The pathobiology of moderate diffuse traumatic brain injury as identified using a new experimental...

www.elsevier.com/locate/ynbdi

Neurobiology of Disease 17 (2004) 29–43

The pathobiology of moderate diffuse traumatic brain injury as

identified using a new experimental model of injury in rats

Ibolja Cernak,a,* Robert Vink,a,b David N. Zapple,c Maria I. Cruz,a Farid Ahmed,a

Taeun Chang,a,d Stanley T. Fricke,a and Alan I. Fadena

aDepartment of Neuroscience, Georgetown University, Washington, DC 20057-1464, USAbDepartment of Pathology, University of Adelaide, Adelaide, SA, AustraliacAdvanced Research Computing, Office of Information Services, Georgetown University, Washington, DC 20057-1464, USAdDepartment of Pediatric Neurology, Children’s Hospital, Washington, DC, USA

Received 5 April 2004; revised 11 May 2004; accepted 28 May 2004

Available online 25 July 2004

Experimental models of traumatic brain injury have been developed to

replicate selected aspects of human head injury, such as contusion,

concussion, and/or diffuse axonal injury. Although diffuse axonal

injury is a major feature of clinical head injury, relatively few

experimental models of diffuse traumatic brain injury (TBI) have been

developed, particularly in smaller animals such as rodents. Here, we

describe the pathophysiological consequences of moderate diffuse TBI

in rats generated by a newly developed, highly controlled, and

reproducible model. This model of TBI caused brain edema beginning

20 min after injury and peaking at 24 h post-trauma, as shown by wet

weight/dry weight ratios and diffusion-weighted magnetic resonance

imaging. Increased permeability of the blood–brain barrier was

present up to 4 h post-injury as evaluated using Evans blue dye.

Phosphorus magnetic resonance spectroscopy showed significant

declines in brain-free magnesium concentration and reduced cytosolic

phosphorylation potential at 4 h post-injury. Diffuse axonal damage

was demonstrated using manganese-enhanced magnetic resonance

imaging, and intracerebral injection of a fluorescent vital dye (Fluoro-

Ruby) at 24-h and 7-day post-injury. Morphological evidence of

apoptosis and caspase-3 activation were also found in the cerebral

hemisphere and brainstem at 24 h after trauma. These results show

that this model is capable of reproducing major biochemical and

neurological changes of diffuse clinical TBI.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Diffuse traumatic brain injury; Rat; Model; Manganese-

enhanced magnetic resonance

Introduction

Traumatic brain injury (TBI) is the most common cause of high

morbidity rates in individuals under 44 years of age and often leads

to severe disability (Bruns and Hauser, 2003; Harris et al., 2003). It

0969-9961/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.nbd.2004.05.011

* Corresponding author. Department of Neuroscience, Georgetown

University, 3970 Reservoir Road NW, Research Building, Room EP04,

Washington, DC 20057-1464. Fax: +1-202-687-0617.

E-mail address: [email protected] (I. Cernak).

Available online on ScienceDirect (www.sciencedirect.com.)

is now well established that neuronal cell death after TBI is caused

by both primary and secondary injury mechanisms and is the major

factor underlying post-traumatic neurological deficits (Faden,

2002; Vink and Nimmo, 2002). Primary injury is defined as

mechanical disruption occurring at the time of trauma and includes

laceration, contusion, shearing, and axonal stretching, among

others. Secondary injury mechanisms include complex biochemi-

cal and physiological processes, which are initiated by the primary

traumatic event, but manifest over a period of hours to days. It has

been acknowledged that such secondary injury may substantially

contribute to chronic neurological disability (DeKosky et al., 1998;

Faden, 1993). Diffuse axonal injury (DAI), which is characterized

by morphologic changes to axons throughout the brain and

brainstem, has contributions from both primary and secondary

mechanisms and is recognized as one of the main consequences of

non-missile TBI leading to the diffuse degeneration of cerebral

white matter (Adams et al., 1989). An important methodological

limitation in neurotrauma research has been the ability to consis-

tently reproduce the pathological mechanisms of clinical injury in

experimental animals, particularly the diffuse component of axonal

injury (Lighthall and Anderson, 1994).

A variety of closed head impact models have been designed to

imitate the biomechanics and pathobiology of human concussive

and diffuse brain injury (Bakay et al., 1977; Beckman and Bean,

1970; Chen et al., 1996; Hall, 1995; Marmarou et al., 1994;

Nilsson et al., 1977; Shapira et al., 1988). This type of trauma is

especially difficult to mimic in small animals using common

experimental models such as fluid-percussion, cortical impact or

focal brain contusion, which cause more focal damage, and are

often accompanied by subdural-, subarachnoid-, and/or intracere-

bral hemorrhages, as well as convulsions. Nonetheless, the

development of such a model is critical for the characterization

of the pathobiology of injury and the subsequent development of

interventional therapies. The first objective of this study was

therefore to develop a highly controlled rat model of diffuse

traumatic brain injury, and to establish the parameters necessary

to cause a reproducible moderate severity of injury, the most

frequent injury group encountered in clinical practice. Having

established these parameters, we characterize the pathobiology of

I. Cernak et al. / Neurobiology of Disease 17 (2004) 29–4330

moderate diffuse traumatic brain injury, describing major mor-

phological, physiological, and biochemical changes comparable

to those observed after moderate clinical TBI.

Materials and methods

All protocols involving the use of animals were in compliance

with the Guide for the Care and Use of Laboratory Animals

published by NIH (DHEW publication NIH 85-23-2985) and

were approved by the Georgetown University Animal Use

Committee.

The injury device

The device consists of an air-driven high-velocity impactor that

is targeted to contact a steel disc cemented onto the rodent skull.

The impactor is of the same diameter as the steel disc, which in the

case of the adult rats used in the present study, is 10 mm. The

animal’s head is supported by a molded, gel-filled base (Hand-

stands, Taiwan), which also acts to decelerate the head after impact.

The middle diameter of the gel is 1.7 cm and its compressibility is

approximately 64 kPa/mm. To ensure precise and even contact

with the steel disc on the animal’s skull, the 14-cm length impactor

can be manually lowered onto the steel disc such that the two

surfaces meet. A laser beam guide is used to confirm that the two

surfaces are parallel and in contact with each other. Having

established this point of contact, the impactor is retracted to a 4-

cm fixed distance above the steel disc. Impact is then initiated

wherein the velocity of the impactor is constant (3.25 m/s),

however, the distance the impactor travels after contacting the

steel disc is under user control. This distance thus determines the

injury severity. Force of impact is controlled and recorded on a

personal computer connected to the device through a PowerLab

(Stoelting, Wood Dakem IL, USA). In this study, the total impact

energy at the end of the impactor was 512.717 N, meaning that

6.528 N/mm2 of impact energy was delivered to the disc.

Induction of injury

Male Sprague–Dawley rats (380F 30 g; n = 360) had access to

food and water ad libitum. Animals were initially anesthetized with

4% isoflurane and 1–1.5% for maintenance; the anesthetic was

evaporated in a gas mixture containing 30% oxygen/70% nitrous

oxide and applied through a nose-mask. The animals were allowed

to breathe spontaneously without tracheal intubation. A midline

incision exposed the dorsal surface of the skull upon which a 10 �3 mm diameter steel disc was cemented centrally between lambda

and bregma using a polyacrylamide adhesive. Rectal temperature

was maintained at 37jC using a thermostatically controlled heating

blanket. After targeting the point of contact between the device

impactor and the steel disc, the impactor was retracted and an

impact displacement was manually entered. Injury was then

induced by activating the air-driven piston.

