Investigative Toxicologic Pathology

Integrated Evaluation of Central Nervous System Lesions:

Stains for Neurons, Astrocytes, and Microglia Reveal
the Spatial and Temporal Features of MK-801-induced
Neuronal Necrosis in the Rat Cerebral Cortex*
ANDREW S. 1 JOSEPH F. ROSS,
FIX, 1 SUSAN R. STITZEL,
1 AND ROBERT C. SWITZER2

The Procter &

1 Gamble Co., Miami Valley Laboratories, Cincinnati, Ohio 45253, and
2
N euroscience Associates, Knoxville, Tennessee 37922

ABSTRACT

Routinely processed, hematoxylin and eosin (H&E)-stained slides are typically used to assess the morphologic integrity of the
central nervous system in neurotoxicity safety studies. However, the value of special stains for improving neuropathologic evaluations

during the assessment of neurotoxicity has been emphasized in the neuroscience literature and by regulatory agencies. The primary
objective of the present study was to characterize the spatial and temporal changes in neurons, astrocytes, and microglia after dizocilpine
maleate (MK-801)-induced focal neuronal necrosis in the posterior cingulate/retrosplenial (PC/RS) cortex of the rat. A secondary
objective was to evaluate the application of special stains and a novel sectioning procedure for detecting neurotoxicity. Sixty adult
male Sprague-Dawley rats were treated with sterile water vehicle or 10 mg/kg MK-801 and perfused through the left ventricle (pumped
at 65 mm Hg pressure) with 10% neutral buffered formalin or 4% paraformaldehyde at 4 hr and on days 1, 3, 7, 14, and 28 after
treatment. For light microscopic evaluation, brain sections were stained with H&E, a special cupric-silver (CS) stain that selectively

impregnates degenerating neurons and makes them readily evident, glial fibrillary acidic protein (GFAP) immunohistochemistry for
astrocytes, and Griffonia simplicifolia isolectin B 4 (GSA) histochemistry for microglia. Brains perfusion-fixed with 4% paraformal-
dehyde were prepared for CS staining with a novel frozen-sectioning procedure for multiple embedding in a composite gelatin block.
In H&E sections from treated rats, necrotic nerve cell bodies were observed in PC/RS cortical layers 3 and 4 on days 1, 3, 7, and
14, but not on day 28. These necrotic neurons required high magnification for detection (X20 objective, X 10 ocular). In contrast,
degenerating neurons selectively stained with CS were observed in the same location as necrotic neurons seen with H&E but at low
magnification (X2 objective, X10 ocular). Cupric-silver staining showed details not seen with H&E, including dendritic and axonal
degeneration with progressive fragmentation. Beginning on day 3, GFAP immunohistochemistry revealed hypertrophic astrocytes in
a diffuse pattern throughout the region of cell body necrosis, a change that persisted throughout the study. However, GSA lectin

histochemistry identified a few reactive microglia on day 1 in a multifocal pattern throughout the region of cell body necrosis. Reactive
microglia were observed on days 3, 7, and 14, but not on day 28. Glial changes observed with H&E staining were limited to an
increase in the cellularity of glial cell nuclei in the area of neuronal necrosis. This study provides a comprehensive and integrated
view of the temporal changes occurring in neurons, astrocytes, and microglia during acute neurotoxic injury. Moreover, advantages
for using new staining and sectioning methodologies to enhance the toxicologic evaluation of the central nervous system are dem-
onstrated.

Keywords. Brain; cupric-silver stain; dizocilpine maleate; glial fibrillary acidic protein (GFAP); GFAP immunohistochemistry;
lectin histochemistry; neuropathology; neurotoxicity; risk assessment; toxicity

INTRODUCTION structural heterogeneity of the central nervous system

Heightened concern about the potential impact of neu-
rotoxicants on human health (29, 32) has prompted re-
vision of guidelines for neurotoxicity testing (25, 56) and
increased the frequency of comprehensive neuropatho-
logic evaluations in safety studies. Most toxicologic pa-
thologists are challenged by the anatomic scope and
*
Address correspondence to: Dr. Andrew S. Fix, Procter and Gamble
Co., Miami Valley Laboratories, P.O. Box 538707, Cincinnati, Ohio
45253-8707.
291
(CNS) and the fact that many neurotoxicants have highly
localized morphologic and/or functional effects. While
the vast majority of safety studies have utilized traditional
sectioning, staining, and evaluation procedures, new ap-
proaches may assist the toxicologic pathologist in meet-
ing the enormous challenge presented in evaluating the
CNS.
A growing body of experimental literature in the field
of neuroscience confirms the ability of special neuronal
and glial markers to demonstrate neuronal injury (31).
292
TABLE I.-Summary
a
Six total time points evaluated after treatment (4 hr; 1, 3, 7, 14, and 28 days).
Several of these special techniques may be appropriate
for identifying neurotoxicant-induced damage in regulat-
ed safety studies, including cupric-silver (CS) stains that
selectively impregnate neurons undergoing primary de-
generation (3, 5, 62) and histochemical stains for astro-
cytes and microglia that reveal secondary responses to
CNS injury (12, 46, 58, 64). In contrast to astrocytes and
microglia, oligodendroglia have not received the attention
in the literature regarding their ability to demonstrate sec-
ondary responses to neurotoxic injury. Special procedu-
res for staining neurons, astrocytes, and microglia may
also have the potential to reveal morphologic changes
that otherwise might not be evident with routine histo-
pathology. Although previous studies have demonstrated
the potential utility of individual special stains for de-
tecting evidence of neurotoxicity, few studies have at-
tempted to integrate information derived from multiple
special stains with routine histopathology to better un-
derstand the varied and complex changes that occur after
acute neurotoxicity.
The induction of focal neuronal necrosis by MK-801
exemplifies the highly selective vulnerability of a specific
CNS site to the action of one neurotoxicant. MK-801 is
a potent noncompetitive antagonist of the N-methyl-D-
aspartate (NMDA) receptor (52) and is a prototype of a
drug class with the potential to treat stroke and related
cerebrovascular diseases (7). In rats, single doses of MK-
801 produce dose-dependent vacuolization and necrosis
of neurons in layers 3 and 4 of the posterior cingulate/
retrosplenial (PC/RS) cortex (16, 18, 19, 51). Cytoplas-
mic vacuoles appear in this small cerebral cortical region
within 4 hr of treatment, but are not seen at 24 hr (51).
Higher doses of MK-801 induce neuronal necrosis in the
same neuronal population; however, this effect is not ev-
ident until 1 or 2 days after vacuolization is no longer
observed (18). Careful light microscopic examination is
required to detect the eosinophilic necrotic neurons in
routine paraffin sections stained with hematoxylin and
eosin (H&E) (18).
Studies of MK-801 neurotoxicity using special stains
are very limited and have not characterized temporal ef-
fects after treatment. Morphologic changes detectable by
the selective CS impregnation stain for neuronal degen-
eration (mentioned above) have not been reported for
MK-801. Moreover, astrocytic and microglial responses
to MK-801-induced neuronal necrosis have only been
characterized superficially with histochemical techniques.
An enzyme-linked immunosorbant assay (ELISA) using
fresh brain tissue has been used to determine dose-related
and temporal changes in glial fibrillary acidic protein
of experimental design and procedures.
(GFAP, the major intermediate filament protein in astro-
cytes) after MK-801 treatment (22). However, that report
only identified increased astrocytic staining with GFAP
immunohistochemistry at 1 time point (9 days after treat-
ment). Similarly, another study reported the dose-related
activation of microglia shortly after treatment with MK-
801, but did not provide information regarding temporal
responses (41).
The primary objective of this study was to use special
stains to characterize the spatial and temporal changes in
neurons, astrocytes, and microglia as a result of MK-801-
induced neurotoxicity. A secondary objective was to
demonstrate the integrated application of special stains
and a novel sectioning procedure for neurotoxicity as-
sessment. Specific goals were: 1) to evaluate the ability
of the CS staining method to demonstrate neurotoxicant-
induced damage relative to H&E findings, 2) to compare
responses made by astrocytes and microglia while relat-
ing those responses to changes in the different parts of
injured neurons, and 3) to assess a novel histologic tech-
nique which generates composite frozen sections from
multiple rodent brains embedded in a single gelatin
block. Portions of this work have been presented previ-
ously in abstract form (20, 21).
METHODS
General. Sixty adult male Sprague-Dawley rats (at
least 90 days old, 361-508 g; Harlan, Indianapolis, IN)
were used. Rats were housed individually in hanging cag-
es with wire mesh floors and maintained on a 12/12 hr
light/dark photoperiod cycle. Food pellets and water were
available ad libitum during the study. All procedures
were conducted in compliance with the National Institute
of Health guidelines for the use of animals in research
(65).
The experimental design for this study is summarized
in Table I. Briefly, rats were treated subcutaneously with
a single injection of sterile water vehicle (n =
24) or 10
mg/kg MK-801 [5R, lOS-(+)-5-methyl-10,11-diben-
zo[a,d]cyclohepten-5,10-imine hydrogen maleate] (n =
36; dizocilpine maleate; Research Biochemicals Inc., Na-
tick, MA). At 4 hr and days 1, 3, 7, 14, and 28 after
on
treatment, rats were anesthetized
deeply with sodium
pentobarbital and perfused through the left ventricle with
fixative solutions specific for embedding individual brain
slices in paraffin or for composite embedding multiple
whole brains in large gelatin blocks. All perfusates were
delivered with a pressure pump system at 65 mm Hg.
Each of the 2 perfusion and embedding procedures was

