergstroke.med.miami.edu M.D. Ginsberg.
These studies were conducted in male Wistar rats
0006-8993 01 – see front matter
2001 Elsevier Science B.V. All rights reserved. P I I : S 0 0 0 6 - 8 9 9 3 0 0 0 3 0 2 3 - 7
108 B
weighing 300–350 g following an overnight fast. The n56 or 3 days n55 in each group after the ischemic
University of Miami’s Animal Care and Use Committee insult. Ten sham animals were similarly studied hy-
approved all procedures. Anesthesia was induced with 3 perglycemic, 24 h survival, n54; hyperglycemic, 3 day
halothane and 70 nitrous oxide. Animals were intubated survival, n53; normoglycemic, 3 day survival, n53. Rats
endotracheally and ventilated mechanically on mixtures of were reanesthetized by halothane and perfused via the
0.5 halothane, 70 nitrous oxide and a balance of ascending aorta with FAM a mixture of 40 formalde-
oxygen. The femoral arteries were catheterized for blood hyde, glacial acetic acid and methanol, 1:1:8 by volume
pressure monitoring and to permit arterial sampling for for 19 min at a pressure of 100–120 mmHg following a 1
blood–gas measurements. Arterial PCO and PO
were min perfusion with physiological saline. The heads were
2 2
maintained in the normal range by ventilatory adjustments. immersed in FAM at 48C for 1 day. The brain was then
Rats were immobilized by pancuronium bromide, 0.75 removed, placed in FAM for another day, and then
mg kg i.v. Rectal temperature was measured continually blocked. Coronal tissue blocks were embedded in paraffin
and maintained at 37.0–37.58C by a warming lamp above for sectioning. Brain sections 10 mm thick were prepared
the rat’s body. Cranial temperature was separately moni- at 250 mm intervals and were stained for the light
tored by a 29-gauge thermocouple implanted into the left microscopic immunocytochemical visualization of the N-
temporalis muscle and was maintained at 36.0–36.58C terminal portion of b-APP695 clone 22C11, 0.1 mg ml,
throughout the experiment by a small warming lamp above Boehringer Mannheim.
the rat’s head. In our previous studies, we have shown For b-APP immunohistochemistry, sections were rehy-
that this corresponds to a brain temperature of 36.5–37.08C drated and placed in 1.5 H O
in methanol to block
2 2
[5]. endogenous peroxidase activity. They were then rinsed in
running water, dipped in 0.05 M phosphate-buffered saline PBS, and incubated with normal horse serum. The
2.1. Hyperglycemia primary antibody 1:500 was applied at 48C for 18 h. To
test for nonspecific staining, negative controls were con- To induce hyperglycemia, rats received an intraperi-
ducted in which the primary antibody was omitted and, toneal injection of 2.5 ml of 25 dextrose solution 30 min
instead, mouse IgG1 1:500 was used No. X0931, Dako prior to initiating the ischemic insult. Normoglycemic rats
Corp., Carpenteria, CA. Sections then were rinsed with were given 2.5 ml of sterile water i.p.
PBS and the secondary antibody applied. The avidin– biotin complex Vector, Burlingame, CA was used for
2.2. Global forebrain ischemia antibody detection followed by peroxidase reaction with
diaminobenzidine DAB and H O . Slides were washed
2 2
Both common carotid arteries were exposed via a in 0.5 Triton X-100 with or without counterstaining by
midline ventral incision and were gently separated from hematoxylin. These procedures were similar to those
the surrounding nerve fibers. Ligatures consisting of recently reported from our laboratory [18].
polyethylene PE-10 tubing contained within a double- lumen Silastic tubing were passed loosely around each
artery. To produce ischemia, arterial blood was gradually 2.4. Identification and quantitation of b-APP
withdrawn into a heparanized syringe to reduce mean immunoreactivity
arterial blood pressure MAP to 45 mmHg. Global ischemia was then produced by tightening the carotid
Representative areas of neocortex, striatum coronal ligatures bilaterally and maintaining blood pressure at 45
level, 0.2 mm posterior to bregma, and the hippocampal mmHg for 12.5 min [17]. The ischemic insult was ended
CA1 sector and thalamus 3.6 mm posterior to bregma by removing the carotid ligatures and immediately reinfus-
were examined by an observer B.L. blinded to the ing the shed blood which had been maintained at body
experimental groups. Specific b-APP staining was iden- temperature to restore MAP to 100–120 mmHg. Catheters
tified by its dark brown appearance, which often exhibited were then removed, incisions closed, and rats were re-
a fine granular punctate pattern. All immunohistochemical turned to their cages and given free access to food and
and histological findings were carefully confirmed by a water.
second observer M.D.G.. Counting of b-APP-positive Sham-operated rats received similar operative prepara-
cells was performed in a blinded fashion within a stan- tion but were not subjected to carotid occlusion or blood
dardized 103 microscopic field of the lateral parietal withdrawal.
neocortex of each hemisphere region Par2 [27], coronal level of 3.6 mm posterior to bregma. Left- and right-
2.3. Tissue preparation and immunohistochemistry hemisphere counts were summed for each rat.
Inter-group differences in b-APP-positive cell counts Animals in the ischemia groups were sacrificed at either
were assessed by the Mann–Whitney rank sum test. The 24 h hyperglycemic group, n56; normoglycemic, group
level of P,0.05 was regarded as statistically significant.