Groups

The animals were divided into eight groups.

Group I (n = 100): the rats were injured with increasing impact

displacement to determine mortality rates.

Group II (n = 14): included 7 animals with moderate DTBI and

7 sham control rats, used for blood pressure and blood gas

analyses after injury.

Group III (n = 60): 35 animals with moderate DTBI and 25

sham control rats, used for edema study.

Group IV (n = 60): 35 moderately injured and 25 sham control

rats, used for blood–brain barrier study.

Group V (n = 30): 15 moderately injured and 15 sham control

rats, used in studies of neurological outcome and magnetic

resonance spectroscopy (MRS).

Group VI (n = 48): 30 animals with moderate DTBI and 18

sham controls used for morphological (histology and immuno-

cytochemistry) studies.

Group VII (n = 40): 25 moderately injured and 15 sham control

rats, used for fluorescent vital staining.

Group VIII (n = 8): 5 animals with moderate DTBI and 3 sham

control rats, used for functional, manganese T1-weighted

magnetic resonance imaging (MRI).

Acute neurological evaluation and mortality

In Group I, the anesthetized animals were subjected to various

levels of injury, depending on the impact displacement: 16, 17, 18,

19, or 20 mm (n = 20/displacement). Immediately following injury,

anesthesia was discontinued and the animals were maintained on

room air alone. No resuscitation was performed after trauma. Acute

neurological recovery was assessed in all rats by recording indi-

cators of sensorimotor function: recovery of hind paw-flexion

following the gradual application of pressure and the latency to

recovery of the righting reflex. The mortality rate was assessed at

24 h post-trauma.

Arterial pressure and blood gases

Seven animals in Group II were anesthetized and a catheter

(PE-50) was placed into the tail artery. After surgical preparation as

previously described, the animals were subjected to moderate

traumatic brain injury (18 mm vertical head displacement). A

further seven animals served as sham (uninjured) controls. Blood

pressure was monitored with an ITT Cannon transducer (Interna-

tional Biomedical, Austin, TX, USA) and a Narcotrace 40 physio-

graph (Narco Biosystems, Houston, TX, USA) before and at 1 min,

30 min, 1 h, 1.5 h, and 2 h after injury. Arterial blood gases were

measured with a NOVA 8 Blood Gas Analyzer (Nova Biomedical,

Waltham, MA, USA) before injury and at 30 min, 1 h, 1.5 h, and 2

h after injury. Values of blood pressure, as well as pH, pO2, SO2,

pCO2, and HCO3�, were compared both with pre-injury findings

and values measured in control animals.

Edema development assessed by diffusion-weighted magnetic

resonance imaging (MRI) and wet/dry method

The 60 animals in Group III were anesthetized and 35 were

subjected to moderate TBI. At 20 min, 40 min, 2 h, 4 h, and 24

h post-trauma, a subgroup of animals (n = 7/time point) were

reanesthetized with isoflurane and assessed for edema formation by

diffusion-weighted image (DWI) as previously described (Albensi

et al., 2000; Hanstock et al., 1994). A further five uninjured

animals were used as sham controls at every time points. Briefly,

animals were placed in the heated plexiglas holder and a respira-

I. Cernak et al. / Neurobiology of Disease 17 (2004) 29–43 31

tory motion detector positioned over the thorax to facilitate

respiratory gating. The plexiglas holder was then positioned in

the center of the 7 T magnet bore (Bruker 7 T/21 cm Biospec-

Avance system; Bruker, Karlsruhe, Germany) where a 72-mm

proton tuned birdcage coil had been positioned. Field homogeneity

across the brain was then optimized, and sagittal and coronal scout

images obtained to orient the transverse slices throughout the brain

region of interest. Diffusion weighted (DWI) images were then

acquired with a spin echo pulse sequence that had diffusion

gradients added before and after the refocusing pulse. Gradient

strength was varied in six steps using sensitization values ranging

from 10 to 1000 s/mm2. A 256 � 256 matrix was used with a 4-cm

field of view, TR 0.5 s, TE 52.7 ms, slice thickness of 2 mm and 4

echoes. The DWI images were converted to diffusion maps by

applying the Stejskal-Tanner equation in association with a Mar-

quart algorithm using the commercially available Paravision soft-

ware (Bruker, Billerica, MA, USA) so as to highlight changes in

MR signal associated with changes in molecular diffusion. Appar-

ent diffusion coefficients (ADCs) were calculated for four regions:

left cortex, right cortex, left hippocampus and right hippocampus,

and expressed as 10�5 mm2/s F SD. ADC values of left and right

cortices, as well as left and right hippocampi were pooled for

statistical analysis due to a lack of any difference between them,

respectively. As there were no differences between sham control

animals at any time point, they were pooled into one group (C) for

statistical analysis (n = 25).

After completion of DWI, the animals were anesthetized with

sodium pentobarbital (100 mg/kg, i.p.) and decapitated. The brains

were removed rapidly and parietal cortex and hippocampal samples

were collected. The samples were then dried for 72 h at 100jC and

reweighed. Water content was calculated and expressed as a per-

centage of a brain tissue (Elliott and Jasper, 1949).

Blood–brain barrier permeability

A time course for blood–brain barrier (BBB) permeability in

cortex and hippocampus was investigated in animals (n = 60)

assigned to Group IV. The integrity of the BBB was assessed using

Evans blue (EB) extravasation according to Uyama et al. (1988).

Evans blue dye (4 ml/kg of 2% solution, i.e., 0.08 g/kg, in

phosphate-buffered saline) was injected intravenously 30 min

before decapitation at 20 min, 40 min, 2 h, 4 h, and 24 h after

injury (n = 7/time point) or sham surgical procedure (n = 5/time

point). At assigned time points, the animals were anesthetized

using sodium pentobarbital (65 mg/kg i.p.) and intracardially

perfused with saline. The meninges and outer vessels were re-

moved, and samples of cortex and hippocampi were then isolated

and weighed. Brain samples were homogenized in 60% trichloro-

acetic acid solution and centrifuged. The absorbance of the

extracted dye in supernatant at 620 nm was determined with a

Perkin-Elmer spectrometer, and the tissue content of EB quantified

from a linear standard curve obtained using known amounts of the

dye, and expressed as Ag/g of tissue. Because there were no

differences between sham controls, they were pooled for statistical

analysis (n = 25).

Magnetic resonance spectroscopy

Magnetic resonance spectroscopy (MRS) was performed on a

Bruker 7.0 T system operating Paravision 2.0 software. Phosphorus

magnetic resonance spectra were obtained before and at 4 h after

injury as previously described (Heath and Vink, 1996; Vink et al.,

1996) in a subgroup (n = 7 injured and n = 7 sham controls) of

randomly chosen animals assigned to Group V. Briefly, the animal

was placed in a plexiglas holder in which a heating pad warmed to

37jC was fixed to maintain the animal’s temperature. A 5 � 9 mm

surface coil was then placed centrally over the skull and skin and

temporal muscle retracted well clear of the coil to ensure that there

was no contribution from these tissues to the brain spectrum. The

plexiglas holder containing the animal was then positioned in the

center of the magnet bore and the B1 field optimized to less than

0.25 ppm at the half height of the proton water signal. Phosphorus

spectra were subsequently obtained in a 20-min block using a 90jpulse width set at a 2-mm cortical depth, a 800-ms delay time, and

a 6000-Hz spectral width containing 2048 data points. Previous

studies have shown that the sensitive volume of this type coil

extends to approximately one-diameter depth (5 mm) directly

under the coil (Vink et al., 1997). The accumulated free induction

decays were analyzed after convolution difference (20/500 Hz

exponential filter) and peak frequencies determined using the

computer peak-picking program.