done with an identical set of brains from similarly treated
rats (2 controls and 3 treated for each of 6 time points).
Paraffin Embedding. Rats ( 12 control, 18 treated) were
perfused with heparinized phosphate buffer to remove
blood from the vasculature ( min) followed by 10% neu-
tral buffered formalin (8-10 min). After postfixation in
10% neutral buffered formalin containing 1 % zinc sulfate
(24 hr), brains were removed and trimmed in 2-mm slices
through the PC/RS cortex using a commercial rat brain
matrix (RBM-4000, ASI Instruments, Inc., Warren, MI).
Since the extent of MK-801-induced neuronal necrosis
has recently been shown to vary significantly along the
rostrocaudal extent of the PC/RS cortex (23), sections
examined were standardized so they were all from ap-
proximately the middle of the susceptible region, near the
caudal termination of the corpus callosum (53). Trimmed
slices were dehydrated through graded alcohols, embed-
ded in paraffin, and sectioned at 5 [Lm. Sections were
stained with H&E.
Detection of astrocytes by GFAP immunohistochem-
istry was based on adaptation of a previously published
procedure (22). Paraffin sections (5 J.Lni) were collected
(ProbeOn Plus slides) and loaded onto an automated im-
munostainer (Code-On, BioTek Solutions, Inc., Santa
Barbara, CA). Following deparaffinization, hydrogen per-
oxide treatment to reduce endogenous peroxidase activi-
ty, and antigen retrieval (Target Unmasking Fluid, Signet,
Dedham, MA), slides were incubated in a 1:1,500 dilu-
tion of primary polyclonal rabbit anti-cow GFAP anti-
body (DAKO, Carpinteria, CA) for 30 min, followed by
a goat anti-rabbit secondary antibody and an avidin biotin
complex (details in Vectastain Elite Kit, Vector, Burlin-
game, CA) for 15 min each. All incubations were at 37°C.
Slides were treated with diaminobenzidine tetrahydro-
chloride (DAB) and counterstained with Mayer’s hema-
toxylin. For positive controls, endogenous GFAP staining
in normal astrocytes was observed throughout the brain
in all sections. Substitution of primary antibody with au-
tomation buffer produced no staining in negative con-
trols.
To detect microglia, an automated lectin histochemical
procedure was developed for paraffin sections based on
previously published procedures for thick vibratome (58)
and paraffin (61) sections. Paraffin sections (5 [Lm) were
prepared as described above for GFAP Following depar-
affinization, slides were immersed in O.1M PBS (Sigma)
containing cations (0.1 mM each of CaCl2, MgCl,, and
MnCl2) and 0.1 % Triton-X 100 for 30 min. Slides were
then incubated for 2 hr in 10 pLg/ml peroxidase-labeled
lectin from Griffonia simplicifolia (GSA; sold by Sigma
as lectin from Bandeiraea simplicifolia B S-Isolectin B4;
catalog # L-5391; specific for a-D-galactosyl residues),
treated with the chromagen DAB, and counterstained
with Mayer’s hematoxylin. All incubations were at room
temperature. For positive controls, endogenous GSA lec-
tin staining in normal microglia was observed throughout
the brain in all sections. Substitution of lectin with au-
tomation buffer produced no staining in negative con-
trols.
Gelatin Embedding for Composite Frozen Sectioning.
Rats (12 controls, 18 treated) were perfused through the
293
FIG. 1.-Composite freeze-cut section of 15 brains at the same ros-
trocaudal level stained with the CS procedure. To produce this section,
15 perfusion-fixed whole brains were embedded with similar orientation
in a large gelatin block. Note the rather consistent anatomic alignment
of individual brain sections, facilitating comparison among individual
sections. x 1.2.
left ventricle with a solution of 0.8% NaCl and 0.8%
sucrose in cacodylate buffer to remove blood from the
vasculature (1 min) followed by a fixative solution of
4.0% paraformaldehyde and 4.0% sucrose in the same
buffer (8-10 min). Brains were left in the skull for 2-4
days while immersed in the same fixative. After removal,
brains were cryoprotected in a glycerol-based solution for
6-8 hr. Whole brains were embedded collectively in 2
large gelatin blocks (15 brains per block) with parallel
alignment of their rostrocaudal axes for coronal section-
ing (MultiBrain Technology, NeuroScience Associates,
Knoxville, TN). Although this procedure has been used
before (13, 63), it has not been described in detail or
illustrated as done in the present study. Gelatin blocks
were hardened in a 4% formaldehyde solution for 48
hours, then frozen on dry ice. With an American Optical
sliding microtome, 20- or 40-pLm composite freeze-cut
sections were taken through the PC/RS cortex in the area
near the caudal termination of the corpus callosum [a
landmark which delineates the approximate middle of the
cortical area targeted by MK-801, see Fix et al (23)]. This
region matched the rostrocaudal level of the sections se-
lected for evaluation in paraffin (see above). Since all 15
brains had similar orientation in each gelatin block, each
composite frozen section provided a set of 15 individual
cross-sections at approximately the same level of the
brain (Fig. 1). Brains were distributed between the 2 gel-
atin blocks so that each composite frozen section con-
tained control and treated rats at each time point. This
equalized staining conditions for all sections and allowed
direct, side-by-side comparison of control and treated rats
on the same slide. Twenty-micron composite sections
were stained routinely with H&E, while 40-pLm sections
were stained by the CS method specific for neuronal de-
generation and by GFAP immunohistochemistry for as-
trocytes (see below).
Cupric-silver staining for the selective identification of
 295

