Journal of Neuropathology and Experimental Neurology Vol. 59, No.

8
Copyright q 2000 by the American Association of Neuropathologists August, 2000
pp. 641–651

Recent Advances in Neurotrauma

D. I. GRAHAM, MBBCH, T. K. MCINTOSH, PHD, W. L. MAXWELL, PHD, AND J. A. R. NICOLL, MD

Abstract. The frequency of and outcome from acute traumatic brain injury (TBI) in humans are detailed together with a
classification of the principal focal and diffuse pathologies, and their mechanisms in extract laboratory models are outlined.
Particular emphasis is given to diffuse axonal injury, which is a major determinant of outcome. Cellular and molecular cascades
triggered by injury are described with reference to the induction of axolemmal and cytoskeletal abnormalities, necrotic and
apoptotic cell death, the role of Ca21, cytokines and free radicals, and damage to DNA. It is concluded that TBI in humans
is heterogeneous, reflecting various pathologies in differing proportions in patients whose genetic background (APOE gene
polymorphisms) contributes to the outcome at 6 months. Although considerable progress has been made in the understanding
of TBI, much remains to be determined. However, a deeper understanding of the pathophysiological events may lead to the

Downloaded from [Link] by guest on 07 March 2024

possibility of improving outcome from rational targeted therapy.

Key Words: ApoE genotype; Apoptosis; Outcome; Traumatic brain injury.

INTRODUCTION diffuse damage. Experience has shown that whereas the

The pathology of head injury is complex, the outcome nature of a focal lesion can be determined in life with a
being a product of different mechanisms, types, and high degree of certainty, it is more difficult in comatose
amounts of brain injury and their anatomical location. cases with diffuse brain damage who do not demonstrate
Epidemiological studies have shown that deaths from an intracranial mass lesion.
head injury comprise 1%–2% of all deaths from all caus-
es, and that between one third to one half of all traumatic Mechanisms of Brain Damage
deaths are due to head injury. The precise incidence of These include static loading and the various types of
head injury is difficult to establish. In the UK about 400 dynamic loading (2). Static loading occurs when forces
patients per 100,000 of the population are admitted to the are applied to the head gradually and a slow time course
hospital each year with head injury, of which 30 per (usually taking more than 200 milliseconds [ms] to de-
100,000 are adults not obeying commands. In a recent
velop). Sufficient force may cause multiple, comminuted,
survey of over 1,000 patients admitted to hospital after
or eggshell fractures of the skull, and may not be asso-
head injury, the European Brain Injury Consortium found
ciated with either coma or severe neurological signs until
31% were dead in 6 months, 30% had made a good re-
the amount of skull deformation is so severe that the
covery, 20% were moderately disabled, 16% were se-
verely disabled, and 3% were vegetative (1). brain itself becomes compressed and distorted. The most
common mechanical input to the head is dynamic loading
Classification of Brain Damage of either impact or impulsive type (act in less than 200
There are 2 main categories of damage after a head ms, and, in most instances, in less than 20 ms). Impact
injury: 1) primary, that occurs or is triggered at the mo- loading occurs when a blunt object strikes the head and
ment of injury (lacerations of the scalp, fracture of the usually initiates a combination of contact and inertial
skull, surface contusions and lacerations of the brain, dif- forces that result in a series of events that vary with the
fuse axonal injury and intracranial hemorrhage); and 2) size of the impacting object and the magnitude of the
secondary, produced by complicating processes that are force delivered to the contact point. In addition, shock
initiated at the moment of injury but do not present clin- waves are propagated throughout the brain. Impulsive dy-
ically for a period of time after injury (damage due to namic loading occurs when the head is set into motion
raised intracranial pressure, ischemia, swelling and infec- or when the moving head is stopped without it striking
tion) (2). Neuroimaging as a means of identifying intra- anything or is arrested by impact. Under these conditions
cranial pathology after head injury has allowed the clin- the resulting head injuries are caused solely by the inertia
ical adoption of an alternative classification of focal and
produced by the manner (translation, rotation, angulation)
in which the head is moved. Strain is the proximate cause
From the University Department of Neuropathology (DIG, JARN),
Institute of Neurological Sciences, Southern General Hospital NHS of tissue injury whether induced by contact or inertia.
Trust, Glasgow; Department of Neurosurgery (TKM), University of Characteristically, biological tissues will stand the strain
Pennsylvania School of Medicine, Philadelphia, Pennsylvania; IBLS Di- better if they are deformed slowly rather than quickly,
vision of Neuroscience and Biomedical Systems (WLM), University of i.e. they become more brittle and will break at lower
Glasgow, Glasgow, United Kingdom.
Correspondence to: Professor D.I. Graham, University Department of
strain levels under rapidly applied loads.
Neuropathology, Institute of Neurological Sciences, Southern General There is now a wealth of information that lesions due to
Hospital, 1345 Govan Road, Glasgow, G51 4TF, United Kingdom. contact include lacerations of the scalp, fracture of the skull

641
642 GRAHAM ET AL

with or without an associated extradural (epidural) hema- To date, experimental models have contributed greatly
toma, surface contusions and lacerations, and intracerebral to our insight of post-traumatic mechanisms of cell death
hematoma. In contrast, inertial forces are responsible for the and behavioral disability. Additionally, they have been
2 most important types of damage encountered in blunt integral in the development of several novel diagnostic
head injury, namely, acute subdural hematoma resulting and/or treatment strategies that have either become part
from tearing of subdural bridging veins and widespread of standard clinical practice or are now under preclinical
damage to white matter in the form of diffuse axonal injury. or clinical investigation.
Analysis of various data sets (2) has shown that focal dam- All models designed to produce experimental TBI
age associated with contact is more likely to be sustained should fulfill several criteria: these include the production
as the result of a fall, while diffuse damage is most com- of a quantifiable and reproducible injury, and the em-
monly associated with acceleration/deceleration occurring ployment of standardized surgical protocols and tech-
after, for example, traffic accidents. niques. The latter should include the use of ‘‘sham’’ (un-
Brain damage after head injury in humans is hetero- injured) animals to control for surgical/systemic variables

