Introduction cognitive processes continues to unfold see Ref. [11] for
Brain Research 882 2000 75–85 www.elsevier.com locate bres
Research report
Estrogen replacement does not prevent the loss of choline acetyltransferase-positive cells in the basal forebrain following either
neurochemical or mechanical lesions
Payal Aggarwal, Robert B. Gibbs
Department of Pharmaceutical Sciences , University of Pittsburgh School of Pharmacy, 1004 Salk Hall, Pittsburgh, PA 15261, USA
Accepted 8 August 2000
Abstract
Recent studies have shown that estrogen replacement can enhance the functional status of basal forebrain cholinergic neurons. Studies have also shown that estrogen has neuroprotective effects both in vitro and in vivo on a variety of cells and against a variety of insults.
The present study examined the ability of estrogen replacement to protect basal forebrain cholinergic neurons from the effects of neurochemical and mechanical injury. Ovariectomized Sprague–Dawley rats received either estrogen replacement or sham surgery, and
then received either a unilateral injection of ibotenic acid into the nucleus basalis magnocellularis, or unilateral transection of the fimbria fornix. Cholinergic neurons in the medial septum and nucleus basalis were detected and quantified using immunohistochemical
techniques. The data show that neither 3 weeks nor 13 weeks of continuous estrogen replacement prevented the loss of choline acetyltransferase ChAT-containing cells in the nucleus basalis following a unilateral injection of ibotenic acid. Likewise, estrogen
replacement did not prevent a decrease in ChAT-positive cells detected in the medial septum following unilateral transection of the fimbria fornix. Notably, increased numbers of ChAT-positive cells were detected in the contralateral nucleus basalis, and in the ipsilateral
and contralateral medial septum, at 2 weeks following a unilateral injection of ibotenic acid into the nucleus basalis; however, these effects were not related to hormone treatment. These data suggest that estrogen replacement does not protect cholinergic neurons in the
medial septum and nucleus basalis from the effects of excitotoxic or mechanical injury.
2000 Elsevier Science B.V. All rights
reserved.
Theme : Disorders of the nervous system
Topic : Trauma
Keywords : Hormone replacement; Cholinergic neuron; Neuroprotection; Ibotenic acid
1. Introduction cognitive processes continues to unfold see Ref. [11] for
review. Several recent prospective studies suggest that es-
Recent studies have show that estrogen has significant trogen replacement can help to reduce the risk of Alz-
neuroprotective effects in vitro on a wide variety of heimer’s-related
dementia in
postmenopausal women
cultured cells exposed to a variety of toxic insults see Ref. [3,34,62]. The mechanisms for this effect are still largely
[39]. For example, estrogen has been shown to reduce cell unknown, although evidence for significant estrogen ef-
death associated with serum deprivation, growth factor fects on neuronal anatomy, biochemistry, and function
deprivation, b-amyloid toxicity, excitatory amino acid within regions of the brain that play an important role in
toxicity, and various inducers of oxidative stress, in cells ranging from primary neuronal cultures to neuroblastoma
and non-neuronal cell lines. In some cases estrogen-me- diated neuroprotection in culture is blocked with selective
estrogen receptor antagonists [5,55], suggesting that es- trogen supports survival via the activation of estrogen
Corresponding author. Fax: 11-412-624-1850. E-mail address
: gibbsr1pitt.edu R.B. Gibbs.
receptor regulated genes e.g. bcl-2 [10]. In other cases,
0006-8993 00 – see front matter
2000 Elsevier Science B.V. All rights reserved. P I I : S 0 0 0 6 - 8 9 9 3 0 0 0 2 8 3 2 - 8
76 P
neuroprotection in culture is related to the structure of the 2. Materials and methods
estrogen molecule and not to any specific estrogenicity of the molecule or to the activation of estrogen receptors
2.1. Animals [4,27–29]. Neuroprotective effects of estrogen in vivo
have also been described. Specifically, several studies have Young female Sprague–Dawley rats were purchased
reported the ability of estrogen replacement to reduce from Hilltop Laboratories and housed in pairs on a 12:12-h
infarct size and neuronal damage following experimentally light dark cycle with food and water available ad libitum.