Intracellular pH was determined from the chemical shift of the

inorganic phosphate peak (yPi) relative to phosphocreatine (PCr)

using the equation:

pH ¼ 6:77þ logyPi � 3:29

5:68� yPi

� �ð1Þ

Similarly, free magnesium concentration was determined from

the chemical shift difference between the a and h peaks of ATP

(Gupta et al., 1978) using the equation:

½Mg2þ� ¼ Kd

10:82� ya�b

ya�b � 8:35

� �ð2Þ

where ya–b is the chemical shift difference between the a and bpeaks of ATP. The Kd for MgATP was initially assumed to be 50

AM at pH 7.2 and 0.15 M ionic strength and was corrected for pH

by multiplying by a correction factor calculated as (Bock et al.,

1987):

f ¼ ð1þ 106:5�pHÞð1þ 106:5�7:2Þ ð3Þ

Cytosolic phosphorylation potential (PP), which is commonly used

as an indicator of cell bioenergetic status, was determined accord-

ing to the equation

PP ¼ ½AATP�½AADP�½APi�

ð4Þ

where A represents all the ionic forms of the free species. The

concentration of ADP was calculated from the creatine kinase

equilibrium equation after correcting the equilibrium constant for

pH and free magnesium concentration as previously described in

detail elsewhere (Vink et al., 1994). Concentrations of the other

metabolites were determined from the integrated peak areas of the

respective MRS peaks, assuming that pre-injury the normal values

for PCr and ATP were 4.72 and 2.59 Amol/g, respectively, and that

the total creatine pool remained constant at 10.83 Amol/g (Siesjo,

1981; Veech et al., 1979). Brain water content was assumed to be

80%, with the intracellular compartment accounting for 78% of the

total water (Siesjo, 1981).

ogy of Disease 17 (2004) 29–43

Neurological outcome

Motor and cognitive functions were assessed in animals

assigned in Group V (n = 15 injured and n = 15 sham controls).

Motor scoring was performed on days 1, 2, 3, and 7 after

moderate TBI using three separate tests (Faden et al., 2003b),

each of which is scored via an ordinal scale ranging from zero

(no function) to five (normal function). Tests included the ability

to maintain position on an inclined plane in two vertical and two

horizontal positions, forelimb flexion, and forced lateral pulsion.

Forelimb flexion measures the reflex extension of the forelimb to

break a fall when suspended by the tail. Lateral pulsion measures

the degree of resistance to a lateral push. Each of seven

individual scores (vertical angle with the animal positioned head

up, vertical angle with the animal positioned head down, right

and left horizontal angle, forelimb flexion, right and left lateral

pulsion) was added to yield a composite neurological score

ranging from 0 to 35.

Cognitive outcome (spatial learning) was determined using the

hidden platform version of the Morris water maze (MWM) as

previously described (Faden et al., 2003b). Briefly, rats are trained

to locate a hidden, submerged platform using constant extra-maze

visual information. The apparatus consists of a large, white circular

pool (165 cm diameter, 500 mm high, water temperature 24 F1jC) with a 110 mm2 plexiglas platform painted white and

submerged 20 mm below the surface of the water. The surface of

the water is rendered opaque with the addition of dilute, white,

nontoxic paint (Crayola). During training, the platform remained in

a constant location hidden in one quadrant 14 cm from the

sidewall. The rat was gently placed in the water facing the wall

at one of four randomly chosen locations separated by 90j. Thelatency to find the hidden platform within a 90-s criterion time was

recorded by a blinded observer. Trials were conducted on days

14,15, 16, and 17 post-injury.

Histology and immunocytochemistry

At assigned time points (24 h, 48 h, 72 h, 7, 15 and 21 days

post-trauma, Group VI: n = 5/time point), the animals were

anesthetized using sodium pentobarbital (65 mg/kg i.p.) and

prepared for histopathological and morphological analysis. Briefly,

the rats were perfused transcardially with isotonic saline, followed

by 4% formaldehyde. After perfusion, the brains were immersed in

30% of sucrose at 4jC for 24 h. The brains were then removed,

blocked and embedded in paraffin for tissue sectioning. Coronal

10-Am sections were then cut with a cryostat and mounted on glass

slides. Sections were either stained with hematoxlin and eosin

(H&E), used for the immunocytochemical visualization of amyloid

precursor protein (APP; monoclonal primary antibody 22C11

obtained as a gift from Prof. Colin Masters, University of Mel-

bourne, Australia), cleaved caspase-3 [cleaved caspase-3(Asp175)

(5A1) monoclonal antibody; Cell Signaling Technology, Beverly,

MA, USA), or to assess in situ internucleosomal DNA fragmen-

tation, used for TUNEL staining. TUNEL was performed using a

commercially available FD NeuroApopTM kit (FD NeuroTechno-

logies, Ellicott City, MD, USA). All sections were examined by

light microscopy (Zeiss Axioplan; Carl Zeiss Inc., Thoronwood,

NY, USA).

Functional integrity of neuronal cell body and visualization of

axons were analyzed using a vital fluorescent dye and anterograde

tracer Fluoro-Ruby (Molecular Probes Inc., Eugene, OR, USA).

I. Cernak et al. / Neurobiol32

Briefly, the animals (Group VII, n = 40) were initially anesthetized

with 4% isoflurane and anesthesia subsequently maintained on

1.5% isoflurane. The animal’s head was then immobilized stereo-

taxically, and small holes were drilled in the skull with a dental

drill at the sites corresponding to the right and left cingulum.

Fluoro-Ruby (3 Al of 15% solution in PBS, pH 7.4) was injected

into each cingulum using a Hamilton syringe with tip diameter of

50 Am. To avoid diffusion of the dye from the cingulum, the

syringe was left in place for 5 min after the injection before slowly

withdrawing it from the brain. Three days following the injection

of the dye, the animals were subjected to TBI as previously

described. At assigned time points (24 h, 72 h, 7, 15, and 21 days

post-trauma, n = 5/time points), the rats were anesthetized, trans-

cardially perfused, and their brains removed as previously de-

scribed. After 24 h incubation in 30% sucrose at 4jC, the brains

were snap-frozen using 2-methyl-butane pre-cooled in dry ice for 2

h. Coronal 30-Am sections were then cut with a cryostat, with

every fifth section mounted on a glass slide and examined under a

fluorescent microscope using a rhodamine filter (Zeiss Axioplan;

Carl Zeiss Inc.). An additional 15 (n = 3/time point) sham control

animals were used for comparison.