Downloaded from [Link] by guest on 07 March 2024

geneous. For example, in a human clinical series, solitary (e.g. the operative procedure, anesthesia, or alterations in
lesions were found in 26% of cases while 2–3 or more core or brain temperature), and any brain damage due to
than 3 lesions occurred in 21% and 27% of cases, re- head restraint. The majority of devices inducing trauma
spectively. In a comprehensive neuropathological study currently use computer-based measurements of the ap-
of 50 cases from the Institute of Neurological Sciences, plied load (e.g. pressure gradients, velocity of impactor,
Glasgow, focal lesions were found in 47 cases (95%) and amount of acceleration/deceleration) to assess for varia-
diffuse lesions in 48 (94%) of cases. Overall there were tion of the mechanical parameters that define injury se-
only 4 cases (8%) in which there was a single type of verity. This information is critical in making adjustments
lesion, 6% in which there were 2, and 14% in which there to the injury device, permitting the maintenance of a nar-
were 3 pathologies. But in over 72% of the cases there row range of injury severity within a given study, and
were 4 or more types of lesion, providing further evi- the generation of a reproducible injury between experi-
dence that as injury severity worsens so the multiplicity mental animals.
of lesions increases. The fluid percussion model of experimental brain in-
jury is the most commonly used and well-characterized
Experimental Models of Human Blunt Traumatic Brain model of clinical traumatic brain injury. This model,
Injury-Matching Models to Humans which was originally developed for the rabbit and cat,
has been modified over the past decade to produce injury
The principal models of experimental head injury and in a wide range of differing animal species including
the major pathological findings in them have been com- mouse, rat, dog, and pig (2, 3). Fluid percussion repro-
prehensively reviewed (3) and the specific attributes of duces contusions as focal grey matter damage with ac-
each model in terms of its ability to reproduce abnor- companying intraparenchymal or subarachnoid hemor-
malities known to occur in human head injury have been rhage. Central or midline fluid percussion in rodents leads
detailed. Replication of diffuse brain injury has been to contusion directly at the site of fluid impact and com-
achieved by angular acceleration in nonhuman primates pression of brainstem structures with resulting behavioral
(4), in the mini-pig (5), and by contact percussion in rats disturbances resembling coma when more severe levels
(6). Typically there is widespread microscopic evidence of injury are induced. The lateral fluid percussion model
of damage to axons (7). Contusion is easy to replicate in the rat produces a lesion cavity that is associated with
and can be induced by suction and local tensile models both necrotic and apoptotic cell death, evolves from the
without the complications of more distant or global contusion over a period of days to weeks, and then pro-
changes: there is a hemorrhagic focus of necrosis, dis- gressively expands for up to one year post-injury (9, 10).
ruption of the blood-brain barrier (BBB), the formation Progressive changes in the hippocampi and thalami have
of vasogenic edema, decreases in regional cerebral blood also been reported in this model, features that have been
flow, and an increase in glucose metabolism. Histologi- accompanied by persistent cognitive (11) and motor im-
cally, neutrophil polymorphs increase in number by 24 h pairment up to one-year post-injury (12). Although the
followed by activation of microglia (8) and the formation lateral fluid percussion model was developed as a model
of macrophages; gliosis develops and a glial limitans is of unilateral damage, a number of studies have shown
formed to separate injured from intact neuropil. A more that the damage is widely distributed.
difficult entity to model is that of acute subdural hema- The plethora of models designed to study brain injury
toma, although this can be achieved by injection and ac- further emphasizes the complexity and heterogeneity of
celeration (3); however, the associated features of skull TBI in humans. In particular, it should be remembered
fracture, brain swelling, and extradural hematoma are not that with few exceptions human head injury is rarely
easily duplicated routinely. pure, in contrast to many of the experimental models that

J Neuropathol Exp Neurol, Vol 59, August, 2000

 HEAD INJURY 643

serve to isolate different aspects of traumatic brain injury. also allowed demonstration of injured axons in humans
Thus, using the purer and extract experimental models, it up to 99 days after mild head injury (17).
may not be possible to adequately address the multiplicity
Concept of Traumatic Axonal Injury
of focal and diffuse pathologies seen in many patients
with traumatic brain injury. As originally described, DAI was defined as a focal
lesion in the corpus callosum, focal lesions in one or both
THE IMPORTANCE OF DIFFUSE BRAIN DAMAGE dorsolateral sectors of the rostral brainstem adjacent to
INCLUDING DIFFUSE AXONAL INJURY the cerebellar peduncles, and diffuse damage to axons
The descriptor ‘‘diffuse brain injury’’ refers to the phe- (2). Such a distribution of injured axons has been
nomenon of widespread damage to either grey or white achieved in only one animal model, the nonhuman pri-
matter (or both) rather than to a localized area of injury— mate (4).
focal injury—where the remainder of the brain is, appar- Recently, the descriptor traumatic axonal injury (TAI)
ently, unaffected. The pathological manifestation of dif- has been applied to experimental studies in animals that

Downloaded from [Link] by guest on 07 March 2024

fuse brain injury will differ with the type of insult that have attempted to elucidate the mechanisms of axonal
the brain has received. The principal types of diffuse pathology after trauma. In TAI, injured axons form focal
brain damage are brain swelling, hypoxic damage, and swellings at intervals along their length within the first 2
diffuse axonal injury (DAI), of which the latter is partic- h after injury. These swellings increase in size until the
ularly important. axon undergoes disconnection between 4 and 6 h after
Various categories of diffuse white matter injury are now injury; a process now termed secondary axotomy (7).
recognized as having a similar pathophysiology, but there This time course parallels that found in DAI but the cur-
rent level of understanding of the events in injured axons
are major differences in the clinical outcome of such injury,
leading to secondary axotomy has been derived from
varying with the severity of the mechanical strain placed
studies of experimental traumatic axonal injury.
upon axons. At the least severe level of injury are the con-
Strich was the first to define ‘‘diffuse degeneration of
cussion syndromes in which there is a short-term loss of
the cerebral white matter’’ where nerve fibers were
neurological function but not of consciousness. A greater
sheared or mechanically torn apart at the time of injury
severity of injury results in ‘‘classical cerebral concussion’’
(18). Good evidence in support of this concept has been
in which there is loss of consciousness for a period of usu-
obtained in only one animal model (7). The concept has
ally less than 6 h. At the most severe level there is pro-
recently arisen that axons are not sheared at the time of
longed traumatic coma that is not associated with an intra-
injury, except perhaps in the most severe form of DAI,
cranial mass lesion but in which there is diffuse axonal
but rather they are damaged and enter a ‘‘pathological
injury (2, 13). Patients with severe DAI are unconscious cascade’’ of events leading to axotomy at least several
from the moment of injury, do not experience a lucid in- hours after the initial injury. Where nerve fibers are
terval, and remain unconscious, vegetative (14), or at least sheared at the time of injury (primary axotomy), the axo-
severely disabled until they die, usually as a result of sec- lemma is fragmented and there is rapid loss of the axonal
ondary, post-traumatic complications. cytoskeleton such that neurofilaments and microtubules
In the head-injured patient, a definitive diagnosis of are replaced by a flocculent ultrastructure suggestive of
DAI can only be made postmortem since the principal autolysis (7). However, where the axolemma has been
criterion for its diagnosis is the presence of injured axons. fragmented there is not a loss of membrane specializa-
Techniques for the demonstration of injured axons in tions such as the dense undercoating characteristic of the
brain have undergone considerable development in recent node of Ranvier.
years. Initially, silver impregnation techniques for the la- In the last several years a consensus has arisen that the
beling of axonal ‘‘retraction’’ bulbs were used. More re- prime site of injury is the axolemma and that pathology
cently, a number of immunocytochemical (ICC) methods of the injured axon occurs because of loss of homeostatic
have been introduced with the general acceptance that mechanisms in the maintenance of the differential ionic
greatest sensitivity of labeling, both for postmortem di- gradients necessary for the electrical activity of the axon.
agnosis and in the experimental situation of damaged ax- Evidence for damage to the axolemma has been obtained
ons, may be obtained using ICC labeling for b-amyloid through use of several experimental techniques. First, the
precursor protein (b-APP) (15, 16). This technique is now electron dense tracer horseradish peroxidase (HRP) is
widely used to identify injured axons scattered among a normally excluded from the axoplasm when the tracer is
population of uninjured ones, and has provided the added injected into the CSF. However, upon TAI, peroxidase
benefit that damaged axons may be demonstrated within floods the axoplasm of injured nerve fibers. The precise
2–3 h of injury (15, 16). Further, the technique has al- mechanism of entry for HRP is unresolved but it has been
lowed discrimination of changes in axonal form prior to suggested that the axolemma becomes leaky or is per-
axotomy in that focal ‘‘axonal swellings’’ (7) precede the turbed (19, 20). Second, quantitative analysis of freeze-
occurrence of ‘‘retraction bulbs.’’ Labeling for b-APP has fracture replicas (21) and cytochemical localization of