degenerating neurons was based on previously published
procedures (10, 11). Briefly, 40-pLm frozen sections were
collected and stained free-floating. Sections were rinsed
in deionized water and placed in an aqueous mixture of
silver and copper nitrate, pyridine, and ethanol for 4 days.
Sections were then processed through acetone for 2 min
followed by a diammine solution of silver nitrate in com-
bination with ammonium and sodium hydroxide for 7
min, then reduced in a weak formaldehyde solution con-
phologic changes were detectable at the end of the study
(day 28), H&E slides were evaluated without knowledge
of treatment condition. After treatment group assign-
ments were provided, interpretation of these data was
made. CS-stained slides were evaluated for degenerative
changes in nerve cell bodies, dendrites, and axons. Since
the CS staining procedure selectively impregnates degen-
erating neurons, specific positive staining was interpreted
as being indicative of a lesion. For sections stained by

taining citric acid and ethanol for 1.5 to 2 min. After
several rinses, sections were bleached in a solution of
potassium ferricyanide and sodium borate to remove any
unreduced silver. Following several additional rinses, sec-
tions were mounted on 2 X 3-inch glass slides (Baxter
Diagnostic Products, Inc., McGaw Park, IL), air-dried,
and coverslipped (Permount, Fisher Scientific, Pittsburgh,
PA).
For GFAP immunohistochemistry, 40-pLm sections
were collected and stained free-floating. After hydrogen
peroxide treatment, sections were immunostained with a
1:2,000 dilution of primary polyclonal rabbit anti-cow
GFAP antibody (DAKO, Carpinteria, CA), a goat anti-
rabbit secondary antibody and an avidin-biotin complex
(details in Vectastain ABC kit, Vector, Burlingame, CA).
Incubation times were 48 hr for the primary antibody (at
4°C), 30 min for the secondary antibody (at room tem-
perature), and 1 hr for the avidin-biotin complex (also at
room temperature). Sections were treated with DAB and
mounted on gelatinized (subbed) glass slides as described
above for CS staining. For positive controls, endogenous
GFAP staining in normal astrocytes was observed
throughout the brain in all sections. Substitution of pri-
mary antibody with buffer produced no staining in neg-
ative controls.
Light Microscopy. All evaluations were by qualitative
light microscopy and were limited to sections from the
posterior cingulate/retrosplenial (PC/RS) cortex. For each
stain at each time point, sections from treated rats were
compared to corresponding control rats. In H&E-stained
slides, neurons were evaluated for cytoplasmic vacuol-
ization, overall size, and tinctorial alterations indicative
of necrosis. The relative number and prominence of per-
ineuronal glial nuclei was also assessed, particularly in
relation to necrotic neurons (reported as glial hypercel-
lularity). These nuclei were considered to be of glial or-
igin based on previously published electron microscopic
data using the same neurotoxin (18). However, the spe-
cific identification of each glial cell type was not attempt-
ed in the H&E-stained slides. To determine whether mor-
GFAP immunohistochemistry, relative staining intensity,
thickness of cellular processes, and extent of perinuclear
cytoplasmic staining were evaluated for astrocytes. When’
increases in these were found, astrocytes were considered
to be hypertrophic and reactive. The relative prominence
and thickness of cellular processes and the extent of peri-
nuclear cytoplasmic staining was also evaluated for mi-
croglia by GSA lectin histochemistry. When increases in
these were found, microglia were considered to be reac-
tive. Morphologic changes in neurons, astrocytes, and mi-
croglia were scored based on the following grading sys-
tem : -, absent or same as controls; +, scant/occasional
presence or slight increase; + +, moderate presence or
increase; + + +, prominent presence or increase. Individ-
ual scores were averaged to determine a group score for
tabular presentation.
RESULTS
Morphologic changes in the posterior cingulate/retro-
splenial (PC/RS) cortex observed by all staining tech-
niques 3 days after treatment with 10 mg/kg MK-801 are
summarized in Fig. 2. Neuronal necrosis (Fig. 2A) was
observed by H&E; selective nerve cell body, dendritic,
and axonal degeneration was seen with CS (Fig. 2B);
increased GFAP immunoreactivity was observed in as-
trocytes (Fig. 2C); and increased Griffonia simplicifolia
BS-1 Isolectin B4 (GSA) staining was seen in microglia
(Fig. 2D). Temporal changes in neurons, astrocytes, and
microglia observed with the different staining procedures
are presented in Table II. Detailed morphologic descrip-
tions of these changes follow.
Hematoxylin and Eosin (Paraffin and Composite
Freeze-Cut Sections)
No structural abnormalities were seen in control rats.
Morphologic changes in treated rats (5-pLm paraffin-em-
bedded sections stained with H&E) were observed only
at high magnification (X20 objective, X 10 ocular) and
involved subtle changes in a small number of neurons.

FIG. 2.-Photomicrographic array summarizing changes in the PC/RS cortex by different histochemical staining techniques 3 days after treatment
with 10 mg/kg MK-801. Stains include H&E (A), CS for neuronal degeneration (B), GFAP immunohistochemistry for astrocytes (C), and GSA
histochemistry for microglia (D). Note the relative obscurity of the eosinophilic necrotic neurons with H&E (arrows in A and the corresponding
inset) compared to the prominent CS-positive degenerating neurons in B. CS staining also reveals degeneration of dendrites (arrows) and superficial
apical tufts (asterisks) not evident with H&E, as well as dense staining of degenerating nerve cell bodies (B, inset). Reactive astrocytes (GFAP

stain in C) and microglia (GSA stain in D) are confined to the area of neuronal degeneration and necrosis. However, within that area, astrocytic
hypertrophy is diffuse while microglial reactivity is more multifocal. Hypertrophic astrocytes (C, right inset) have thickened processes compared
to those from a control rat (C, left inset). Reactive microglia (D, right inset) surround necrotic neuronal debris and have abundant cytoplasm
compared to those from a control rat (D, left inset). x 178; insets x357.
296

TABLE II.-Summary of morphologic changes in neurons, astrocytes, and microglia in the posterior cingulate/retrosplenial cortex of rats treated
with 10 mg/kg MK-801 ,a

a Average group scores determined by the following grading system (2 control, 3 treated rats per time point): -, absent or same as controls; +, scant/occasional
presence slight increase; + +, moderate presence or increase; + + +, prominent presence or increase.
or
b
Morphologic changes confined to PC/RS cortical layers 3 and 4 (5-[Lm paraffin sections).
I
Stains degenerating components of neurons (40-~Lm composite freeze-cut sections-see text for details).
d
Reactive astrocytes diffusely distributed throughout area of neuronal necrosis (40->m composite freeze-cut and 5-f-Lm paraffin sections).
e
Reactive microglia multifocally distributed throughout area of neuronal necrosis (5-f-Lm paraffin sections).