Manganese-dependent contrast magnetic resonance imaging

(MRI)

It has been demonstrated that MnCl2 shows potential as an MRI

contrast agent that may reflect neuronal pathway activation (Lin

and Koretsky, 1997). MnCl2 (120 mM solution, 88 mg/kg) was

administered intravenously, with a flow rate of 2.25 ml/h using a

syringe pump (Harvard Bioscience Inc., Holliston, MA, USA), to

eight rats assigned to Group VIII (n = 5 injured and n = 3 uninjured

sham controls) 48 h before injury or sham procedure. This regime

of MnCl2 administration has been shown to be safe for the animal,

without any toxic effects (Lin and Koretsky, 1997). Moderate 18-

mm injury was then induced as previously detailed. Twenty-four

hours following injury, the animals were prepared for MRI analysis

(see detailed description above). Following optimization of the

field homogeneity across the brain, and obtaining sagittal and

coronal scout images to orient the transverse slices throughout the

brain region of interest, T1-weighted images were acquired (128 �128 matrix, TR 0.5 s, TE 6.4 ms, field of view 3.40 cm, slice

thickness 30 mm, slice separation 30 mm). Regions of hyper-

intensity showed localization of activated neural fibers.

Statistical analysis

All continuous data are shown as mean F SD and were

analyzed by repeated measures analysis of variance followed by

individual Student–Neuman–Keuls tests. Wet weight/dry weight

ratios were compared using Student’s t test. Neurological motor

outcome was analyzed using nonparametric repeated measures

Kruskal–Wallis ANOVA followed by individual Mann–Whitney

U tests.

Results

Acute neurological evaluation and mortality

The 16-mm vertical displacement of the head did not result in

any mortality, whereas the 17-mm head depression produced a

Fig. 1. Systemic, mean arterial and diastolic blood pressures expressed in

mm Hg following moderate traumatic brain injury, induced by 18-mm head

depression in rats. Significant transient increase (5 s post-injury) followed

by a significant reduction in all blood pressures occurred at 1 min following

injury. *P < 0.05 compared to pre-injury levels.


mortality rate of 12%. Moderate severity of head injury (26%

mortality) was caused by 18-mm head depression. Severe TBI was

produced by 19 mm (56% mortality) and 20 mm (90% mortality)

of head depression. Death occurred within 5 min of injury and

resulted from cardiorespiratory depression. The moderate injury

level (18 mm) was subsequently used in all further experiments.

There was no mortality later than 24 h post-trauma.

Following cessation of anesthesia, sham-injured control rats

recovered the pedal withdrawal reflex after a mean 1.56 F 0.18

min and righting reflex after 2.67 F 0.30 min. The 16-mm vertical

displacement did not significantly change these parameters com-

pared to controls (pedal withdrawal reflex: 1.63 F 0.32 min;

righting reflex: 2.90 F 0.31 min). However, animals subjected to

a 17-, 18-, or 19-mm injury levels recovered the pedal withdrawal

reflex more slowly: pedal withdrawal reflex 3.07 F 0.65 min,

6.92 F 0.83 min, or 12.12 F 1.17 min, respectively (for all

groups P < 0.0001 vs. sham-injured group). Similarly, recovery of

the righting reflex took significantly longer (P < 0.0001 vs. sham-

control) in these injury groups: 5.79 F 1.13 min for 17 mm,

11.52 F 1.84 min for 18 mm, and 20.46 F 3.27 min for 19 mm

injury levels.

Arterial pressure and blood gases

Arterial blood pressure increased immediately after induction

of moderate injury, and then declined, reaching a minimum at 1

min post-trauma (Fig. 1). This was followed by a gradual

normalization such that there was no significant difference at

30 min post-injury compared to control values. These changes are

similar to what has been previously described after weight-drop

injury (Marmarou et al., 1994). There were no significant changes

in arterial blood pH, pCO2 or HCO3� at any time point compared

to control animals as well as to pre-trauma levels (Table 1). Since

the animals inhaled a mixture of 30% oxygen and 70% nitrous

oxide, the baseline of pO2 was significantly higher (approximate-

ly 140 mm Hg) than it would be if they would inhale air. Hence,

although pO2 levels were significantly decreased compared to the

baseline, even the lowest values were over 100 mm Hg suggest-

ing that the animals were never hypoxic after injury. Indeed, SO2

remained normal during the 2-h observation period following

injury (Table 1).

Edema development and blood–brain barrier integrity: vasogenic

vs. cytotoxic edema

The percentage water content in cortex of sham control animals

was 78.45 F 0.24%, whereas in hippocampus, it was 77.94 F0.18%; this is consistent with values reported in rats by others

(Bareyre et al., 1997; Kita and Marmarou, 1994). At 20 min after

injury, the water content in cortex of injured animals was 79.20 F0.17% (P < 0.05) and in hippocampus 79.77 F 0.26% (P < 0.001)

indicating acute edema development after moderate diffuse trau-

matic brain injury (Fig. 2A). Increased cortical water content was

also seen at 4 h (79.40F 0.10%, P < 0.05) and 24 h (80.25F 0.22)

post-trauma. In hippocampus, the water content was also increased

at 4 h (80.52 F 0.22, P < 0.001) and 24 h (82.38 F 0.64, P <

0.001) after injury.

The changes in edema measured by wet/dry weight method

were confirmed using diffusion-weighted MRI. Additionally, this

method is accepted as a sensitive tool for distinguishing between

vasogenic and cytotoxic edema following traumatic brain injury

(Li and Fisher, 1996). Increased apparent diffusion coefficient

(ADC) in the brain indicates vasogenic edema whereas reduced

ADC suggests cytotoxic edema (Li and Fisher, 1996). Significant

changes in ADC were found in both cortex and hippocampus at

20 min, 4 h, and 24 h following injury, compared to control

values (Fig. 2B). ADC maps derived from the DWI are shown in

Fig. 2C. ADC was significantly increased in cortex and hippo-

campus of injured animals at 20 min and 4 h following injury,

which reflects vasogenic edema. This is manifested as areas of

hyperintensity on the ADC maps. At 24 h post-trauma, decreased

ADC suggested development of cytotoxic edema. Indeed, the

ADC map showed regions of hypointensity diffusely scattered

throughout the brain, reflecting a decrease in the molecular

motion of water, consistent with accumulation of water within

the intracellular space. Such a biphasic development of edema

has been previously reported in other diffuse models of brain

trauma (Barzo et al., 1997).

Increase in ADC was associated with a significant increase in

the permeability of the blood–brain barrier (BBB) at 20 min and 4

h after injury as assessed by penetration of Evans blue dye (Fig. 3).

Since increased BBB permeability associated with edema at early

time points following TBI has been seen to correspond to vaso-

genic edema (Tanno et al., 1992), these results support our DWI

findings suggesting a vasogenic origin of acute edema. Moreover,

Table 1

Blood gas changes in spontaneously breathing animals after moderate diffuse traumatic brain injury

Post-trauma

Variable Group Pre-trauma 30 min 1 h 1.5 h 2 h

pH Control 7.44 F 0.04 7.45 F 0.03 7.47 F 0.04 7.45 F 0.04 7.46 F 0.05

Injured 7.45 F 0.03 7.43 F 0.07 7.45 F 0.05 7.47 F 0.02 7.44 F 0.06

pO2 Control 141.6 F 7.5 147.5 F 9.1 142.4 F 8.8 139.3 F 9.7 132.7 F 7.5

Injured 139.3 F 7.7 115.1 F 17.3* 107.8 F 16.8* 104.2 F 16.1* 102.7 F 14.3*

SO2 Control 98.9 F 0.4 99.3 F 0.6 98.3 F 0.5 98.9 F 0.6 98.2 F 0.7

Injured 98.7 F 0.4 98.1 F 0.4 97.9 F 0.5 97.9 F 0.7 97.3 F 1.3

pCO2 Control 30.2 F 4.4 30.4 F 5.6 32.3 F 6.7 34.9 F 6.6 33.2 F 2.7

Injured 33.0 F 7.6 36.7 F 5.2 37.2 F 5.5 36.5 F 3.5 39.7 F 8.7

HCO3� Control 24.0 F 2.2 25.3 F 1.8 25.6 F 1.7 24.0 F 2.6 26.0 F 2.1

Injured 25.9 F 2.7 26.5 F 1.8 27.1 F 2.3 27.4 F 0.7 26.9 F 2.7

Statistical significance of difference: *P < 0.05, **P < 0.01, ***P < 0.001 compared to pre-trauma level.


significantly reduced ADC and the lack of a significant increase in

Evans blue dye concentration in brain structures at 24 h post-

trauma confirm cytotoxic edema formation at this time point.