J Neuropathol Exp Neurol, Vol 59, August, 2000

644 GRAHAM ET AL

certain axolemma membrane pumps has demonstrated the brain to both impact and inertial forces. With direct
changes in the structure of the axolemma in injured nerve impact to the skull, a local bending occurs with under-
fibers and alteration in the localization of Ca21 -ATPase lying tissue strain and gross movement of brain. Con-
and p-NPPase activity after TAI (7, 22). These studies versely, inertial injury does not create local contact ef-
provide strong support for the hypothesis that after TAI fects but produces a nonuniform distribution of pressure
there is an uncontrolled influx of Ca21 into injured nerve and tissue strain that cause primary tissue damage (2).
fibers, a morpho-pathological manifestation of which is Intracranial pressure changes and brain motion due to
swelling of axonal mitochondria (22). Elevated intra-ax- translational acceleration have been linked to specific fo-
onal levels of Ca21 are currently hypothesized to result cal lesions such as coup and contrecoup contusions, in-
in depolymerization of microtubules, as reflected in the tracerebral and/or subdural hematomas, and brainstem le-
rapid acute reduction in their number (7), activation of sions. Other lesions, such as DAI and gliding contusions,
calpains that lead to alterations in axonal inter-neurofil- are more related to rotational acceleration forces (2).
ament relationships, and changes in axonal transport that

Downloaded from [Link] by guest on 07 March 2024

Secondary Events
allow the development of axonal swellings and potentiate
secondary axotomy. Recent ICC demonstrating spectrin Superimposed on trauma-induced mechanical injury,
breakdown products in injured axons has served also to ‘‘secondary’’ or delayed neuronal or cellular damage de-
greatly strengthen this hypothesis (23). velops over a period of hours, days, or weeks after the
Early in the 1990s it was hypothesized that the appli- initial trauma. This type of damage appears to be asso-
cation of transient tensile strain to axons directly disrupts ciated with trauma-induced neurochemical alterations,
the relationships between components of the axonal cy- which can exert either direct pathogenic effects on re-
toskeleton after traumatic brain injury. This has now been gional cerebral blood flow, BBB function, cerebral me-
superseded by the view that injury to the axolemma is tabolism, and ion homeostasis, or have direct neurotoxic
the principal cause of the ensuing axonal pathology. effects on regional populations of neurons or glial cells
Where swollen axons are labeled for b-APP, it is hypoth- (26). These post-traumatic cellular and molecular changes
esized that there is a focal disruption of axonal transport. may involve alterations in the synthesis and release of
Support for this hypothesis has been provided by quan- both neuroprotective and autodestructive or neurotoxic
titative analyses of changes in the axonal cytoskeleton cascades (27). Identification of these pathways will pro-
after TAI, after which there is a dramatic loss of the num- vide important information about mechanisms of damage
ber of axonal microtubules (25) and a reduced spacing, and stimulate the development of novel therapeutic strat-
termed compaction, between neurofilaments within in- egies designed to prevent or attenuate trauma-induced
jured axons (19, 24). Loss of microtubules has been hy- damage to the CNS. In this review, we have chosen to
pothesized to result from their depolymerization when detail selected secondary injury pathways that have been
intra-axonal Ca21 levels are elevated post-trauma (7). currently and most convincingly related to the pathobi-
Loss of microtubules will clearly result in loss of fast ology of neuronal death associated with traumatic brain
axonal transport and allow the accumulation of both b- injury.
APP and membranous organelles at foci that then form Is ‘‘Cytoskeletal Collapse’’ a Major Feature of the
axonal swellings. The significance of compaction of neu- Response to Traumatic Brain Injury?
rofilaments is presently unclear, except in that slow ax-
onal transport will, probably, be compromised. However, Alterations in brain Ca21 homeostasis (13, 22, 28) and
compaction of neurofilaments has been noted within min- receptors/channels associated with Ca21 entry (voltage
utes of TAI, and neurofilaments remain in a compacted sensitive channels or ionophore-associated glutamate re-
state for at least several hours prior to the axon under- ceptors such as N-methyl-D-aspartate-NMDA receptors)
going secondary axotomy. At the point of axotomy there have been associated with regional cerebral edema, va-
is loss of integrity of the axolemma and dissolution of sospasm, and delayed cell death. Traumatic, ischemic, or
the compacted neurofilaments (24). It is also clear now anoxic injury to neurons is associated with widespread
that degeneration of the axonal cytoskeleton is a result neuronal depolarization (including the induction of cor-
of damage to the axolemma rather than a direct effect of tical spreading depression) and release of excitatory ami-
mechanical trauma. no acid neurotransmitters such as glutamate, leading to
the opening of NMDA receptor-associated ion channels
CURRENT CONCEPTS OF THE PATHOPHYSIOLOGY and influx of Ca21. The post-traumatic Ca21storm has
OF TRAUMATIC BRAIN INJURY been documented using 45Ca-autoradiography (29) and
indirectly via cytochemical evidence for redistribution of
Initial Events
membrane pump calcium-ATPase and ecto- Ca21-ATPase
The biomechanics of ‘‘primary’’ or mechanical dam- activity, and Ca21 influx in myelinated nerve fibers of the
age are linked to the response of bone, blood vessels, and guinea-pig optic nerve after stretch (22) and the analysis

J Neuropathol Exp Neurol, Vol 59, August, 2000

 HEAD INJURY 645

of calcium-mediated gene expression (28). Both Ca21- transgenic mouse tumor necrosis factor (TNF/Lympho-
channel blockers and competitive and noncompetitive toxin-d (LT-a) and interleukin 6 (IL-6) and in wild-type
NMDA receptor antagonists have been shown to be ef- litter mates subjected to experimental closed TBI has
ficacious in the treatment of experimental TBI (27), but shown that in spite of an increased post-traumatic mor-
to date have been disappointing in human studies (30). tality in TNF/LT-a-deficient mice, there is a neuroprotec-
Pathological elevations in intracellular Ca21 after TBI tive effect of these cytokines (40). The results suggest
can precipitate an attack on the lipid bilayer cell mem- that these cytokines may play a role in facilitating long-
brane via the activation of calcium-dependent phospho- term behavioral recovery, thereby underscoring the po-
lipases and generation of reactive oxygen species (26, tential for post-injury cellular and molecular changes to
27). These highly reactive molecules cause direct per- be either pathological or protective, depending on when
oxidative destruction of the cell membrane, oxidize cel- they are expressed during the post-injury cascade.
lular proteins and nucleic acids, and destroy the cerebral
vasculature. Calcium can also activate nonlysosomal cys- Is There Neurodegeneration after Traumatic Brain