induced transient forebrain ischemia [7,53,61,64]. In one All animals were housed for a minimum of 2 weeks before
study, the reduction in infarct size was associated with the use. All animals then underwent bilateral ovariectomy and
prevention of injury-induced down-regulation of bcl-2 and 2 weeks later, received either 3-mm silastic capsules
estrogen receptor beta gene expression [7]. Ischemic injury containing 17b-estradiol crystals implanted s.c. or blank
was also shown to dramatically upregulate estrogen re- capsules. Studies have shown that these estrogen capsules
ceptor alpha gene expression on the injured side. Estrogen produce mean circulating levels of estradiol in the range of
has also been shown to reduce kainic acid-induced cell loss 60–120 pg ml serum.
in the dentate gyrus [2]. These studies suggest that at least some neuroprotective effects of estrogen in vivo may
2.2. Neurochemical lesions involve the selective regulation of estrogen receptors and
estrogen receptor target genes. Seven to ten days following capsule implantation,
One of the consistent biochemical features of Alzheim- animals n541 were anesthetized with a combination of
er’s disease is the loss of cholinergic neurons in the medial ketamine 14 mg 100 g body wt. and xylazine 2.8
septum, the diagonal band of Broca, and the nucleus mg 100 g body wt. and placed into a standard stereotaxic
basalis magnocellularis see Ref. [33] for review. These device. Twenty animals ten estrogen E-treated, ten non-
neurons are the major source of cholinergic innervation to E-treated received unilateral lesions of the nucleus basalis
the hippocampus and neocortex [69] and play an important magnocellularis NBM by intracerebroventricular infusion
role in learning, memory, and attentional processes of ibotenic acid. Briefly, the skull was exposed and a hole
[6,8,36,47,63,66]. Recent studies have shown that estrogen was drilled 1.3 mm posterior to Bregma and 2.4 mm to the
replacement can enhance the functional status of basal left of the midline. A stainless steel cannula 28 gauge;
forebrain cholinergic projections as indicated by increases Small Parts, Inc. containing ibotenic acid 10 mg ml in
in choline acetyltransferase mRNA and protein in the sterile filtered saline was lowered 7.5 mm from the skull
medial septum
and nucleus
basalis magnocellularis
into the NBM. One microliter 10 mg of ibotenic acid [16,12,38], and by increases in high affinity choline uptake
solution was infused into the NBM at a rate of 12 ml h. [46,56] and acetylcholine release [24] in the hippocampus
Following infusion, the cannula was left in place for 5 min and overlying cortex. Treatment with estrogen and with
and then withdrawn very slowly. The skin overlying the estrogen plus progesterone have also been shown to
skull was sutured shut and the animals were placed onto a increase the expression of trkA nerve growth factor
warm heating pad during recovery. Two separate control receptor mRNA [18,40] in the medial septum and to
groups were prepared. Cannula controls consisting of five increase relative levels of brain-derived neurotrophic factor
E-treated and five non-E-treated animals had an empty mRNA in the hippocampus [18,20,57], suggesting that
infusion cannula lowered into the NBM to control for estrogen may enhance basal forebrain cholinergic projec-
damage produced by lowering the cannula. Sham controls tions in part by increasing trophic support for the choliner-
consisting of six E-treated and five non-E-treated animals gic neurons.
were anesthetized and placed into the stereotaxic ap- These findings suggest that one mechanism by which
paratus, the skull was exposed and then the skin was estrogen replacement can help to reduce the risk of
sutured shut. Following surgery, all animals were given Alzheimer’s-related dementia in postmenopausal women is
Torbugesic 0.02 cc 100 g twice a day for 3 days to by increasing the functional status of cholinergic projec-
reduce discomfort associated with the surgical procedures. tions to the hippocampus and neocortex. Whether estrogen
replacement also reduces the loss of cholinergic neurons 2.3. Fimbria fornix transections
associated with Alzheimer’s disease, and or protects the cholinergic neurons from neurotoxic and mechanical in-
Seven to ten days following capsule implantation, jury, is currently unknown. The purpose of this study was
animals n530 were anesthetized with a combination of to examine potential neuroprotective effects of estrogen
ketamine 14 mg 100 g body wt. and xylazine 2.8 replacement on basal forebrain cholinergic neurons by
mg 100 g body wt. and placed into a standard stereotaxic examining the ability of estrogen replacement to reduce
device. Unilateral transections of the fimbria fornix were losses of ChAT positive neurons following ibotenic acid
performed as previously described [22]. Briefly, the skull injection into the nucleus basalis magnocellularis and
was exposed and a trench was drilled extending laterally following transection of the fimbria fornix.