These findings were consistent with those previously re-
ported at comparable time points (18) and are described
here in order to provide a basis for comparison with spe-
cial stains. Briefly, changes were confined to a subset of
pyramidal neurons in the PC/RS cortex (cortical layers 3
and 4 only). At 4 hr, affected neurons had pale, foamy
cytoplasm containing variably sized cytoplasmic vacu-
oles discernible only by careful evaluation. At day 1, vac-
uoles were not seen, but occasional neurons in layers 3
and 4 were necrotic (shrunken and eosinophilic cell bod-
ies). On days 3 (Fig. 2A), 7, and 14, more necrotic neu-
rons were evident throughout the susceptible region, al-
though these were smaller and more shrunken at the later
time point. There was slight variability in the extent of
necrosis among treated rats. An increase in the number
and prominence of perineuronal glial nuclei around some
of the necrotic cell bodies was evident as early as day 3.
These nuclei were more prominent on days 7 and 14;
however, this change did not persist to the end of the
study (day 28). Evaluation of day 28 slides without
knowledge of treatment condition failed to identify
changes in the morphology or number of neurons be-
tween control and treated rats.
In 20-pLm composite freeze-cut sections stained with
H&E, necrotic neurons were evident in treated rats as
described above for paraffin sections, except that finding
these neurons was more difficult. The necrotic neurons
were less obvious because the thicker composite sections
reduced the contrast between the necrotic neurons and
the background neuropil.
Cupric-Silver (Composite Frozen Sections)
All silver-stained slides had a uniform light golden
brown background color that revealed sufficient cytoar-
chitecture to determine neuroanatomy without the need
for a counterstain (Fig. 3). Some nonspecific background
silver-positive staining was observed in control and treat-
ed rats at all time points. Fine silver-positive stippling
was observed in meninges and perivascular areas. The
exact cellular origin of this staining could not be deter-
mined. Isolated red blood cells retained in clear, dilated
capillaries were also silver-positive (Fig. 3A). The loca-
tion of these cells was confirmed by focusing up and
down through the entire thickness of the 40-pLm frozen
sections. Short portions of axons deep in the PC/RS cor-
tex near or in the cingulum and corpus callosum were
stained also. These latter axons were dark brown and typ-
ical of the silver-positive unaffected axons (called &dquo;nor-
mal fiber staining&dquo;) previously reported with this proce-
dure (10).
Cupric-silver staining revealed selective degeneration
of neurons in the PC/RS cortex (Fig. 3). No specific sil-
ver-positive staining was observed in PC/RS cortical neu-
rons from control rats at any time point (Fig. 3A). In
treated rats at 4 hr (time of vacuolization), no specific
silver-positive staining was observed. However, on day
1, all treated rats had isolated, individual silver-positive
degenerating neurons in cortical layers 3 and 4 (Fig. 3B).
The cell bodies of these degenerating neurons exhibited
variable staining intensity (light gray to black) and were
seen easily at low magnification (X2 objective, X 10 oc-
ular). In addition, slightly positive silver staining was not-
ed for degenerating apical dendrites (the orienting den-
dritic processes that extend superficially through cortical
layer 2 into layer 1) and apical tufts (the branched ter-
mination of each apical dendrite in layer 1). Degenerating
basal dendrites (processes that extend laterally from cell
bodies within layers 3 and 4) stained slightly positive on
day 1. On day 3, the overall staining of degenerating cell
bodies and dendrites was more intense and involved more
neurons than on day 1 (Fig. 2B). Degenerating neurons
were more shrunken and angular than on day 1. In ad-
dition, isolated silver-positive, corkscrew-shaped degen-
erating axons were evident in deeper cortical layers near
the corpus callosum. On day 7, axons had a prominent
corkscrew or beaded appearance, but staining of cell bod-
ies and associated dendrites was less intense than at ear-
lier times (Fig. 3C). Neurons were no longer intact, as
 297

FIG. 3.-Time course of neuronal changes demonstrated by CS staining in the PC/RS cortex of rats treated with vehicle (A) or 10 mg/kg MK-
801 (B, C, D). No CS-positive staining is evident in vehicle-treated controls, except occasional remnant erythrocytes (marked with asterisks in A
and B). These were consistently located within dilated capillaries and were observed clearly by focusing up and down through the 40-)JLm-thick
sections. At day 1 (B), isolated silver-positive degenerating neurons are evident in the same location as necrotic neurons found with H&E. Note
the fine staining of irregular degenerating apical (a) and basal (b) dendrites. By day 7 (C), silver-positive degenerating cell bodies (d) are shrunken
and dendritic fragmentation is prominent (fine stippling around degenerating nerve cell bodies). At day 14 (D), cell bodies and dendrites remain as
silver-positive debris (d). Note that silver-positive staining of degenerating axons is evident in the deeper portions of the cortex on day 7 (arrow in
C and inset) and is still present at day 14 (arrow in D and inset). x 178; insets X357.
298
degenerating cell bodies appeared slightly fragmented
and had lost direct connections with both apical and basal
dendrites. On day 14, silver-positive debris persisted in
cortical layers 3 and 4 and cell bodies remained as small
dense structures or partial fragments (Fig. 3D). Staining
of basal dendrites appeared dustlike around moderately
fragmented cell bodies. Small portions of silver-positive
apical dendrites were broken and beaded, while apical
tufts were difficult to observe. In contrast to the overall
fragmentation and light staining observed in cell bodies
and dendritic processes, silver-positive axons remained
prominent near the corpus callosum (Fig. 3D). On day
28, fragmented cell bodies and dendrites remained as
scant clumps of silver-positive debris in cortical layers 3
and 4 while virtually no silver-positive staining was ev-
ident in cortical layer 1. However, axonal staining in the
area of the corpus callosum was still evident, but slightly
lessprominent than on day 14. There was a slightly great-
er degree of variability in the number of silver-positive
neurons with CS staining than with H&E staining.
GlialFibrillary Acidic Protein (Paraffin and
Composite Frozen Sections)
Changes observed in astrocytes with GFAP immuno-
histochemistry are presented in Fig. 4. In control rats,
paraffin-embedded sections contained thin processes and
slight perinuclear cytoplasmic staining in normal, non-
reactive astrocytes (Fig. 4A). Staining was similar be-
tween control and treated rats at 4 hr and day 1, but a
slight increase in GFAP immunoreactivity was evident in
treated rats beginning on day 3 (Fig. 2C). This increase
was diffuse, but confined to a relatively narrow band in
cortical layers 3 and 4 (region of necrotic cell bodies). In
this area, all hypertrophic astrocytes had prominent peri-
nuclear cytoplasm and thickened processes (Fig. 2C, right
inset) compared to the thin, pale staining processes in
astrocytes from control rats (Fig. 2C, left inset). At all
later time points, GFAP immunoreactivity remained dif-
fusely prominent in the same region of treated rats (Fig.
4B). No difference in the intensity of GFAP immuno-
reactivity or the thickness of astrocytic processes was ap-
parent between control and treated rats in other layers of
the PC/RS cortex.
Changes in GFAP immunoreactivity in composite
freeze-cut sections paralleled those described above for
paraffin sections (Fig. 4C, D). However, the PC/RS cor-
tical zone of hypertrophic astrocytes was more prominent
in composite sections than in paraffin sections (compare
Fig. 4B and D). This was particularly evident at lower
magnification and was likely due to the increased thick-
ness of the composite freeze-cut sections, since more as-
trocytes were present per given area. The corresponding
greater density of astrocyte processes throughout the neu-
ropil produced prominent GFAP immunoreactivity in the
area of PC/RS cortical neuronal cell body necrosis that
resembled a localized increase in background staining.
Griffonia simplicifolia B4 Lectin
(Paraffin Sections)
Changes in microglia observed with Griffonia simpli-
cifolia B4 lectin staining are presented in Fig. 5. In control
rats, GSA I-B4 staining revealed nonreactive microglia
with fine, highly ramified processes extending out from
scant perinuclear cytoplasm (Fig. 5A), consistent with
previous reports using the same lectin (60). In treated
rats, lectin histochemistry revealed a slight increase in
the number and staining intensity of reactive microglia
beginning day 1. These treated rats had isolated GSA I-
B4 lectin-positive microglia that were slightly enlarged
and more clumped in appearance than those in control
rats. Like hypertrophic astrocytes, reactive microglia
were confined to PC/RS cortical layers 3 and 4 (region
of necrotic cell bodies). On day 3, prominent reactive
microglia were observed in the same area from treated
rats (Fig. 2D). These cells had prominent perinuclear cy-
toplasm and thickened processes (Fig. 2D, right inset)
compared to the thin, pale staining processes in microglia
from control rats (Fig. 2D, left inset). In some cases, a
small ring of positive lectin staining surrounded a central
clear area containing necrotic neuronal debris. Frequent-
ly, reactive microglia were located immediately adjacent
to ramified, nonreactive microglia. This observation in-
dicated that microglial reactivity was occurring in a mul-
tifocal pattern, rather than the more diffuse pattern seen
for astrocytes. Reactive microglia remained prominent on
day 7 and day 14 (Fig. 5B, C), but were smaller and
more compact at the later time point. By day 28, reactive
microglia had only a scant increase in cytoplasmic stain-
ing (Fig. 5D).
DISCUSSION
This light microscopic study characterized morpholog-
ic changes in neurons, astrocytes, and microglia after se-
lective MK-801-induced neuronal vacuolization and ne-
crosis in the rat cerebral cortex. Data from the integrated
approach described in this study demonstrate the value in
combining routine histopathology with special sectioning
and staining procedures for a more complete toxicologic
evaluation of the central nervous system. The CS stain
selective for neuronal degeneration accurately identified
the cell population injured by MK-801 and better dem-
onstrated the associated neurotoxic changes than routine
histopathology. This is consistent with the historic use of
CS stains, which were developed originally by neuroan-
atomists for use in tracing neuroanatomic tracts after ex-
perimentally induced injury to the CNS (10). The con-
clusion that the silver-positive neurons are degenerating
as a result of MK-801 treatment is based on the absence
of staining in corresponding controls and on data from
previous studies reporting morphologic changes in the
same neuronal population after similar treatment. The
vacuolization originally reported in PC/RS cortical neu-
rons (51) and further characterized by light and electron
microscopy (1, 2, 17, 18) was seen with routine histo-
pathology in the present study. Moreover, the subsequent
temporal changes in eosinophilic necrotic neurons de-
scribed herein confirm previous data (18). In addition to
these overt morphologic effects, others have described
the immunocytochemical expression of heat shock pro-
tein 72 (57) and neuronal hypermetabolism by autora-
diography (35) in the same neurons after MK-801 treat-
 299