Downloaded from [Link] by guest on 07 March 2024

teine protease calpain, which can degrade a wide range Injury?
of cytoskeletal protein substrates, including spectrin (23),
tubulin, microtubule-associated proteins MAP-1B, MAP- It is well established that TBI is a major epidemiolog-
2, and the neurofilament protein family. The activation ical risk factor for Alzheimer disease (AD) (see below)
of calpain can be determined by direct detection of au- and a pathological hallmark of this disease (diffuse b-
tolyzed calpain or indirectly via detection of calpain-spe- amyloid deposits) has been reported to occur between 4
cific proteolytic fragments. Several recent studies have h and 2.5 yr in human head-injured patients (41). At-
documented both acute calpain activation and regional tempts to reproduce these clinical observations using ex-
calpain-induced cytoskeletal proteolysis (23, 31, 32). perimental models of TBI in rodents have largely failed
Both the neuronal and axonal cytoskeleton appears to be with exception of a diffuse increased expression of b-
vulnerable to calpain-induced proteolysis (7, 33) in lab- APP, although neurofilament inclusions in a pig model of
oratory models and in human (34), and therapeutic strat- diffuse TBI have been reported (42). Recent studies em-
egies to block or antagonize the proteolytic effects of ploying transgenic mice engineered to over-express hu-
calpain on the cytoarchitecture of the cell have proven man APP 2-fold (APP-YAC mice) failed to produce am-
effective (35) by attenuating both post-traumatic motor yloid plaques or altered behavioral outcome (43). A
and cognitive deficits (31). subsequent study with transgenic mice over-expressing
human APP 10-fold (PD-APP mice) reported evidence of
Is Inflammation a Major Feature of the Response to
exacerbated hippocampal cell loss, significantly increased
TBI?
regional concentrations of Ab-PP 1–42, and worsened
There remains little doubt that a role exists for post- cognitive function after traumatic brain injury (44). TBI
injury inflammation in mediating delayed neuronal dam- in young PD-APP mice has also been shown to induce
age. Alterations in blood-borne immunocompetent cells marked ipsilateral hippocampal atrophy with diminished
have been described after trauma, and since the BBB is Ab-PP deposition during aging (45), suggesting that the
opened during the acute post-traumatic period, entry into
vulnerability of brain to Ab-PP toxicity increases while
the brain of these cells may directly influence neuronal
the accumulation of Ab-PP deposits decreases over a pe-
death and/or survival. Infiltration and accumulation of
riod of months after traumatic brain injury.
polymorphonuclear leukocytes (PMNs) into brain paren-
The brain damage associated with dementia pugilistica
chyma has been documented in experimental TBI in rats
(see below) and its associated memory loss, Parkinso-
(26) and, in the first 3 days after human TBI (2, 36). The
entry of PMNs into injured brain and activation of mi- nian-like tremors and gait, underscores the potential re-
croglia/macrophages is believed to be both pathogenic in lationship between TBI and other neurodegenerative dis-
mediating the local inflammatory response, and a partic- eases such as Parkinson disease. Because increasing
ipant in reparative and/or regenerative processes. How- evidence suggests that neurofilament-rich inclusions have
ever, the role of the marked increase in regional concen- deleterious effects on neuronal function and survival,
trations of the cytokines interleukin-1 (IL-1B), transgenic mice expressing excessive levels of heavy-
interleukin-6 (IL-6), and tumor necrosis factor (TNF) ob- chain neurofilament protein fused to a beta-galactosidase
served after experimental TBI remains uncertain. Recent reporter gene (NFH-LacZ mice) were subjected to ex-
studies documenting the beneficial effects of pharmaco- perimental TBI and found to be more behaviorally and
logical blockade of the complement cascade and the cy- histologically vulnerable to TBI than wild type mice (46).
tokines IL-1B (37, 38) and TNF (39) suggest that the These data suggest that the presence of NF-rich inclu-
release and/or upregulation of these pathways may be sions may exacerbate neuromotor dysfunction and cell
indeed pathogenic. However, more recent work with the death after traumatic brain injury.

J Neuropathol Exp Neurol, Vol 59, August, 2000

646 GRAHAM ET AL

Does the Response to TBI Recapitulate Developmental experimental traumatic brain injury (55; Barey, personal
Changes? communication). Caspase-3 itself has been also shown to
be activated in injured cortex in the acute period after
While necrotic cell death has been extensively docu-
experimental (49, 56) and human (53) brain injury, and
mented after both clinical and experimental TBI, it has
a reduction in post-traumatic apoptosis and neurological
been suggested there is also an induction of neuro-de-
velopmental cascades including programmed cell death deficits have been reported after central administration of
(PCD). Unlike necrosis, PCD involves the initiation and the caspase inhibitor z-DEVD-fmk after lateral fluid per-
active expression of transcription and translation-depen- cussion brain injury in rats (49).
dent pathways in which apoptosis is regarded as the pri- In the CNS, neurotrophic molecules have a profound
mary hallmark. Apoptosis has been classically associated influence on the development and maintenance of neu-
with the formation of the normal CNS during develop- ronal innervation, differentiation, and process outgrowth.
ment and, although necrosis and apoptosis have been For example, nerve growth factor (NGF) plays a major
role during normal development in supporting neurons

Downloaded from [Link] by guest on 07 March 2024

more traditionally considered as distinct mechanisms, it
may be possible to consider them to be part of the same that have made appropriate connections, while basic fi-
continuum of cell death, particularly within the context broblast growth factor (bFGF) has trophic effects in CNS
of traumatic CNS injury. Using a combination of terminal development and promotes neurite outgrowth and astro-
deoxynucleotidyl transferase (TdT)-mediated biotinylated cyte/oligodendroglial proliferation. Several recent studies
deoxyuridine triphosphate (dUTP) nick end labeling (TU- suggest that tissue content of growth factors is altered
NEL) histochemistry with electron microscopy and DNA after TBI: both NGF gene and protein expression have
gel electrophoresis, it has been possible to identify apo- been shown to markedly increase in the acute post-trau-
ptotic cells after experimental lateral fluid percussion matic period in rats (57) and after mechanical injury to
brain injury in the rat (47). These observations have been organotypic hippocampal slices in vitro (58). It has been
extended by demonstrating regional and temporal differ- hypothesized that the upregulation of NGF after TBI
ences in apoptotic cell death cascades, in astrocytes, ol- serves as a mediator of oxidative homeostasis by induc-
igodendrocytes, and neurons some months after experi- ing the production of free radical scavengers such as glu-
mental traumatic brain injury in a number of TBI models tathione (59). Although the bFGF response has not been
(48–50), and by the presence of apoptotic, TUNEL-pos- evaluated in models of TBI, concentrations of both NGF
itive neurons and oligodendrocytes in the tissue of sur- and bFGF in the CSF of head-injured patients have been
face contusions of human head-injured patients (51). reported to be increased (60). Hippocampal expression of
The Bcl-2 superfamily of cell death regulatory genes, brain-derived neurotrophic factor (BDNF) and RNA has
originally associated with developmental cell death, has been shown to increase bilaterally by 3 h after lateral
been implicated as a mediator of post-traumatic apopto- fluid percussion brain injury (61).
sis. Increased expression of the anti-apoptotic protein In addition to the injury-induced response of trophic
Bcl-2 has been observed in surviving neurons after both factors, convergent evidence suggests that an upregula-
experimental and clinical traumatic brain injury (52, 53), tion of several growth-related proteins such as growth-
while Raghupathi et al (personal communication) have associated protein 43, MAP1B, and polysialylated neural-
found an acute downregulation of bcl-2 in hippocampal cell adhesion molecule (PSA-NCAM) occurs after brain
and cortical neurons destined to die after experimental injury (62, 63), providing further support for a ‘‘recapit-
CNS injury. Transgenic mice over-expressing human Bcl- ulation’’ of early CNS developmental events. It is pos-
2 protein exhibit significantly less neuronal loss in the sible that this developmentally appropriate injury re-
injured cortex and hippocampus (54). Bcl-2 proteins may sponse represents an attempt at regeneration by the adult
also participate in the control of cell death and survival CNS to traumatic injury.
by regulating the release of cytochrome c from mito-
Does the Response to TBI Mimic Oncogenesis?
chondria, which itself participates in the activation of
members of the caspase family of death-related proteases. The association between TBI and apoptotic cell death
This gene family contains up to 12 known members that, and alterations in cell death/survival pathways including
during CNS development, are associated with the final the Bcl-2 family of genes have been reviewed (52, 53).
steps in the apoptotic cascade. Specific caspases, includ- Other parallels such as single- or double-strand DNA
ing caspase-3, may cleave substrates associated with breaks associated with apoptotic cell death and oncogen-
DNA damage and repair, including DNA-fragmentation esis have also been described (49, 50, 54). Activation of
factor (DFF45/40), poly(ADP-ribose), polymerase endonucleases can result in double-strand DNA breaks,
(PARP), and cytoskeletal proteins actin and laminin. It whereas single-strand breaks are often associated with
has been reported recently that caspase-associated cleav- oxidative damage and formation of reactive oxygen spe-
age products of all of these substrates are expressed after cies (ROS). These ROSs, such as the superoxide anion,