Magnetic resonance spectroscopy

The phosphorus magnetic resonance spectroscopy (MRS) at

4 h after trauma demonstrated that there were no significant

Fig. 2. (A) Changes in water content measured in cortex in hippocampus by wet

Apparent diffusion coefficients (ADC) obtained in cortex and hippocampus follo

***P < 0.001 compared to sham control values. (C) Apparent diffusion coeffici

obtained in sham control animals and in animals with moderate traumatic brain in

animals at 4 h postinjury indicates vasogenic edema, while scattered regions of

cytotoxic edema.

changes in ATP concentration at this time point compared to

sham controls (results not shown). Similarly, there was no

significant difference in pH at 4 h post-trauma (7.15 F 0.04)

compared to sham values (7.13 F 0.04). This lack of significant

change in ATP concentration and pH after trauma is consistent

with previous studies of moderate diffuse brain injury (Heath and

Vink, 1995; Smith et al., 1998). There was, however, a significant

decline (P < 0.05) in intracellular brain-free magnesium concen-

weight/dry weight method following moderate traumatic brain injury. (B)

wing moderate traumatic brain injury in rats; *P < 0.005, **P < 0.01, and

ent maps derived from the diffusion weighted magnetic resonance images

jury at 4 and 24 h postinjury. Areas of subcortical hyperintensity in injured

hypointensity (white arrows) at 24 h postinjury suggests development of

Fig. 3. Changes in blood–brain barrier (BBB) permeability to Evans

blue dye in cortex and hippocampus of rats subjected to moderate

traumatic brain injury. *P < 0.05 and **P < 0.01 compared to sham

controls (C).


tration, with the free concentration of the ion falling to 0.31 F0.02 mM by 4 h post-trauma compared to the sham control value

of 0.50 F 0.06 mM (Fig. 4A). The change in free magnesium

Fig. 4. (A) Brain-free magnesium concentration at 4 h after moderate

traumatic brain injury in rats. (B) Brain cytosolic phosphorylation potential

at 4 h following diffuse traumatic brain injury in rats. *P < 0.05 compared

to preinjury levels.

concentration paralleled a significant reduction in phosphorylation

potential 4 h following injury (21 F 2 in injured vs. 40 F 4

mM�1 (Fig. 4B). Previous studies have shown that this decline is

typical in traumatic brain injury and occurs irrespective of the

model or the species used (for review, see (Vink and Cernak,

2000). The preservation of ATP and the lack of acidosis at 4 h

after injury also indicate an absence of ischemic indicators,

consistent with the vasogenic origin of the observed edema at

this time point following moderate traumatic brain injury.

Motor and cognitive outcomes

Motor and cognitive outcomes are summarized in Fig. 5. All

animals demonstrated a significant decline (P < 0.05) in motor

function after moderate (18 mm) trauma (Fig. 5A). Rats subjected

to an 18-mm head trauma exhibited similar reduction in com-

Fig. 5. (A) Changes in composite neuroscore 1, 2, 3, and 7 days after

moderate traumatic brain injury in rats; (B) Water maze cognitive score

after moderate traumatic brain injury in rats measured at 14, 15, 16, and

17 days postinjury. *P < 0.05 and ***P < 0.001 compared to sham control

animals.

Fig. 6. The brain of a moderately brain-injured rat 24 h following injury.

(A) Superior surface; (B) inferior surface; (C) coronal section. No focal

lesions or contusions were observed. Subarachnoid hemorrhage and

intraventricular hemorrhage (black arrows) were frequent findings.


posite neuroscore at 1, 2, and 3 days post-injury (30.67 F 0.8,

30.33 F 0.5, and 29.67 F 0.4, respectively, vs. 35 F 0 before

injury), and showed a trend toward neurological improvement at

7 days after injury (31.33 F 0.3). Although the deficits are

moderate, they are consistent with those previously described in

other rodent models of diffuse trauma as determined by the

composite motor tests used in the present study (Heath and Vink,

1995). We found that the forelimb flexion and forced lateral

pulsion tests lacked sensitivity to show motor deficits after

diffuse traumatic brain injury. This can be explained by the fact

that these methods are based upon the ‘‘lateralization effect’’

induced by unilateral brain injury, which develops as the undam-

aged hemisphere attempts to compensate for the damaged or

disconnected region. On the other hand, the impact energy in our

model is delivered centrally to the skull, and distributed homo-

genously over the brain surface. However, the method of inclined

plate examining fine motor coordination and balance proved to be

sensitive to indicate motor deficits following diffuse TBI induced

by our model, and the reduction in composite neuroscore reflects

changes measured by these tests.

The Morris Water Maze was performed over 4 consecutive days

starting at 14 days after sham surgery or TBI (Fig. 5B). The mean

latency to locate the platform immersed below the surface of the

water was significantly increased in injured animals at 15 (P <

0.001), 16 (P < 0.05) and 17 (P < 0.001) days postinjury, compared

to sham control animals.

Pathological changes: histology and immunocytochemistry

Gross pathological observation of the brains following moder-

ate (18 mm) brain injury show an absence of focal lesions or

contusion. Subarachnoid hemorrhage was usually observed in the

basal cisterns, cisterna magna, and over the cerebral hemispheres

(Fig. 6).

Histological examination of hematoxylin and eosin (H&E)-

stained tissue sections demonstrated profound dark cell change

and reduced neuronal density in the CA3 pyramidal layer of the

hippocampus (Fig. 7). There were also shrunken neurons

associated with perineuronal vacuolation in supraventricular

cortical areas and the dentate gyrus after injury, together with

pericapilllary edema in supraventricular regions of the cerebral

cortex, brainstem, and thalamus; in these areas, several con-

gested capillaries also were seen (results not shown). A pro-

found increase in APP immunoreactivity in neuronal cell bodies

was also present at 24 h, particularly in the brainstem (Fig. 8).

This increase in APP has been noted in other models of diffuse

traumatic brain injury and is thought to represent an acute

phase post-traumatic stress response (Van den Heuvel et al.,

1999).

Induction of moderate (18 mm) TBI resulted in internucleo-

somal DNA fragmentation as assessed by the positive Terminal

Deoxynucleotidyl Transferase Biotin-dUTP Nick End Labeling

(TUNEL) histochemistry. TUNEL-stained cells were clearly ap-

parent throughout the brain at 1, 2, and 3 days post-trauma.

These TUNEL-positive cells were found mainly in cortex, thala-

mus, and brainstem. The largest number of TUNEL-positive cells

was seen 24 h following injury (Fig. 9B), whereas no TUNEL-

positive cells were detected after 7 days post-trauma. Sham

control (non-injured) tissue did not demonstrate any TUNEL-

positive cells at any time point. The TUNEL findings were

confirmed by the Fluoro-Ruby technique, which showed cells

with condensed nuclei and marginated chromatin in cortex and

thalamus (Fig. 9D).