from midline 1.0 mm posterior to the coronal suture. An
P . Aggarwal, R.B. Gibbs Brain Research 882 2000 75 –85
77
incision was made through the dura to expose the brain. method as previously described [23] to verify that the
An ultrathin microknife Tieman was lowered 6.0 mm unilateral transections were complete.
into the brain at midline. Care was taken to avoid the midsagittal sinus which was gently pushed to the side. The
2.5. ChAT immunostaining knife was moved laterally along the trench to bone and
then removed. The trench was packed with gelfoam soaked ChAT was detected as previously described [12] using
with sterile saline. The skin was sutured shut and the an affinity purified goat polyclonal antibody raised against
animals were placed onto a warm heating pad during rat ChAT Chemicon Pharmaceuticals, AP144P; diluted
recovery. 1:3500 in 50 mM PBS pH 7.2 containing 0.05 Triton
X-100 and 1 normal rabbit serum. Briefly, sections were 2.4. Tissue preparation
placed in the primary antibody solution for 3 days at 48C. The sections were rinsed well with PBS and then incubated
Ten of the ibotenate-treated animals five E-treated and with a biotinylated rabbit-anti-goat antibody Vector Lab-
five non-E-treated, 18 of the fimbria fornix transected oratories, diluted 1:1000 for 1 h at room temperature.
animals nine E-treated, nine non-E-treated, and all of the Sections were rinsed again with PBS and then incubated
controls were killed 2 weeks post-surgery. The remaining with an avidin–HRP complex Vectastain Elite kit; Vector
ten ibotenate-treated animals five E-treated and five non- Laboratories for 1 h at room temperature. Sections were
E-treated and 12 fimbria fornix transected animals six rinsed with PBS, placed in 50 mM Tris–HCl pH 7.6
E-treated, six non-E-treated were killed 12 weeks post- containing 3,39-diaminobenzidine DAB; 0.5 mg ml, and
surgery. For euthanasia, the animals were anesthetized then reacted with 50 mM Tris containing DAB 0.5 mg
with pentobarbital 100 mg kg and perfused with saline ml, H O 0.01, and NiCl 0.032 for 7–10 min.
2 2
2
followed by 4 paraformaldehyde in 50 mM sodium The sections were then rinsed with PBS, mounted onto
acetate pH 6.5, and then 4 paraformaldehyde in 50 mM Superfrost plus slides Fisher Scientific, counterstained
Tris pH 9.0. Blood was collected by cardiac puncture with cresyl violet Sigma, dehydrated, and coverslipped.
from two E-treated animals killed after 3 weeks of estradiol treatment to verify that the estrogen capsules
2.6. Data analysis were producing the expected levels of circulating estradiol.
All capsules were inspected and the presence of estradiol 2.6.1. Ibotenate-treated animals and related controls
crystals in the E-capsules was verified. Note that these Sections through the MS and NBM of ibotenate-treated
capsules have been shown to continue releasing estradiol animals and related controls were analyzed using unbiased
even after long-term implantation [19]. The brains were stereological techniques. Briefly, the total number of
removed and postfixed for 3–4 h with 4 paraformal- ChAT-positive profiles in the NBM and MS on each side
dehyde in phosphate buffer 50 mM, pH 7.2 at 48C and of the brain was estimated using the Optical Fractionator
then stored overnight in 15 sucrose in 50 mM phosphate- method [30]. The analysis was performed using a Leitz
buffered saline PBS; pH 7.2 at 48C. Laborlux-S microscope equipped with a Ludle Biopoint
For the ibotenate-treated animals and controls, 40-mm motorized x-y-z-stage connected to a Power Macintosh G3
coronal sections were cut through the MS including plates computer running the Neurozoom software [43]. Three
15–17 of Paxinos and Watson [48] and the NBM includ- equivalently spaced sections through the MS 120 mm
ing plates 24–25 of Paxinos and Watson [48] of each apart, beginning at approximately plate 15 of Paxinos and
brain and every third section was processed for immuno- Watson [48] and six equivalently spaced sections through
cytochemical detection
of choline
acetyltransferase the NBM 120 mm apart, beginning at approximately plate
ChAT as described below. Unbiased stereological pro- 23 of Paxinos and Watson [48] were analyzed from each
cedures were then used to quantify the number of ChAT- brain. The dorsal and ventral boundaries of the septum
positive cells detected in the MS and NBM of each animal were defined at low magnification by the corpus callosum