J Neuropathol Exp Neurol, Vol 59, August, 2000

 HEAD INJURY 647

the hydroxyl radical, or singlet oxygen have all been as- component of DFF), which then triggers both DNA frag-
sociated with traumatic CNS injury. In addition to caus- mentation and chromatin condensation during apoptosis.
ing DNA strand breaks, ROS can mediate the formation Although these events are believed to play a role in ap-
of DNA-protein adducts and/or oxidative adducts of the optosis associated with oncogenesis, it has recently been
nitrogen bases, which can participate in DNA fragmen- demonstrated that caspase-3-associated cleavage of
tation. DFF45/DFF40 occurs after lateral fluid percussion brain
Damage to intracellular DNA may lead to apoptosis, injury (67), which may also play a role in acute neuronal
arrest of the cell cycle, or cell growth and/or elimination/ death. The relationship between cell death pathways as-
repair of the damaged DNA. The tumor suppressor gene sociated with oncogenesis and those activated by TBI
p53, often termed the ‘‘guardian of the genome,’’ has remain intriguing and warrant further investigation.
been shown to be induced and upregulated in response
LONG TERM SEQUELAE OF TBI AND THE
to DNA damage. Induction of this gene has been reported
INFLUENCE OF GENETIC FACTORS
after both fluid percussion and controlled cortical impact

Downloaded from [Link] by guest on 07 March 2024

brain injury in the rat, where increased p53 mRNA and Long-Term Outcome after TBI
protein were observed in regions exhibiting TUNEL-pos- Long-term follow-up studies after head injury docu-
itive apoptotic cells (64, 65). Under conditions associated ment persisting problems even after relatively mild in-
with cell cycle arrest, p53 has been suggested to mediate jury. There is relatively little direct high quality infor-
the induction of cyclin D1, a regulator of cyclin-depen- mation concerning the outcome after many years or
dent kinases that is essential for progression through the decades post-injury. However, studies from several dif-
G1 phase of the cell cycle. However alterations in mRNA ferent perspectives point to the occurrence of late neu-
for cyclin D1 were not observed in the acute period after rodegenerative events many years after injury (68), al-
TBI (65). though comprehensive neuropathological studies have
Base excision repair (BER) and nucleotide excision re- not been undertaken.
pair (NER) are the primary mechanisms responsible for There has been much interest in the long-term outcome
the repair of damaged DNA. The BER pathway involves in sporting participants who have experienced relatively
activation of specific repair enzymes, including glyco- minor but repeated head injury. Most obviously these in-
sylases, endonucleases, and ligases and has been associ- clude boxers, a proportion of whom develop a late cog-
ated with oxidative DNA damage. The NER pathway in- nitive decline (dementia pugilistica) often with Parkin-
volves the activation of both exo- and endonucleases to sonian features (69). The neuropathology comprises a
remove the patch of injured or mutated DNA strand that loss of neurons in the substantia nigra and some features
is subsequently replaced through the activation of DNA in common with AD, including neurofibrillary tangles
polymerases. It has been proposed that during oncogen- and Ab-PP containing plaques (70). Of recent interest has
esis and tumor formation, these repair mechanisms are been the demonstration of neurofibrillary tangles in the
compromised. Interestingly, a decrease in nuclear endo- cerebral cortex of young adults in their twenties who had
nuclease activity associated with BER pathways has been sustained relatively minor repeated head injury due to
reported after experimental global cerebral ischemia, sug- boxing and other activities (71).
gesting that neuronal apoptosis may occur from a failure Further information on the long-term effect of head
of DNA repair (66). Moreover, poly(ADP ribose) poly- injury has come from a meta-analysis of studies per-
merase (PARP), a nuclear protein that acts as an endog- formed in the 1980s that showed a significant association
enous detection system for DNA damage repaired by between a history of head injury with loss of conscious-
BER mechanisms (including both single- and double- ness and the risk of subsequent AD, a finding confirmed
strand DNA breaks) has been shown to be activated in by subsequent studies (72). Among patients who devel-
the acute post-traumatic period after lateral fluid percus- oped AD and who had a head injury in the past, the age
sion brain injury (55). Subsequent inhibition of PARP of onset of which was 8 yr less than among patients who
activation in the later stages of TBI in this same study had developed AD without a history of head injury (73).
may be related to caspase-3 induced proteolysis associ-
ated with apoptosis and impaired repair of damaged Evidence For a Genetically Determined Influence on
DNA. Outcome after Traumatic Brain Injury
Recently, a heterodimeric protein composed of 40- and Exploration of the apparent association between head
45-kDA subunits, designated DNA fragmentation factor injury and the subsequent development of dementia re-
(DFF), which is required for the initiation of endonucle- vealed a potential mechanism to explain the link. Plaque-
ase-mediated DNA fragmentation when DFF45 is like deposits of Ab-PP (Ab) were found in the cerebral
cleaved by caspase-3, has been identified. Activated cas- cortex of approximately 30% of patients dying in the
pase-3 will dissociate DFF45 from DFF40 (the active acute stage after head injury: deposits of Ab-PP occurred

J Neuropathol Exp Neurol, Vol 59, August, 2000

648 GRAHAM ET AL

in a higher proportion of head-injured patients than in of a head injury and possession of the APOE e4 allele as
age-matched controls (41). From these findings it was risk factors for AD (79).
inferred that head injury can trigger the deposition of Ab
in the cerebral cortex and that this, in keeping with the Potential Underlying Pathological Mechanisms
‘‘amyloid hypothesis’’ of AD, could render survivors vul- It is tempting to focus on the deposition of Ab-PP as
nerable to later neurodegenerative processes. After the the pathological event triggered by head injury and pre-
recognition that possession of the e4 allele of the apoli- disposed to by possession of APOE e4. Other studies im-
poprotein E (APOE) gene is a major genetic risk factor plicate apoE in binding Ab-PP and promoting aggrega-
for sporadic AD, the relation of the APOE gene poly- tion of fibrils, and APOE e4 shows a robust association,
morphism to the Ab-PP deposits in fatal head injury was in AD and in aging, with Ab-PP deposition both in
explored (74). Patients with APOE e4 who died from plaques in the cerebral cortex and in blood vessels as
head injury were more than 4 times as likely to have cerebral amyloid angiopathy. Particularly with the knowl-
cortical Ab-PP deposits than patients without APOE e4. edge that Ab-PP deposition is an early event in AD, and

Downloaded from [Link] by guest on 07 March 2024

In view of the previous studies this was interpreted to by analogy with dementia pugilistica, the possibility was
indicate that head injury can act as a trigger for Ab-PP raised that head injury related Ab-PP deposition in those
deposition in those people, about a third of the popula- who survive may be followed by the development of AD
tion, with a genetic susceptibility conferred by APOE e4. later in life (41, 74). However, the original interpretation
This evidence suggesting that the neuropathology of head that head injury can trigger the acute deposition of Ab-
injury may be influenced by the APOE gene polymor- PP has been questioned and the suggestion raised that the
phism was a prelude to clinical studies of outcome. Ab-PP deposits could predate the head injury. It was pro-
Several studies have now shown that patients with posed that individuals with APOE e4 were more likely to
APOE e4 have a worse outcome after traumatic brain have pre-existing Ab-PP deposits at a young age and that
injury. A small study (n 5 16) of patients with prolonged individuals with APOE e4 may be more likely to die after
post-traumatic coma revealed a higher frequency of their injury, therefore permitting their inclusion in a post-
APOE e4 among patients who did not recover conscious- mortem study (80). This issue has not been resolved and
it may not be possible to do so within the constraints of
ness than those who did (75). In a prospectively recruited
studies of human autopsy tissue. However, it is of interest
series of patients admitted after a head injury to a neu-
from animal models that plaque-like deposits of Ab-PP
rosurgical unit (n 5 93), it was found that 57% of patients
can indeed form within days (81, 82). This experimental
with APOE e4 had an unfavorable outcome (dead, veg-
finding supports the hypothesis that Ab deposition in hu-
etative state, or severe disability according to the Glas-
mans could be triggered by head injury and the possibil-
gow Outcome Scale) 6 months after injury, compared
ity, in survivors of head injury with APOE e4, that this
with 27% of the patients without APOE e4 (76). This
could predispose to AD-like neurodegeneration later in
remained significant when controlling for age, the sever-
life.
ity of the initial injury as assessed by the Glasgow Coma
However, exploration of other potential mechanisms
Score (GCS), and initial CT scan findings. However, it is
that may mediate the influence of APOE genotype is war-
worthy of note that a higher proportion of the patients ranted. Indeed, it is remarkable the number of processes
with APOE e4 had a low initial GCS, raising the possi- in which apoE appears to be involved, both in the CNS
bility that their poorer response to the injury may be man- and the circulatory system which supports it. For exam-
ifest within hours. A subsequent study found that APOE ple, of potential relevance to the acute response to injury
e4 is strongly associated with more than 7 days of post- is evidence that apoE is involved in the maintenance of
traumatic unconsciousness and patients with APOE e4 are cytoskeletal integrity, protection from oxidative stress,
unlikely to have a good functional outcome (77). It has and modulation of the response to excitotoxicity. APOE
also been shown that APOE genotype influences outcome e4 predisposes to atherosclerosis, ischemic heart disease,
after traumatic brain injury associated with boxing. A and cerebral amyloid angiopathy, which could compro-
study of 30 professional boxers (aged 23–76 yr) showed mise cardiovascular integrity in patients with injuries re-
that high exposure boxers (i.e. those with 12 professional sulting from trauma. Of potential relevance to longer-
bouts or more) who carried APOE e4 had significantly term outcome apoE is involved in the delivery of lipids
greater scores on a clinical scale of chronic traumatic to neurons required for neurite outgrowth and synapto-
brain injury (78). genesis (83), clearance of degeneration products, microg-
Studies addressing the influence of APOE genotype on lial activation, and in maintenance of the cholinergic sys-
the longer-term outcome after trauma, and particularly tem (84).
whether APOE e4 predisposes to a late neurodegenerative As yet it is unclear which specific mechanism or mech-
decline, are awaited. However, of note is the evidence anisms in which apoE is involved are responsible for the
that there is a synergistic interaction between a history poorer outcome in head-injured patients with APOE e4.