To elucidate whether caspase-3-mediated apoptosis was asso-

ciated with the TBI-induced cell death, the distribution of a

cleaved (active) caspase-3 in the brain if injured animals was

measured by qualitative immunocytochemistry (Fig. 10). The

antibody used [cleaved caspase-3(Asp175) (5A1) monoclonal

antibody; Cell Signaling Technology] does not detect the pro-

caspase-3 form; it identifies endogenous levels of the large

fragment (17/19 kDa) of activated caspase-3 resulting from

cleavage adjacent to Asp175. As such, this antibody can be used

as a cellular marker of caspase activation. The cleaved form of

caspase-3 was expressed in cortex, hippocampus (not shown),

and widely distributed throughout the entire region of brainstem

by 24 h after injury (Fig. 10).

Analysis of morphological characteristics and distribution of

Fluoro-Ruby labeled cells showed microglia and cell debris in

supraventricular cortical areas (Figs. 11C and D) and in brain-

stem (not shown) at 24 h after injury. At 7 day post-trauma, a

significant reduction of neurons and further increase in micro-

glia were observed (Figs. 11G and H). Similarly, Fluoro-Ruby-

Fig. 7. Representative photomicrograph of the hippocampal CA3 region 24 h after the 18-mm injury (C and D) compared to a sham injured animal (A and B);

H&E staining. Magnifications A and C �3, B and D �20. Reduced neuronal density and cell loss as well as profound dark cell changes are shown.


labeled axons in the caudate putamen showed mild axonal

pathology 24 h after injury (Fig. 12B), whereas retraction balls,

broken fibers, and disrupted axons were observed at 7-day post-

injury (Fig. 12D).

Manganese-dependent contrast magnetic resonance imaging

(MRI)

Manganese-enhanced T1-weighted MRI was used to demon-

strate and compare the regional difference of manganese accumu-

lation in injured vs. non-injured animals. Our results demonstrated

hyperintense areas corresponding to dentate gyrus and interpedun-

Fig. 8. Representative photomicrograph of amyloid precursor protein (APP) accu

injury 24 h after the insult (C and D), compared to a sham-injured animal (A and

and D �20. The APP protein accumulation in axons and neuronal cell body is s

cular nuclei (Fig. 13) at 24 h post-injury, whereas absence of

positive contrast enhancement was shown in the brains of naive

(not shown) and sham-injured animals (Fig. 13).

Discussion

Methodological aspects

Several models of traumatic brain injury have been developed

to reproduce diffuse axonal injury (DAI), which is one of the

primary aspects of human head trauma (Chen et al., 1996;

mulation in the brain of an animal with moderate (18 mm) traumatic brain

B). APP 22C11 monoclonal antibody used. Magnifications A and C � 3, B

hown as darkly staining immunoreactivity.

Fig. 9. (A and B) TUNEL immunocytochemistry in rat brain cortex at 24 h

after moderate traumatic brain injury. Darkly stained TUNEL-positive

nuclei (arrows) are clearly observed distributed throughout the cortex (B).

Such dark staining nuclei are absent in sham control (A). Scale bar = 100

Am. (C and D) Representative photomicrographs of Fluoro-Ruby-labeled

cells in thalamus. Cells with condense nuclei and marginated chromatin

(arrows) are clearly observed at 24 h postinjury (D) compared to a sham

control (C) Scale bar = 50 Am.


Gennarelli, 1994; Lighthall and Anderson, 1994; Povlishock et al.,

1994; Shapira et al., 1988). While many earlier brain injury models

used higher species such as primate, sheep, and cat, recent models

Fig. 10. Immunocytochemical staining of paraffin-embedded rat brain shows dar

injury using cleaved Caspase-3 (Asp175) (5A1) Rabbit Monoclonal Antibody, whi

caspase-3. Scale bar = 100 Am.

have often focused on rodents (Povlishock et al., 1994). The non-

impact acceleration injury models in nonhuman primates and

miniature swine appear to closely reproduce the complex pathobi-

ology of human traumatic brain injury (Smith et al., 1997).

Although these techniques offer valuable information on morpho-

logical, cellular and molecular responses to diffuse brain injury,

their use is difficult for most laboratories due to the cost and size of

the animal, as well as the sophisticated technical requirements

(Povlishock et al., 1994). Moreover, the models of acceleration

head injury using large experimental animals lack established

functional outcome tests, which are critical to the preclinical

assessment of neuroprotective strategies. Since rodent models are

inexpensive, easy to use, and generally involve highly inbred

strains of one sex, they have been found to be useful tools to

reduce intersubject and outcome variability (Faden, 2001).

One of the most frequently used rodent models of impact

acceleration head injury is Marmarou’s weight drop model (Foda

and Marmarou, 1994; Marmarou et al., 1994). The trauma device

utilized in this model consists of a column of brass weights falling

freely by gravity from a designated height through a Plexiglas tube

onto a stainless steel disc fixed to the skull. This model has been

shown to produce graded brain injury in rodents depending upon

the mass and the height from which the brass weight is released

(De Mulder et al., 2000; Marmarou et al., 1994; Piper et al., 1996).

The popularity of the weight drop model is based on the fact that it

is inexpensive, easy to perform, and capable of producing graded

diffuse axonal injury. However, the biomechanics of the impact

produced by this model are not strictly controlled. For example,

there is a possibility of a ‘‘second hit’’ induced by the weight after

rebounding from the skull of the animal resting on the flexible

sponge. Furthermore, although the weight is enclosed within a

Plexiglas tube, some lateral movement occurs during its fall, which

has the potential to cause lateralization of the impact, uneven

kly stained apoptotic cells (white arrows) in the brainstem 24 h following

ch detects endogenous levels of the large fragment (17/19 kDa) of activated

Fig. 11. Representative photomicrographs of Fluoro-Ruby-labeled cells in cortex. A, B, E and F show cortex in a sham control animal, while C, D, G, and H show

cortex in an injured animal. A, C, E, and G are with a lower magnification (the scale bar represents 50 Am), whereas B, D, and H demonstrate a region (white box)

with a higher magnification (the scale bar represents 200 Am). At 24 h postinjury (C), there are few microglia (white arrows) and cell debris (yellow arrows); at 7-

day postinjury (F) decrease in neuronal cell number, increased number of microglia (white arrows), and presence of debris (yellow arrows) are observed.

Fig. 12. Representative photomicrographs of Fluoro-Ruby labeled axons in

the caudate putamen. A and C show caudate putamen in a sham control

animal, while B and D represent caudate putamen in a moderately injured

rat. At 24-h postinjury (B), there is no noticeable change compared to a

sham control (A); at 7 days (D) postinjury, retraction balls (black arrow)

and broken fibers (white arrow) are visible. Scale bar = 50 Am.


distribution of the impact energy and increased variability of

outcome measures. As with Marmarou’s model, we also use a

stainless-steel disk to protect the skull, as well as to disperse the

impact energy over the surface of the brain. However, our device

enables control of the velocity as well as of the dwell time during

which the impactor is in contact with the disk, that is, with the

animals’ head. It is well known that these biomechanical factors

are the most important determinants of the injury severity (Gold-

smith, 2001); thus, their inconsistency inevitably increases the

variability of the injury outcome. Moreover, using a laser-beam

guide incorporated in our device permits precise centralization of

the hit, preventing uneven delivery of impact energy to the skull.