FIG. 4.-GFAP staining on paraffin (A, B) and composite freeze-cut (C, D) sections in the PC/RS cortex of rats 28 days after treatment with
vehicle (A, C) or 10 mg/kg MK-801 (B, D). Note that the increase in GFAP immunoreactivity in treated rats (B, D) is diffuse throughout the region
of neuronal necrosis as observed with H&E (area between arrows) and is more prominent in the composite freeze-cut section (D), probably due to
increased section thickness. Mayer’s hematoxylin counterstain (A, B). x 178.

ment. Collectively, these histologic, electron microscopic, These data support the hypothesis that CS stains can ac-
immunohistochemical, and autoradiographic studies have curately identify selective neurotoxic damage (3, 5, 62).
characterized a variety of changes in the same neurons In addition, the clear demonstration of neuronal injury by
selectively identified in the present study by CS staining. CS staining suggests benefits this approach may have
300

FiG. 5.-Time course of microglial reactivity demonstrated by Griffonia simplicifolia BS-1 Isolectin B4 staining in the PC/RS cortex of rats
treated with vehicle (A) or 10 mg/kg MK-801 (B, C, D). Normal microglia have ramified, lightly staining processes in control rats (arrows in A;
also see left inset of Fig. 2D), but have short, thickened processes and abundant cytoplasm in treated rats at day 7 (B). The cytoplasm and processes
of reactive microglia are more compact by day 14 (C) and are barely detectable by day 28 (arrows in D). Note that at day 28, reactive microglia
are barely evident while the astrocytic response is still prominent (compare to Fig. 4B). Mayer’s hematoxylin counterstain. X178.