J Neuropathol Exp Neurol, Vol 59, August, 2000

 HEAD INJURY 649

However, both human and animal neuropathological 10. Bramlett HM, Dietrich WD, Green EJ, Busto R. Chronic hispath-
ological consequences of fluid-percussion brain injury in rats: Ef-
studies have highlighted the importance of apoE in the
fects of post-traumatic hypothermia. Acta Neuropathol 1997;93:
acute response of brain injury. After brain injury there 190–99
are marked alterations in the cellular localization of apoE, 11. Dixon CE, Kochanek PM, Yan HQ, et al. One-year study of spatial
most notably with an increase in neuronal immunoreac- memory performance, brain morphology, and cholinergic markers
tivity (85, 86). Studies of genetically modified mice have after moderate controlled cortical impact in rats. J Neurotrauma
1999;16:109–22
provided further evidence that apoE modulates the re-
12. Pierce JES, Smith DH, Trojanowski JQ, McIntosh TK. Enduring
sponse to injury. For example, apoE deficient mice have cognitive, neurobehavioral, and histopathological changes persist
more severe lesions than wild type mice (86), and mice for up to one year following severe experimental brain injury in
bearing the human e4 allele have more severe lesions rats. Neurosci 1998;87:359–36
than those with e3 (87). ApoE appears to have a direct 13. Gennarelli TA, Thibault LE, Graham DI. Diffuse axonal injury: An
important form of traumatic brain damage. The Neuroscientist
neuroprotective role since neuronal damage in APOE-de- 1998;4:202–15
ficient mice is ameliorated after entorhinal cortex lesion-

Downloaded from [Link] by guest on 07 March 2024

14. Adams JH, Jennett B, McLellan DR, Murray LS, Graham DI. The
ing and global ischemia by intraventricular infusion of neuropathology of the vegetative state after head injury. J Clin
apoE (88). These animal models may give an indication Pathol 1999;52:804–6
of the mechanisms that are important in humans. 15. Sherriff FE, Bridges LR, Gentleman SM, Sivaloganathan S, Wilson
S. Markers of axonal injury in post mortem human brain. Acta.
If early events such as head injury do precipitate AD, Neuropathol. 1994;88:433–39
a potent acute-phase molecule such as IL1 and the upre- 16. McKenzie KJ, McLellan DR, Gentleman SM, Maxwell WL, Gen-
gulation of neuronal b-APP are necessarily involved with narelli TA, Graham DI. Is b-APP a marker of axonal damage in
the appearance of diffuse plaques of Ab-PP widely dis- short surviving head injury? Acta Neuropathol 1996;92:608–13
17. Blumbergs PC, Scott G, Manavis J, Wainwright H, Simpson DA,
tributed throughout the cerebral cortex in genetically sus-
McLean AJ. Staining of amyloid precursor protein to study axonal
ceptible (APOE e4-positive) individuals (89). These mo- damage in mild head injury. Lancet 1994;344:1055–56
lecular and cellular events are mirrored in the 18. Strich SJ. Diffuse degeneration of the cerebral white matter in se-
neurodegeneration of AD. vere dementia following head injury. J Neurol Neurosurg Psychi-
atry 1956;19:163–74
ACKNOWLEDGMENTS 19. Pettus EH, Povlishock JT. Characterization of a distinct set of intra-
axonal ultrastructural changes associated with traumatically induced
The authors are grateful to Mrs. M. Hughes and Mrs. G. McFarlane alteration in axolemma permeability. Brain Res 1996;772:1–11
for secretarial help. This review has in part been funded by the Medical 20. Povlishock JT, Marmarou A, McIntosh T, Trojanowski JQ, Moroi
Research Council (UK), the Wellcome Trust, and the National Institutes J. Impact acceleration injury in the rat: Evidence for focal axolem-
of Health (USA). ma change and related neurofilament sidearm alteration. J Neuro-
pathol Exp Neurol 1997;56:347–59
REFERENCES 21. Maxwell WL, Kosanlavit R, McCreath BJ, Reid O, Graham DI.
Freeze-fracture and cytochemical evidence for structural and func-
1. Murray GD, Teasdale GM, Braackman R, et al. The European brain tional alteration in the axolemma and myelin sheath of adult guinea
injury consortium survey of head injuries Acta Neurochir (Wien) pig optic nerve fibres after stretch injury. J Neurotrauma 1999;16:
1999;141:223–36 273–84
2. Graham DI, Gennarelli TA. Trauma. In: Graham DI, Lantos PL, 22. Maxwell WL, McCreath BJ, Graham DI, Gennarelli TA. Cyto-
eds. Greenfield’s neuropathology 6th Edition. London: Arnold, chemical evidence for redistribution of membrane pump calcium-
1997:197–262 ATPase and ecto-Ca-ATPase activity, and calcium influx in mye-
3. Gennarelli TA. Animate models of human head injury. J Neuro- linated nerve fibres of the optic nerve after stretch injury. J
trauma 1994;11:357–68 Neurocytol 1995;24:925–42
4. Gennarelli TA, Thibault LE, Adams JH, Graham DI. Diffuse axonal 23. Buki A, Siman R, Trojanowski JQ, Povlishock JT. The role of cal-
injury and traumatic coma in the primate. Ann Neurol 1982;12: pain-mediated spectrin proteolysis in traumatically induced axonal
564–74 injury. J Neuropathol Exp Neurol 1999;58:365–75
5. Smith DH, Chen X-H, Xu B-N, McIntosh TK, Gennarelli TA, Mea- 24. Jafari SS, Neilson M, Graham DI, Maxwell WL. Axonal cytoskel-
ney DF. Characterization of diffuse axonal pathology and selective etal changes after non-disruptive axonal injury. II. Intermediate
hippocampal damage following inertial brain trauma in the pig. J sized axons. J Neurotrauma 1998;15:955–66
Neuropathol Exp Neurol 1997;56:822–34 25. Maxwell WL, Graham DI. Loss of axonal microtubules and neu-
6. Foda MA, Marmarou A. A new model of diffuse brain injury in rofilaments after stretch-injury to guinea pig optic nerve fibers. J
rats. Part II: Morphological characterization. J Neurosurg 1994;80: Neurotrauma 1997;14:603–14
301–13 26. McIntosh TK, Smith DH, Meaney DF, Kotapka MJ, Gennarelli TA,
7. Maxwell WL, Povlishock JT, Graham DI. A mechanistic analysis Graham DI. Neuropathological sequelae of experimental brain in-
of non-disruptive axonal injury: A review. J Neurotrauma 1997;14: jury: Relationship to neurochemical and biomechanical mecha-
419–40 nisms. Lab Invest 1996;74:315–42
8. Gehrmann J, Banati RB, Weisser C, Hossman KA, Kreutzberg GW. 27. McIntosh TK, Juhler M, Wieloch T. Novel pharmacologic strategies
Reactive microglia in cerebral ischaemia: An early mediator of tis- in the treatment of experimental traumatic brain injury: 1998. J
sue damage. Neuropathol Appl Neurobiol 1995;21:277–89 Neurotrauma 1998;15:731–69
9. Smith DH, Chen X-H, Pierce JES, et al. Progressive atrophy and 28. McIntosh TK, Saatman KE, Raghupathi R. Calcium and the path-
neuronal death for one year following brain trauma in the rat. J ogenesis of traumatic CNS injury: Cellular and molecular mecha-
Neurotrauma 1997;14:715–27 nisms. The Neuroscientist 1997;3:169–75