Our model also demonstrates a direct relationship between the

injury outcomes/mortality and the vertical displacement of the head

(i.e., intensity of the impact energy inflicting the injury), thus

fulfilling one of the most important criteria for a reliable experi-

mental model of TBI.

In the present studies, recovery latencies for pedal withdrawal

and righting reflexes measured immediately after injury were

dependent upon vertical displacement of the head, that is, levels

of injury. In clinical practice, severity of brain injury is routinely

graded by the Glasgow Coma Scale, which is based upon eye

opening, as well as verbal and motor responses (Teasdale and

Jennett, 1976). Although translation of neurological parameters

observed in experimental animals into outcome parameters of

patients has proven to be difficult (Teasdale et al., 1999), our

results demonstrate a causal relationship between acute neurolo-

gical/motor dysfunction and injury severity.

Physiological and functional responses to diffuse traumatic brain

injury

The systemic response to traumatic head injury can significant-

ly modify complex cellular and molecular changes in the brain

induced by impact. Indeed, the hypertensive response to fluid

Fig. 13. Manganese-dependent contrast T1-weighted MRI in sham control and moderately injured rats at 24 h postinjury. Hyperintense areas correspond to

dentate gyrus (white arrow) and interpeduncular nuclei (black arrows).


percussion injury, in which blood pressure exceeds 180 mm Hg for

several minutes (Yuan et al., 1991), has been shown to cause loss

of autoregulation and a marked elevation of cerebral blood flow

(Lewelt et al., 1980). Such pathological alterations due to hyper-

tension could significantly compromise investigation of trauma-

induced changes in blood–brain barrier integrity and cerebral

autoregulation. In contrast, TBI induced using our model caused

only a small and transient elevation of blood pressure (5–10 s)

followed by mild hypotension at 1 min post-trauma, with the blood

pressure returning to normal, pre-injury levels by 30 min post-

trauma. Similar dynamics of blood pressure were shown after

impact injury inflicted by Marmarou’s model (Marmarou et al.,

1994). Hypotension was also shown in patients with isolated brain

injury (Mahoney et al., 2003).

Because many of the critical energetic processes occurring in

the cell are essentially phosphate transfer reactions, phosphorus

MRS is one of the most widely used techniques used in the study

of brain metabolism. It has been established that trauma, which

results in mild to severe levels of motor dysfunction does not cause

loss of ATP or sustained lactic acidosis, but rather induces more

subtle changes in bioenergetics reflected as reduction in calculated

cytosolic phosphorylation potential and decrease in free magne-

sium concentration (Vink et al., 1988, 1994). In our study, MRS

analysis of the brain confirmed the absence of indicators of

cerebral ischemia/hypoxia, and demonstrated significant decrease

in cytosolic phosphorylation potential parallel with a reduction in

free Mg2+, suggesting increased energy demand in response to

attempts to maintain ionic gradients after injury (Vink and Cernak,

2003). Our findings are comparable with changes found after

impact-acceleration brain injury induced by Marmarou’s model

(Heath and Vink, 1995). The similarity in metabolic events

associated with production of neurologic deficits in our and

Marmarou’s model, as well as in other classical experimental

models of traumatic brain injury (Vink et al., 1988) suggests that

these bioenergetic changes may be common to major experimental

models of brain trauma. Indeed, MRS showed similar alterations in

the brain of head-injured patients (Garnett et al., 2001).

Currently, two major forms of brain edema are recognized

(Klatzo, 1987b): (1) vasogenic, which involves a breakdown of

the blood–brain barrier (BBB) leading to accumulation of edema

fluid in the extracellular spaces, and (2) cytotoxic, where swelling

of cellular elements of brain parenchyma occurs without BBB

breakdown. While the wet and dry weight method measures total

tissue water content (Elliott and Jasper, 1949) without giving

information on the type of edema, diffusion-weighted imaging

(DWI), and calculation of apparent diffusion coefficient (ADC)

offer the opportunity of a real-time detection and differentiation of

the edema formation (Hanstock et al., 1994). In our study, wet and

dry weight results show significant increase in water content at 20

min, 4 h and 24 h after injury (Fig. 2A). Increased ADC (Fig. 2B)

together with BBB breakdown (Fig. 3) at 20 min and 4 h post-

trauma suggest an early and rapid opening of the BBB and

development of vasogenic brain edema at these time points. Early

vasogenic edema was also shown after impact acceleration injury

induced by Marmarou’s weight-drop model (Barzo et al., 1997),

controlled cortical impact injury (Baskaya et al., 1997), closed

head injury (Assaf et al., 1999), and following major cerebral

artery occlusion (Klatzo, 1987a). Although total tissue water

content, measured by wet and dry method, is significantly in-

creased at 24 h post-trauma, absence of increased BBB permeabil-

ity at the same time point implies that the brain edema is not of

vasogenic origin. Indeed, significantly decreased ADC at 24 h after


injury suggests cytotoxic edema (Hanstock et al., 1994). Complex

secondary injury factors initiated at the time of traumatic brain

injury (TBI) and manifesting for minutes or even days after the

traumatic insult have been shown to induce cytotoxic edema in

various models of TBI (Assaf et al., 1999; Ito et al., 1996;

Loubinoux et al., 1997). These findings suggest that our traumatic

brain injury model induces early vasogenic edema followed by

more delayed cytotoxic edema. The fact that alterations in water

content and ADC values showed similar trends in both cortex and

hippocampus suggests that the trauma in this model results in

diffuse injury rather than a more focal brain damage. Brain edema

has been shown as one of the main consequences of TBI in

patients. Rozsa et al. (1989) analyzed 107 patients with head

injury and indicated that diffuse swelling was present in 91% of

patients. These findings are consistent with the results published by

Marmarou et al. (2000), showing brain edema development in 87%

(66 of 76) of head-injured patients. The temporal profile of edema

development following our experimental TBI is comparable with

the temporal profile of the brain swelling in injured patients

(Bullock et al., 1991; Marmarou et al., 2000).

Our model, similar to various rodent models of TBI including

fluid percussion (Faden et al., 2003b; Hamm et al., 1993),

controlled cortical impact (Faden et al., 2003a; Scheff et al.,

1997), and impact-acceleration brain injury (Chen et al., 1996;

Schmidt et al., 2000; Zohar et al., 2003), induces cognitive deficits

using the Morris water maze test. The model also induces moderate

motor deficits measured by composite neuroscore, which are

comparable to motor impairments observed in certain rodent TBI

models (Chen et al., 1996; Heath and Vink, 1995; Raghupathi et al.,

1998).

Pathological changes induced by diffuse traumatic brain injury

Adams et al. (1989) identified diffuse axonal injury as major

pathological feature of human fatal non-missile head injury, and

classified it into three grades. Grade 1 is associated with histolo-

gical evidence of axonal injury in the white matter of the cerebral

hemispheres, the corpus callosum, and the brain stem, and less

commonly, the cerebellum. In Grade 2, there is also a focal lesion

in the corpus callosum. Grade 3 is characterized by an additional

focal lesion in the dorsolateral parts of the rostral brain stem.