when combined with more routine approaches in neuro-
pathology screening studies, particularly concerning the
detection of focal effects in isolated neuroanatomic sites.
The detailed nerve cell body, dendritic, and axonal
changes revealed by CS staining in this study are in
agreement with previous reports describing a variety of
insults to the central nervous system produced by drugs,
chemicals, hypoxia, and trauma (3-5, 8, 45, 55). Al-
though several of these previous studies included parallel
morphologic evaluations, those morphologic comparisons
were limited to thicker sections stained with cresyl violet
and did not use the standard histologic approach familiar
to toxicologic pathologists (as done concurrently in the
present study). Thus, the lack of concurrent paraffin-em-
bedded, H&E-stained sections has made it difficult for
toxicologic pathologists to consider the use of CS stains
for neuropathologic evaluation. Moreover, CS stains,
which have been used traditionally for tracing neuroan-
atomic tracts (10), have undergone significant improve-
ment in recent years and are now more reliable and have
improved suppression of background staining (9). These
stains are still, however, very complex and technically
demanding. This complexity could be a limiting factor
for the ultimate implementation of CS stains into broader
scale neurotoxicity testing, since routine histology labo-
ratories are not properly equipped for these procedures.
Additional investment of personnel and equipment would
likely be required. Our comparative H&E findings help
validate the CS stain while demonstrating the potential
value this approach may have in neurotoxicity screening
assessments.
In this both hypertrophic astrocytes and reactive
study,
microglia wereconfined to the cortical layers where the
cell bodies of necrotic neurons were located, but were
not observed in regions containing apical dendrites or
axons of affected neurons as identified by CS staining.
This suggests that the presence or absence of a glial re-
sponse was, at least in part, a function of location relative
to affected components of injured neurons (i.e., cell body,
dendrite, or axon). Although previous studies have not
made concurrent evaluations of CS staining with histo-
chemical staining for glial cells, some studies have cor-
related CS staining changes with astrocytic responses as
measured with an enzyme-linked immunosorbent assay
(ELISA) for GFAP content in fresh brain homogenate (4,
33, 45). Such an ELISA has been proposed for routine
use in assessing neurotoxicity (42, 47). In general, studies
with trimethyltin (4), 3,3’-iminodipropionitrile (33), and
1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (45) have
found relatively good correlation between neuronal injury
(detected with CS staining) and increases in GFAP (mea-
sured by ELISA). However, since those previous studies
used fresh brain homogenates and not tissue immuno-
histochemistry for GFAP evaluation, they were unable to
reveal the anatomic relationship between neuronal
changes and reactive glial changes (as done in the present
study). A more recent study using the GFAP ELISA to
evaluate injury induced by MK-801 reported temporal
changes in GFAP similar to those found by immunohis-
tochemistry in the present study (22). That study also
used immunohistochemistry to identify a localized in-
301
crease in GFAP in PC/RS cortical layers 3 and 4, but did
it atonly 1 time point after treatment (day 9). As in the
present study, changes were not observed in other cortical
layers. Our present results indicate that various parts of
injured neurons (i.e., nerve cell bodies, dendrites, and
axons) may have significant differences in ability to ini-
tiate activation of astrocytes or microglia. This finding
would not be possible with routine H&E histopathology
alone, but required the identification of detailed neuronal
degeneration evident with CS staining and specific iden-
tification of glial cell types with special stains. Factors
involved in this differential activation are not currently
known.
Although astrocytic and microglial reactions were con-
fined to the same cortical region where injured neuronal
cell bodies were located, the morphologic character of
the response made by each of these glia differed some-
what. Hypertrophic astrocytes with prominent immuno-
reactive processes were scattered diffusely throughout the
region of cell body injury. These changes, which ap-
peared to involve all astrocytes in the area, are consistent
with classic astrocytic hypertrophy in response to local
injury (12, 14, 38). In contrast, reactive microglia were
mixed sporadically with nonreactive microglia through-
out the affected area and were associated with individual
necrotic neuronal cell bodies. This association represents
phagocytosis of remnant cell debris by reactive microglia,
as confirmed previously by us with electron microscopy
after MK-801 treatment (18), and is consistent with the
role microglia have in many aspects of nervous system
injury (54, 64). Simultaneous identification of diffuse as-
trocytic hypertrophy and multifocal microglial phagocy-
tosis in the same region of neurotoxic damage illustrates
fundamental differences in the responses made by astro-
cytes and microglia, especially considering the rather
widespread presence of both cell types throughout the
CNS. Although this difference is evident using the cur-
rent model of acute neuronal injury, additional studies
are needed to better define the character of glial responses
during different types of neurotoxic injury. Furthermore,
it is important to interpret morphologic changes in glial
cells with the knowledge that they (astrocytes in partic-
ular) can be primary targets of specific neurotoxins (6)
and may be influenced directly by the regulatory effects
of steroid hormones (43, 44).
Not only were anatomical differences identified in the
responses of astrocytes and microglia, but temporal dif-
ferences were noted as well. In general, microglia re-
sponded earlier and were less persistent than astrocytes,
which had a gradual response that remained evident to
the end of the 28-day study. These data support findings
from other studies with different models of nervous sys-
tem injury that show microglial activation preceding the
astrocytic response (30, 34, 37, 50). Collectively, these
time-course data support the hypothesis that cytokines,
particularly interleukin-1, initially are produced by mi-
croglia and subsequently activate astrocytes (26, 27).
However, it is important to consider the type of detection
signal used to identify &dquo;activation&dquo; of different glia cell
types. In contrast to increased GFAP immunoreactivity,
which requires production of new protein by hypertrophic
302
astrocytes, reactive microglia identified by either an
are
upregulation of immunologic proteins on the cell mem-
brane, an overt change in shape, or a combination of both
(59, 60, 64). The inherent differences in these approaches
should be considered when interpreting data that compare
temporal responses made by microglia and astrocytes in
an area of nervous system injury. As more information
is generated on glial cell biology, the complex and di-
verse responses these cells make will require continual
reassessment, particularly regarding their role in identi-
fying areas of neurotoxic injury.
Methods for identifying microglia in tissue section are
relatively new and include immunohistochemistry for
major histocompatibility antigens and complement recep-
tors, as well as specific membrane staining with the lectin
used in the present study (28, 40, 60). Approaches such
as these are beginning to be used for characterizing the
reactive responses microglia make to different nervous
system insults, including toxicants (41, 48, 49, 66), hyp-
oxia (36), and trauma (24, 59). What is clearly evident
from these and other data is that microglial responses
occur along a gradient that involves a range of molecular
and morphologic changes. Furthermore, many variables
are involved in the character of these changes, such as
type of insult (24), brain location (39), time after insult
(39), and involvement of the blood-brain barrier (15). As
research with microglia progresses, our understanding of
the overall response to neurotoxicants will grow and ad-
ditional opportunities for identifying neurotoxic insult
should emerge.
The gelatin-embedded composite freeze-cut sections
used for CS and GFAP staining proved advantageous for
convenient screening of multiple brains with special
stains. This sectioning approach produced orientation so
that similar neuroanatomic sites were represented across
all brain sections. The opportunity to have up to 16 brains
in one block allows simultaneous processing of different
treatment groups, thus minimizing the potential slide to
slide variability that can occur with special stains. Com-
posite sections can also improve the efficiency of obser-
vations, especially when subtle effects require precise
comparisons among treatment groups. In addition, serial
sectioning of the original gelatin block allows the inves-
tigator to determine the interval between sections pre-
pared for examination.
The thick gelatin-embedded composite sections had
advantages or disadvantages compared with thinner par-
affin sections, depending on the stain and endpoint eval-
uated. For example, composite sections stained for GFAP
clearly demonstrated a diffuse, localized band of reactive
astrocytes specific to the cortical layers involved in MK-
801-induced neuronal necrosis. These same changes were
observed concurrently in GFAP-stained thinner paraffin
sections; however, the increased section thickness in the
composite sections made astrocytic reactivity easier to
identify. In contrast, detection of individual eosinophilic
necrotic neurons by H&E staining was more difficult in
the thicker composite sections. In this case, the increased
resolution gained with paraffin sectioning proved neces-
sary for recognizing this subtle tinctorial change. More-
over, paraffin-embedded H&E sections were required for
demonstrating cytoplasmic vacuolization at 4 hr after
treatment. These datasuggest that the composite section-
ing procedure may be best used for special stains that
produce reasonable contrast between positive cells and
background, rather than for identifying subtle tinctorial
or morphologic changes.
This study provided an integrated view of the multi-
faceted changes in neurons, astrocytes, and microglia af-
ter acute neuronal injury, while demonstrating the utility
of combining multiple morphologic approaches for as-
sessing neurotoxic injury in the CNS. Linking the proven
value of routine histopathology with additional morpho-
logically based endpoints can strengthen our ability to
interpret data for hazard identification and characteriza-
tion during the risk assessment process. Studies with oth-
er neurotoxicants acting at different sites and through oth-
er mechanisms are needed to better understand the role
special staining and sectioning techniques may play in
future neurotoxicity testing. In addition, dose response
studies will better define the degree of neurotoxic injury
required induce the changes measurable by these tech-
niques.
ACKNOWLEDGMENTS
The authors thank Dave Barnett, Gerry Lawhorn, Cin-
dy Lewis, Tracy Pesut, Julie Switzer, and Dana Wheat
for their technical contributions to this manuscript. Drs.
James Maurer, Jay Nash, and Laurence Whiteley provid-
ed valuable scientific review.
REFERENCES
1. Auer RN (1994). Assessing structural changes in the brain to eval-
uate neurotoxicological effects of NMDA receptor antagonists.
Psychopharmacol. Bull. 30(4): 585-591.
2. Auer RN and Coulter KC (1993). The nature and time course of
neuronal vacuolization induced by the N-methyl-D-aspartate antag-
onist MK-801. Acta Neuropathol. 87: 1-7.
Balaban CD (1992). The use of selective silver degeneration stains
3.
in neurotoxicology—Lessons from studies of selective neurotoxi-
cants. In: The Vulnerable Brain and Environmental Risks, Vol. 1:
Malnutrition and Hazard Assessment, RL Isaacson and KF Jensen
(eds). Plenum Press, New York, pp. 223-238.
4. Balaban CD, O’Callaghan JP, and Billingsley ML (1988). Trime-
thyltin-induced neuronal damage in the rat brain: Comparative
studies using silver degeneration stains, immunocytochemistry, and
immunoassay for neurotypic and gliotypic proteins. Neuroscience
26: 337-361.
5. Beltramino CA, de Olmos JS, Gallyas F, Heimer L, and Zaborszky
L (1993). Silver staining as a tool for neurotoxic assessment. In:
Assessing Neurotoxicity of Drugs of Abuse, L Erinoff (ed). National
Institutes of Health Research Monograph Series, U.S. Department
of Health and Human Services, Public Health Service, pp. 101-
126.
6. Butterworth RF, Giguere JF, Michaud J, Lavoie J, and Pomier LG
(1987). Ammonia: Key factor in the pathogenesis of hepatic en-
cephalopathy. Neurochem. Pathol. 6: 1-12.
7. Choi DW (1992). Bench to bedside: The glutamate connection.
Science 258: 241-243.
8. Commins DL, Vosmer G, Virus R, Wollverton W, Schuster C, and
Seiden L (1987). Biochemical and histological evidence that meth-
ylenedioxymethylamphetamine (MDMA) is toxic to neurons in the
rat brain. J. Pharmacol. Exp. Ther. 241: 338-345.
9. de Olmos JS, Beltramino CA, and de Olmos de Lorenzo S (1994).