J Neuropathol Exp Neurol, Vol 59, August, 2000

650 GRAHAM ET AL

29. Fineman I, Hovda DA, Smith M, Yoshino A, Becker DP. Concus- behavioral deficits and cell loss. J Cereb Blood Flow Metab
sive brain injury is associated with a prolonged accumulation of 1999;19:762–70
calcium: A 45Ca autoradiographic study. Brain Res 1993;624: 47. Rink AD, Fung K-M, Trojanowski JQ, Lee V, Neugebauer E, Mc-
94–102 Intosh TK. Evidence of apoptotic cell death after experimental trau-
30. Doppenberg EMR, Bullock R. Clinical neuroprotection trials in se- matic brain injury in the rat. Am J Pathol 1995;147:1575–83
vere traumatic brain injury. J Neurotrauma 1997;14:71–80 48. Colicos MA, Dash PK. Apoptotic morphology of dentate gyrus
31. Saatman KE, Murai H, Bartus RT, Hayward NJ, Perri BR, McIntosh granule cells following experimental cortical impact injury in rats:
TK. Calpain inhibitor AK295 attenuates motor and cognitive defi- Possible role in spatial memory deficits. Brain Res 1996;739:
cits following experimental brain injury in the rat. Proc Natl Acad 120–31
Sci 1996;93:3428–33 49. Yakovlev AG, Knoblach S, Fan L, Fox GB, Goodnight R, Faden
32. Kampfl A, Posmantur RM, Zhao X, Schmutzhard E, Clifton GL, AI. Activation of CPP32-like caspases contributes to neuronal ap-
Hayes RL. Mechanisms of calpain proteolysis following traumatic optosis and neurological dysfunction after traumatic brain injury. J
brain injury: implications for pathology and therapy: a review and Neurosci 1997;17:7415–24
update. J Neurotrauma 1997;14:121–34 50. Conti AC, Raghupathi R, Trojanowski JQ, McIntosh TK. Experi-
33. Saatman KE, Bozyczko-Coyne D, Marcy V, Siman R, McIntosh mental brain injury induces regionally distinct apoptosis during the

Downloaded from [Link] by guest on 07 March 2024

TK. Prolonged calpain-mediated spectrin breakdown occurs region- acute and delayed post-traumatic period. J Neurosci 1998;18:
ally following experimental brain injury in the rat. J Neuropathol 5663–72
Exp Neurol 1996;55:850–60 51. Smith F, Raghupathi R, MacKinnon M-A, Saatman KE, McIntosh
34. McCracken E, Hunter AJ, Patel S, Graham DI, Dewar D. Calpain TK, Graham DI. TUNEL staining of contusions after fatal head
activation and cytoskeletal protein breakdown in the corpus callos- injury. Acta Neuropathol (in press)
um in head-injured patients. J Neurotrauma 1999;16:749–61 52. Clark RSB, Chen J, Watkins SC, et al. Apoptosis-suppressor gene
35. Posmantur RM, Kampfl A, Liu SJ, et al. Cytoskeletal derangements bcl-2 expression after traumatic brain injury in rats. J Neurosci
of cortical neuronal processes three hours after traumatic brain in- 1997;17:9172–82
jury in rats: An immunofluorescence study. J Neuropathol Exp 53. Clark RS, Kochanek PM, Chen M, et al. Increases in Bcl-2 and
Neurol 1996;55:68–90 cleavage of caspase-1 and caspase-3 in human brain after head
36. Holmin S, Biberfeld P, Mathiesen T. Intracerebral inflammation af- injury. FASEB Journal 1999;13:813–21
ter human brain contusion. Neurosurgery 1998;42:291–99 54. Raghupathi R, Fernandez SC, Murai H, et al. Bcl-2 overexpression
37. Toulmond S, Rothwell NJ. Interleukin-1 receptor antagonist inhibits attenuates cortical cell loss following traumatic brain injury in
neuronal damage caused by fluid percussion injury in the rat. Brain transgenic mice. J Cereb Blood Flow Metab 1998;18:1259–69
Res 1995;671:261–66 55. LaPlaca MC, Raghupathi R, Verma A, et al. Temporal patterns of
38. Sanderson KL, Raghupathi R, Martin D, Miller G, McIntosh TK. Poly(ADP-Ribose) polymerase activation in the cortex following
Systemic administration of interleukin-1 receptor antagonist (IL- experimental brain injury in the rat. J Neurochem 1999;73:205–13
1ra) attenuates neuronal death and cognitive dysfunction but not 56. Pike BR, Zhao X, Newcomb JK, Posmantur RM, Wang KKW,
neurological motor deficits following lateral fluid-percussion brain Hayes RL. Regional calpain and caspase-3 proteolysis of a-spectrin
injury in the rat. J Cereb Blood Flow Metab 1999;19:1118–25 after traumatic brain injury. NeuroReport 1998;9:2437–42
39. Shohami E, Beit-Yannai E, Horowitz M, Kohen R. Oxidative stress
57. Oyesiki NM, Evans CH, Houston S, et al. Regional changes in the
in closed-head injury: Brain antioxidant capacity as an indicator of
expression of neurotrophic factors and their receptors following
functional outcome. J Cereb Blood Flow Metab 1997;17:1007–19
acute traumatic brain injury in the adult rat brain. Brain Res
40. Stahel PF, Shohami E, Younis FM, et al. Experimental closed head
1999;833:161–72
injury: Analysis of neurological outcome, blood-brain barrier dys-
58. Morrison III B, Eberwine JH, Meaney DF, McIntosh TK. Traumatic
function, intracranial neutrophil infiltration, and neuronal cell death
injury induces differential expression of cell death genes in organ-
in mice deficient in genes for pro-inflammatory cytokines. J Cereb
otypic brain slice cultures determined by cDNA array hybridization.
Blood Flow Metab 2000;20:369–80
Neurosci 2000;96:131–39
41. Roberts GW, Gentleman SM, Lynch A, Murray L, Landon M, Gra-
59. Goss JR, Taffe KM, Kochanek PM, DeKosky ST. The antioxidant
ham DI. b-amyloid protein deposition in the brain after severe head
enzymes glutathione peroxidase and catalase increase following
injury: implications for the pathogenesis of Alzheimer’s disease. J
Neurol Neurosurg Psychiatry 1994;57:419–25 traumatic brain injury in the rat. Exp Neurol 1997;146:291–94
42. Smith DH, Chen XH, Nonaka M, et al. Accumulation of amyloid 60. Patterson SL, Grady SM, Bothwell M. Nerve growth factor and a
beta and tau and the formation of neurofilament inclusions follow- fibroblast growth factor-like neurotrophic activity in cerebrospinal
ing diffuse brain injury in the pig. J Neuropathol Exp Neurol fluid of brain injured human patients. Brain Res 1993;605:43–49
1999;58:982–92 61. Hicks RR, Numan S, Dhillon HS, Prasad MR, Seroogy KB. Alter-
43. Murai H, Pierce JES, Raghupathi R, et al. Two-fold overexpression ations in BDNF and NT-3 mRNAs in rat hippocampus after exper-
of human b-amyloid precursor proteins in transgenic mice does not imental brain trauma. Mol Brain Res 1997;48:401–6
affect the neuromotor, cognitive, or neurodegenerative sequelae fol- 62. Hulsebosch CE, DeWitt DS, Jenkins LW, Prough DS. Traumatic
lowing experimental brain injury. J Comp Neurol 1998;392: brain injury in rats results in increased expression of GAP-43 that
428–38 correlates with behavioral recovery. Neurosci Lett 1998;255:83–86
44. Smith DH, Nakamura M, McIntosh TK, et al. Brain trauma induces 63. Emery DL, Raghupathi R, Saatman KE, Fischer I, McIntosh TK.
massive hippocampal neuron death linked to a surge in b-amyloid Bilateral growth-related protein expression suggests a transient in-
levels in mice overexpressing mutant amyloid precursor protein. crease in regenerative potential following brain trauma. J Comp
Am J Pathol 1998;153:1005–10 Neurol (in press)
45. Nakagawa Y, Nakamura M, McIntosh TK, et al. Traumatic brain 64. Kaya SS, Mahmood A, Li Y, Yavuz E, Goksel M, Chopp M. Ap-
injury in young amyloid-b peptide overexpressing transgenic mice optosis and expression of p53 response proteins and cyclin D1 after
induces marked ipsilateral hippocampal atrophy and dimininished cortical impact in rat brain. Brain Res 1999;818:23–33
Ab deposition during aging. J Comp Neuro 1999;411:390–98 65. Napieralski JA, Raghupathi R, McIntosh TK. The tumor-supressor
46. Nakamura M, Saatman KE, Galvin JE, et al. Increased vulnerability gene, p53, is induced in injured brain regions following experi-
of NFH-LacZ transgenic mouse to traumatic brain injury-induced mental traumatic brain injury. Mol Brain Res 1999;71:78–86