see below. Brains from the fimbria fornix transected and by a line connecting the anterior commisures. The
animals were processed prior to the use of stereological medial boundary was defined by drawing a line down the
procedures in our laboratory. For these animals, 25-mm middle of the septum Fig. 1. The boundaries of the NBM
coronal sections through the MS were cut and every third were defined at low magnification by drawing lines
section was processed for immunocytochemical detection connecting the globus pallidus to the reticular nucleus of
of ChAT. These sections were analyzed by counting the the thalamus and the substantia innominata Fig. 2. Once
number of ChAT-positive profiles in the MS of three the borders of each region were outlined, 20 of the total
matched sections animal and then comparing the number area was analyzed in each tissue section on each side. The
of immunoreactive cells section see below. Sections dissector, which consists of two planes separated by a
through the hippocampus of the fimbria fornix-transected specified distance, was set at 10 mm less than the
animals were stained for the detection of acetylcholinester- minimum diameter of a ChAT-positive cell. A counting
ase using a modified copper thiocholine identification box was set at 10.37 mm by 7.44 mm and consisted of two
78 P
Fig. 1. Bright-field photomicrographs showing ChAT-positive profiles detected in the MS at low A and high B power. The borders used for
stereology are also outlined in A. Scale bars: A51 mm; B562.5 mm.
forbidden lines and two inclusive lines. Objects completely inside the counting frame were counted whereas objects
completely outside the counting frame were excluded. Objects which crossed through any of the three forbidden
planes bottom, front, and left sides of the frame were excluded. Objects which crossed through any of the
remaining planes top, right, and back sides of the frame were included if they either came into focus or went out of
focus when focusing through the dissector [68]. The nucleus of each labeled cell was used as the reference
point for counting. An estimate of the total number of
Fig. 2. Bright-field photomicrographs showing ChAT staining detected in
ChAT-positive neurons in each region was then calculated
the NBM contralateral A and ipsilateral B to the ibotenate infusion.
with the following formula: Total ChAT-positive cells
Arrows in A and B indicate the borders used for stereology. C
region5Total profiles counted351 percentage of total
High-power magnification of ChAT-positive profiles in the NBM con- tralateral to the lesion. Scale bars: A,B51 mm; C562.5 mm.
area3separation factor 3 for the NBM and 2 for the MS. Statistical analyses were performed using SYSTAT 5.2
for Macintosh. The effects of surgical manipulation ibote- relative to the controls was analyzed by one-way ANOVA.
nate vs. cannula vs. sham and hormone treatment E vs. Individual group comparisons were made using the Tukey
no E on the number of ChAT-positive cells in the MS and test.
NBM at 2 weeks post-surgery were analyzed by three- factor Hormone Treatment3Surgical Condition3Side
2.6.2. Fimbria fornix transected animals ANOVA with repeated measures on ‘Side’. The effects of
Sections through the medial septum of animals with a ibotenic acid 2 weeks and 3 months following injury
unilateral transection of the fimbria fornix were analyzed
P . Aggarwal, R.B. Gibbs Brain Research 882 2000 75 –85
79
as previously described [12,17,25]. Briefly, the average treated animals. Significant interactions between Surgical
number of ChAT-positive profiles on each side of the Condition3Side F2,255112.7, P,0.001, and Hor-
MS section was determined for each animal by counting mone Treatment3Surgical Condition3Side F2,255
the number of labeled profiles with an identifiable nucleus 7.01, P,0.005 were also detected.
in each of three matched sections animal. Group means Subsequent post hoc analyses revealed no significant
were calculated and the effects of hormone treatment and interaction between Hormone Treatment3Surgical Con-
fimbria transection were compared using analysis of dition on either the intact F2,2552.60, P50.10 or the
variance with repeated measures on ‘side’. lesioned F2,2550.26, P50.78 side, contrary to the
three-way interaction detected in the ANOVA. Additional post hoc analyses revealed a significant effect of Surgical