Use of an amino-cupric-silver technique for the detection of early
and semiacute neuronal degeneration caused by neurotoxicants,
hypoxia, and physical trauma. Neurotoxicol. Teratol. 16(6): 545-
561.
10. de Olmos JS, Ebbeson SO, and Heimer L (1981). Silver methods
for the impregnation of degenerating axoplasm. In: Neuroanatom-
ical Tract-Tracing Methods, L Heimer and MJ Robards (eds). Ple-
num Press, New York, pp. 117-170.
11. de Olmos JS and Ingram WR (1971). An improved cupric-silver
method for impregnation of axonal and terminal degeneration.
Brain Res. 33: 523-529.
12. Eddleston M and Mucke L (1993). Molecular profile of reactive
astrocytes—Implications for their role in neurologic disease. Neu-
roscience 54(1): 15-36.
13. Ellison G and Switzer RC (1993). Dissimilar patterns of degener-
ation in brain following four different addictive stimulants. Neu-
roreport 5 : 17-20.
14. Eng LE and Ghirnikar RS (1994). GFAP and astrogliosis. Brain
Pathol. 4: 229-237.
15. Espinosa IF, Sampedro MN, and Bovolenta P (1993). Differential
activation of microglia and astrocytes in aniso- and isomorphic
gliotic tissue. Glia 8: 277-291.
16. Fix AS (1994). Pathological effects of MK-801 in the rat posterior
cingulate/retrosplenial cortex. Psychopharmacol. Bull. 30(4): 577-
583.
17. Fix AS, Horn JW, Truex LL, Smith R, and Gomez E (1994). Neu-
ronal vacuole formation in the rat posterior cingulate/retrosplenial
cortex after treatment with the N -methyl-D-aspartate (NMDA) an-
tagonist MK-801 (dizocilpine maleate). Acta Neuropathol. 88(6):
551-519.
18. Fix AS, Horn JW, Wightman KA, Johnson CA, Long GG, Storts
RW, Farber N, Wozniak DF, and Olney JW (1993). Neuronal vac-
uolization and necrosis induced by the noncompetitive N -methyl-
D-aspartate (NMDA) antagonist MK(+)801 (dizocilpine maleate):
A lightand electron microscopic evaluation of the rat retrosplenial
cortex. Exp. Neurol. 123: 204-215.
19. Fix AS, Long GG, Wozniak DF, and Olney JW (1994). Pathomor-
phologic effects of N-methyl-D-aspartate antagonists in the rat pos-
terior cingulate/retrosplenial cerebral cortex: A review. Drug Dev.
Res. 32: 147-152.
20. Fix AS, Ross JF, Stitzel SR, Barnett DM, and Switzer RC (1995).
Histological approaches for detecting central nervous system neu-
rotoxicity : Comparison of routine paraffin procedures with a novel
embedding and freeze-sectioning technique for multiple brains.
Toxicol. Pathol. 23(6): 763-764 (abstract).
21. Fix AS, Ross JF, Stitzel SR, Barnett DM, and Switzer RC (1995).
MK-801-induced neurotoxicity: Evaluation by cupric silver stain-
ing for neuronal degeneration and lectin histochemistry for mi-
croglia. Vet. Pathol. 32: 576 (abstract).
22. Fix AS, Wightman KA, and O’Callaghan JP (1995). Reactive gli-
osis induced by MK-801 in the rat posterior cingulate/retrosplenial
cortex: GFAP evaluation by sandwich ELISA and immunocyto-
chemistry. Neurotoxicology 16(2): 229-238.
23. Fix AS, Wozniak DF, Truex LL, McEwen M, Miller JP, and Olney
JW (1995). Quantitative analysis of factors influencing neuronal
necrosis induced by MK-801 in the rat posterior cingulate/retro-
splenial cortex. Brain Res. 696: 194-204.
24. Flaris NA, Densmore TL, Molleston MC, and Hickey WF (1993).
Characterization of microglia and macrophages in the central ner-
vous system of rats: Definition of the differential expression of
molecules using standard and novel monoclonal antibodies in nor-
mal CNS and in four models of parenchymal reaction. Glia 7: 34-
40.
25. Food and Drug Administration (1982). Toxicologic Principles for
the Safety Assessment of Direct Food Additives and Color Additives
Used in Food. FDA’s Center for Food Safety and Applied Nutrition,
Division of Health Effects, Washington, D.C.
26. Giulian D, Li J, Leara B, and Keenen C (1994). Phagocytic mi-
croglia release cytokines and cytotoxins that regulate the survival
303
of astrocytes and neurons in culture. Neurochem. Int. 25(3): 227-
233.
27. Giulian D, Woodward J, Young DG, Krebs JE and Lachman LB
(1988). Interleukin-1 injected into mammalian brain stimulates as-
trogliosis and neovascularization. J. Neurosci. 8(7): 2485-2490.
28. Graeber MB, Streit WJ, and Kreutzberg GW (1990). The third glial
cell type, the microglia: Cellular markers of activation in situ. Acta
Histochem. 38(suppl.): 157-160.
29. Isaacson RL and Jensen KF (eds) (1992). The Vulnerable Brain
and Environmental Risks. Plenum Press, New York. 266 pp.
30. Jensen MB, Gonzalez B, Castellano B, and Zimmer J (1994). Mi-
croglial and astroglia reactions to anterograde axonal degeneration:
A histochemical and immunocytochemical study of the adult rat
fascia dentata after entorhinal perforant path lesions. Exp. Brain
Res. 98: 245-260.
31. Johannessen JN (1993). Markers of neuronal injury and degenera-
tion. The New York Academy of Sciences, New York.
32. Langston JW and Young A (eds) (1992). Neurotoxins and Neuro-
degenerative Disease. Annals of the New York Academy of Sci-
ences, New York. 385 pp.
33. Llorens J, Crofton KM, and O’Callaghan JP (1993). Administration
of 3-3’-iminodipropionitrile to the rat results in region-dependent
damage to the central nervous system at levels above the brain
stem. J. Pharmacol. Exp. Ther. 265(3): 1492-1498.
34. McCann MJ, O’Callaghan JP, Martin PM, Bertram TA, and Streit
WJ (1996). Differential activation of microglia and astrocytes fol-
lowing trimethyltin-induced brain injury. Neuroscience (in press).