J Neuropathol Exp Neurol, Vol 59, August, 2000

 HEAD INJURY 651
66. Kawase M, Fujimura M, Morita-Fujimura Y, Chan PH. Reduction 79. Mayeux R, Ottman R, Maestre G, et al. Synergistic effects of trau-
of apurinic/apyrimidinic endonuclease expression after transient matic head injury and apolipoprotein-epsilon 4 in patients with Al-
global cerebral ischemia in rats: Implication of the failure of DNA zheimer’s disease. Neurology 1995;45:555–57
repair in neuronal apoptosis. Stroke 1999;30:441–48 80. Roses AD, Saunders A. Head injury, amyloid beta and Alzheimer’s
67. Zhang C, Raghupathi R, Saatman KE, LaPlaca MC, McIntosh TK. disease. Nature Med 1995;1:603–4
Regional and temporal alterations in DNA fragmentation factor 81. Smith DH, Chen XH, Nonaka M, et al. Accumulation of amyloid
(DFF)-like proteins following experimental brain trauma in the rat. beta and tau and the formation of neurofilament inclusions follow-
J Neurochem 1999;73:1650–59 ing diffuse brain injury in the pig. J Neuropathol Exp Neurol
68. Macfarlane DP, Nicoll JAR, Smith C, Graham DI. APOE e4 allele 1999;58:982–92
and amyloid b-protein deposition in long term survivors of head 82. Iwata N, Tsubuki S, Takaki Y, et al. Identification of the major
injury. NeuroReport 1999;10:1–4 Abeta1-42-degrading catabolic pathway in brain parenchyma: sup-
69. Corsellis JA. Boxing and the brain. BMJ 1989;298:105–9 pression leads to biochemical and pathological deposition. Nat Med
70. Roberts GW, Allsop D, Bruton CJ. The occult aftermath of boxing. 2000;6:143–50
J Neurol Neurosurg Psychiatr 1990;53:373–78 83. Bellosta S, Nathan BP, Orth M, Dong LM, Mahley RW, Pitas RE.
71. Geddes JF, Vowles GH, Nicoll JA, Revesz T. Neuronal cytoskeletal Stable expression and secretion of apolipoproteins E3 and E4 in
changes are an early consequence of repetitive head injury. Acta

Downloaded from [Link] by guest on 07 March 2024

mouse neuroblastoma cells produces differential effects on neurite
Neuropathol 1999;98:171–78 outgrowth. J Biol Chem 1995;270:63–71
72. Mayeux R, Ottman R, Tang MX, et al. Genetic susceptibility and 84. Poirier J, Sevigny P. Apolipoprotein E4, cholinergic integrity and
head injury as risk factors for Alzheimer’s disease among com- the pharmacogenetics of Alzheimer’s disease. J Neural Transm
munity-dwelling elderly persons and their first-degree relatives. 1998;53:199–207
Ann Neurol 1993;33:494–501 85. Horsburgh K, Graham DI, Stewart J, Nicoll JAR. Influence of apo-
73. Nemetz PN, Leibson C, Naessens JM, et al. Traumatic brain injury
lipoprotein E genotype on neuronal damage and apoE immunore-
and time to onset of Alzheimer’s disease: A population-based study.
activity in human hippocampus following global ischaemia. J Neu-
Am J Epidemiol 1999;149:32–40
ropathol Exp Neurol 1999;58:227–34
74. Nicoll JAR, Roberts GW, Graham DI. Apolipoprotein E epsilon 4
86. Horsburgh K, Kelly S, McCulloch J, Higgins GA, Roses AD, Nicoll
allele is associated with deposition of amyloid beta-protein follow-
JAR. Increased neuronal damage in apolipoprotein E deficient mice
ing head injury. Nature Med 1995;1:135–37
following global ischaemia. Neuroreport 1999;10:837–41
75. Sorbi S, Nacmias N, Piacentini S, et al. ApoE as a prognostic factor
for post-traumatic coma. Nature Med 1995;1:852 87. Buttini M, Orth M, Bellosta S, et al. Expression of human apoli-
76. Teasdale GM, Nicoll JA, Murray G, Fiddes M. Association of apo- poprotein E3 or E4 in the brains of apoE -/- mice: Isoform-specific
lipoprotein E polymorphism with outcome after head injury. Lancet effects on neurodegeneration. J Neurosci 1999;19:4867–80
1997;350:1069–71 88. Horsburgh K, McCulloch J, Nilsen M, et al. Intraventricular infu-
77. Friedman G, Froom P, Sazbon L, et al. Apolipoprotein E-epsilon4 sion of apolipoprotein E ameliorates acute neuronal damage after
genotype predicts a poor outcome in survivors of traumatic brain global cerebral ischaemia in mice. J Cereb Blood Flow Metab
injury. Neurology 1999;52:244–48 2000;20:458–62
78. Jordan BD, Relkin NR, Ravdin LD, Jacobs AR, Bennett A, Gandy 89. Griffin WS, Sheng JG, Royston MC, et al. Glial-neuronal interac-
S. Apolipoprotein E epsilon4 associated with chronic traumatic tions in Alzheimer’s disease: The potential role of a ‘cytokine cycle’
brain injury in boxing. JAMA 1997;278:136–40 in disease progression. Brain Pathol 1998;8:65–72

J Neuropathol Exp Neurol, Vol 59, August, 2000