B . Lin et al. Brain Research 888 2001 107 –116
109
2.5. Quantitative western blots 2.6 fold by dextrose administration hyperglycemic–is-
chemic rats, 340666 mg dL; normoglycemic–ischemic In a separate series, relative levels of expression of
group, 133621 mg dl. b-APP were determined by quantitative western blot
analysis. The following experimental groups were studied: 3.2. Summary of light-microscopic histopathology
sham normoglycemic n53 and sham hyperglycemic n53 with 24 h recovery; 12.5 min ischemia 124 h
The light-microscopic neuropathological alterations in recovery normoglycemic n54, 12.5 min ischemia 124 h
normoglycemic and hyperglycemic rats surviving for 1 and recovery hyperglycemic n54; 12.5 min ischemia 13
3 days following a 12.5 min ischemic insult have been day recovery normoglycemic n54; and 12.5 min is-
previously reported and illustrated in detail [16]. To chemia 13 day recovery hyperglycemic n54. Tissue
summarize these findings, brains of hyperglycemic –is- samples from hippocampus |100 mg, thalamus |40
chemic rats contained extensive ischemic neuronal altera- mg, and cortex |500 mg of each brain were isolated and
tions and foci of infarction within neocortex, striatum, and rapidly frozen on dry-ice. Total protein was isolated by
thalamus; and widespread hippocampal damage extending homogenization in buffer composed of 50 mM Tris, pH
beyond the CA1 sector. These changes were already 8.0, 100 mM NaCl, 0.1 SDS, 0.01 mg ml leupeptin, and
apparent at 24 h. Vascular changes observed in the hy- 100 mM phenylmethylsulfonyl fluoride PMSF. Protein
perglycemic group consisted of endothelial thickening, concentration was determined using the Bio-Rad DC
vascular occlusion, prominent peri- and intravascular poly- protein assay system.
morphonuclear and monocytic accumulation, and foci of Protein samples were added to 10 SDS and boiled to
perivascular rarefaction and microinfarction. By contrast, denature them. Twenty five mg of protein per sample was
normoglycemic–ischemic rats showed no injury at 24 h. loaded onto a 12 SDS polyacrylamide stacking gel and
At 3 days, the cerebral neocortex and striatum of nor- electrophoresed in 25 mM Tris, 250 mM glycine pH 8.3,
moglycemic–ischemic animals showed only mild damage, 0.1 SDS at 30 mA. Proteins were transferred by standard
with few or no necrotic neurons; the thalamus was spared electroblotting techniques to a PVDF transfer membrane
from injury; hippocampal damage was confined to the CA1 Polyscreen, NEN Research Products in 25 mM Tris, 192
sector; and no endothelial changes were present. Numbers mM glycine, pH 8.3, and 10 methanol. After confirma-
of ischemic cells in the striatum of hyperglycemic animals tion of transfer by Ponceau S Red staining, the membrane
were increased by 5 fold or more compared to nor- was blocked with 5 milk in 13PBS plus Tween-20 for 2
moglycemic rats, and surviving normal neurons in the h at room temperature.
vulnerable hippocampal CA1 sector of hyperglycemic– Incubation with a monoclonal antibody to the N-termi-
ischemic rats were reduced to one third of the number seen nal portion of b-APP695 clone 22C11, 0.1 mg ml,
in normoglycemic animals [16]. Boehringer Mannheim was performed in the same buffer
at 1:1000 dilution at room temperature for 2 h. Following a 3.3. Immunohistochemical observations
brief wash, incubation with HRP-linked anti-mouse IgG 1:1000 was carried out for 45 min. Detection of signal
3.3.1. Sham-operated group was performed with a Phototope-HRP Western blot de-
In four of six sham brains 3 normoglycemic, 1 hy- tection kit New England BioLabs utilizing LumiGlo
perglycemic studied at 3 days, scattered faintly stained, reagent and was visualized on X-ray film. The film was
apparently extracellular foci of b-APP immunoreactivity then digitized, and densitometric measurements were made
were observed on non-counterstained sections, randomly to quantify signals from each sample. Following stripping
distributed within the middle layers of neocortex. These of the antibody, the blot was re-incubated as above with an
foci, however, lacked both a dark brown appearance and a antibody against actin Sigma Chemicals at 1:5000. The
punctate or granular texture in hematoxylin-counterstained actin signal was used as an internal control for evenness of
sections and were only rarely intraneuronal. In three of loading.
four hyperglycemic sham-operated rats with 24 h survival, widespread thalamic neurons tended to exhibit light brown
cytoplasmic b-APP immunostaining that was not present
3. Results in the other sham-operated groups. This faint staining