35. McMulloch J and Iversen LL (1991). Autoradiographic assessment
of the effects of N -methyl-D-aspartate receptor antagonists in vivo.
Neurochem. Res. 16: 951-963.
36. McRae A, Gilland E, Bona E, and Hagberg H (1995). Microglial
activation after neonatal hypoxic-ischemia. Dev. Brain Res. 84:
245-252.
37. Mitchell J, Sundstrom LE, and Wheal HV (1993). Microglial and
astrocytic cell responses in the rat hippocampus after an intrace-
rebroventricular kainic acid injection. Exp. Neurol. 121: 224-230.
38. Montgomery DL (1994). Astrocytes: Form, functions, and roles in
disease. Vet. Pathol. 31: 145-167.
39. Morioka T, Kalehua AN, and Streit WJ (1993). Characterization of
microglial reaction after middle cerebral artery occlusion in rat
brain. J. Comp. Neurol. 327: 123-132.
40. Nakalima K and Kohsaka S (1993). Functional roles of microglia
in the brain. Neurosci. Res. 17: 187-207.
41. Nakki R, Koistanaho J, Sharp FR, and Sagar SM (1995). Cerebellar
toxicity of phencyclidine. J. Neurosci. 15(3): 2097-2108.
42. O’Callaghan JP (1993). Quantitative features of reactive gliosis fol-
lowing toxicant-induced damage in the CNS. In: Markers of Neu-
ronal Injury and Degeneration, J Johannessen New York Academy
of Sciences, New York, pp. 195-210.
43. O’Callaghan JP, Brinton RE, and McEwen BS (1989). Glucocorti-
coids regulate the concentration of glial fibrillary acidic protein
throughout the brain. Brain Res. 494: 159-161.
44. O’Callaghan JP, Brinton RE, and McEwen BS (1991). Glucocorti-
coids regulate the synthesis of glial fibrillary acidic protein in intact
and adrenalectomized rats but do not affect its expression following
brain injury. J. Neurochem. 57(3): 860-869.
45. O’Callaghan JP and Jensen KF (1992). Enhanced expression of
glial fibrillary acidic protein and the cupric silver degeneration re-
action can be used as sensitive and early indicators of neurotoxicity.
Neurotoxicology 13: 133-122.
46. O’Callaghan JP, Jensen KF, and Miller DB (1995). Quantitative
aspects of drug and toxicant-induced astrogliosis. Neurochem. Int.
26(2): 115-124.
47. O’Callaghan JP and Miller DB (1993). Quantification of reactive
gliosis as an approach to neurotoxicity assessment. In: Assessing
Neurotoxicity of Drugs of Abuse, L Erinoff (ed). U.S. Government
Printing Office, Washington, D.C., pp. 188-212.
48. Ogawa M, Araki M, Niijima K, and Yoshida M (1992). Microglial
reactions in the inferior olive and its tract induced by intraperito-
304
neal injection of 3-acetylpyridine in the rat. Neurodegeneration 1 :
37-43.
49. O’Hearn E, Long DB, and Molliver ME (1993). Ibogaine induces
glial activation in parasagittal zones of the cerebellum. Neuroreport
4: 299-302.
50. Ohno M, Aotani H, and Shimada M (1995). Glial responses to
hypoxic/ischemic encephalopathy in neonatal rat cerebrum. Dev.
Brain Res. 84: 294-298.
51. Olney JW, Labruyere J, and Price MT (1989). Pathological changes
induced in cerebrocortical neurons by phencyclidine and related
drugs. Science 244: 1360-1362.
52. Olney JW, Price MT, Salles KS, Labruyere J, and Frierdich G
(1987). MK-801 powerfully protects against N -methyl aspartate
neurotoxicity. Eur. J. Pharmacol. 141: 357-361.
53. Paxinos G and Watson C (1986). The Rat Brain in Stereotaxic
Coordinates, 2nd ed. Academic Press, San Diego.
54. Perry VH and Gordon S (1991). Macrophages and the nervous
system. Int. Rev. Cytol. 125: 203-244.
55. Ritter S and Dinh T (1988). Capsaicin-induced neuronal degener-
ation : Silver impregnation of cell bodies, axons, and terminals in
the central nervous system of the adult rat. J. Comp. Neurol. 271:
79-90.
56. Sette WF (1991). Pesticide Assessment Guidelines, Subdivision F;
Hazard Evaluation: Human and Domestic Animals, Addendum 10,
Neurotoxicity, Series 81, 82, and 83.
57. Sharp FR, Jasper P, Hall J, Noble L, and Sagar SM (1991). MK-
801 and ketamine induce heat shock protein HSP72 in injured neu-
rons in posterior cingulate and retrosplenial cortex. Ann. Neurol.
30: 801-809.
58. Streit WJ (1990). An improved staining method for rat microglial
cells using the lectin from Griffonia simplicifolia (GSA I-B4). J.
Histochem. Cytochem. 38: 1683-1686.
59. Streit WJ, Graeber MB, and Kreutzberg GW (1989). Peripheral
nerve lesion produces increased levels of major histocompatibility
complex antigens in the central nervous system. J. Neuroimmunol.
21: 117-123.
60. Streit WJ and Kreutzberg GW (1987). Lectin binding by resting
and reactive microglia. J. Neurocytol. 16: 249-260.
61. Streit WJ and Kreutzberg GW (1988). Response of endogenous
glial cells to motor neuron degeneration induced by toxic ricin. J.
Comp. Neurol. 268: 248-263.
62. Switzer RC (1991). Strategies for assessing neurotoxicity. Neurosci.
Biobehav. Rev. 15: 89-93.
63. Switzer RC, Bogo V, and Mickley GA (1994). Histologic effects
of high energy electron and proton irradiation of rat brain detected
with a silver-degeneration stain. Adv. Space Res. 14(10): 443-451.
64. Thomas WE (1992). Brain macrophages: Evaluation of microglia
and their functions. Brain Res. Rev. 17: 61-74.
65. U.S. Department of Health and Human Services (1995). Guide for
the Care and Use of Laboratory Animals. Animal Resources Pro-
gram, Division of Research Resources, National Institutes of
Health, Bethesda, MD.
66. Wilson MA, Molliver M (1994). Microglial response to degenera-
tion of serotonergic axon terminals. Glia 11: 18-34.