Amongst experimental models of TBI, Marmarou’s model (Foda

and Marmarou, 1994) successfully replicated Grade 1 DAI ob-

served in cerebral hemisphere and brain stem following mild

injury, without focal lesion in the corpus callosum or in the brain

stem. Furthermore, Marmarou’s severe injury model has been

shown to reproduce Grade 3 DAI, which was in cerebellum and

in the brain stem. However, there was a lack of evidence for similar

changes in the corpus callosum (Foda and Marmarou, 1994). The

authors hypothesized that focal lesions in the corpus callosum

could be caused only by a head acceleration oriented in coronal

plane and not in the sagittal plane (Foda and Marmarou, 1994).

Indeed, Gennarelli et al. (1982), using the acceleration-rotational

head injury model in primates, has demonstrated that coronal head

acceleration is a major cause of prolonged traumatic coma and

underlying DAI, whereas acceleration in sagittal plane caused less

neurological deficits and less DAI. In our study, we demonstrated

diffuse axonal injury in the cerebral hemispheres and in the brain

stem, and observed reduced numbers of intact fibers in the corpus

callosum without signs of focal lesions. Furthermore, we found

progressive reduction of neuronal cells with a parallel increase in

microglia. Neuronal cell death due to impact showed features of

both necrosis and apoptosis, similar to previously developed

models of closed head injury (Stahel et al., 2000). Additionally,

the involvement of caspase-3 activation in injury-induced cell

death was demonstrated, which is consistent with our previous

findings using Marmarou’s model of diffuse traumatic brain injury

(Cernak et al., 2002), as well as with fluid percussion induced

injury (Knoblach et al., 2002; Yakovlev et al., 1997).

Recently, Lin and Koretsky (1997) demonstrated that focal

injections of manganese (Mn2+) within the mouse central nervous

system combined with in vivo high-resolution magnetic resonance

imaging (MRI) demarcate neuronal tracts originating from the site

of injection. Others, using alternative (intraarterial) route of injec-

tion coupled with artificial opening of the blood–brain barrier,

confirmed the usefulness of Mn2+ contrasted magnetic resonance

imaging in rats (Aoki et al., 2003; Morita et al., 2002). Manganese,

similarly to Ca2+ in various biological systems (Shibuya and

Douglas, 1993), is known to enter glial cells and/or neurons via

voltage-gated Ca2+ channels following the triggering of an action

potential (Narita et al., 1990). Indeed, this new visualization

method is described as a useful tool to detect manganese accumu-

lation caused by permanent ischemia in rat brain, whereas manga-

nese hyperintensity reflected the increased Ca2+ influx (Aoki et al.,

2003). Here, we demonstrate for the first time the utilization of this

method in experimental TBI, and show its usefulness in differen-

tiation of neuronal structures activated by trauma. Since traumatic

brain injury induces blood–brain barrier breakdown, which is well

documented by literature data as well as our own results, in this

study, there was no need for artificial opening the blood–brain

barrier by administration of a hyperosmolar agent, necessary for

the entry of MnCl2 into brain parenchyma. Although future

experimental work is necessary to fully define the potential and

limitations of this methodology in TBI, these findings indicate that

Mn2+ contrasted MRI may be a useful technique for investigating

the trauma-induced functional changes in the brain.

In conclusion, we have developed a new model for diffuse

traumatic brain injury, which induces biochemical and neurological

changes consistent with the diffuse axonal injury produced in other

models and comparable to alterations found after clinical TBI. This

model offers a high level of control of mechanical force inflicting

the injury, thereby reducing the intersubject and outcome variabil-

ity. Moreover, by reproducing both features of necrosis and

apoptosis, this model is suitable for studying complex mechanisms

of post-traumatic cell death and testing novel therapeutic

approaches targeting diffuse axonal damage.

Acknowledgments

We acknowledge Jill MacLeod, Lilia Ileva and Thomas Krupica

Jr. for their outstanding technical assistance, along with Drs. Jung

Hee Lee and Alan P. Koretsky (Laboratory of Functional and

Molecular Imaging, NINDS, NIH) for their valuable directions

concerning manganese-enhanced MRI.

References

Adams, J.H., Doyle, D., Ford, I., Gennarelli, T.A., Graham, D.I., McLellan,

D.R., 1989. Diffuse axonal injury in head injury: definition, diagnosis

and grading. Histopathology 15, 49–59.


Albensi, B.C., Knoblach, S.M., Chew, B.G., O’Reilly, M.P., Faden, A.I.,

Pekar, J.J., 2000. Diffusion and high resolution MRI of traumatic brain

injury in rats: time course and correlation with histology. Exp. Neurol.

162, 61–72.

Aoki, I., Ebisu, T., Tanaka, C., Katsuta, K., Fujikawa, A., Umeda, M.,

Fukunaga, M., Takegami, T., Shapiro, E.M., Naruse, S., 2003. Detec-

tion of the anoxic depolarization of focal ischemia using manganese-

enhanced MRI. Magn. Reson. Med. 50, 7–12.

Assaf, Y., Holokovsky, A., Berman, E., Shapira, Y., Shohami, E., Cohen,

Y., 1999. Diffusion and perfusion magnetic resonance imaging follow-

ing closed head injury in rats. J. Neurotrauma 16, 1165–1176.

Bakay, L., Lee, J.C., Lee, G.C., Peng, J.R., 1977. Experimental cerebral

concussion. Part 1: an electron microscopic study. J. Neurosurg. 47,

525–531.

Bareyre, F., Wahl, F., McIntosh, T.K., Stutzmann, J.M., 1997. Time course

of cerebral edema after traumatic brain injury in rats: effects of riluzole

and mannitol. J. Neurotrauma 14, 839–849.

Barzo, P., Marmarou, A., Fatouros, P., Hayasaki, K., Corwin, F., 1997.

Contribution of vasogenic and cellular edema to traumatic brain

swelling measured by diffusion-weighted imaging. J. Neurosurg. 87,

900–907.

Baskaya, M.K., Rao, A.M., Dogan, A., Donaldson, D., Dempsey, R.J.,

1997. The biphasic opening of the blood–brain barrier in the cortex

and hippocampus after traumatic brain injury in rats. Neurosci. Lett.

226, 33–36.

Beckman, D.L., Bean, J.W., 1970. Pulmonary pressure–volume changes

attending head injury. J. Appl. Physiol. 29, 631–636.

Bock, J.L., Wenz, B., Gupta, R.K., 1987. Studies on the mechanism of

decreased NMR-measured free magnesium in stored erythrocytes. Bio-

chim. Biophys. Acta 928, 8–12.

Bruns Jr., J., Hauser, W.A., 2003. The epidemiology of traumatic brain

injury: a review. Epilepsia 44 (Suppl. 10), 2–10.

Bullock, R., Maxwell, W.L., Graham, D.I., Teasdale, G.M., Adams, J.H.,

1991. Glial swelling following human cerebral contusion: an ultrastruc-

tural study. J. Neurol., Neurosurg. Psychiatry 54, 427–434.

Cernak, I., Chapman, S.M., Hamlin, G.P., Vink, R., 2002. Temporal char-

acterisation of pro- and anti-apoptotic mechanisms following diffuse

traumatic brain injury in rats. J. Clin. Neurosci. 9, 565–572.

Chen, Y., Constantini, S., Trembovler, V., Weinstock, M., Shohami, E.,

1996. An experimental model of closed head injury in mice: patho-

physiology, histopathology, and cognitive deficits. J. Neurotrauma 13,

557–568.

DeKosky, S.T., Kochanek, P.M., Clark, R.S., Ciallella, J.R., Dixon, C.E.,

1998. Secondary injury after head trauma: subacute and long-term

mechanisms. Semin. Clin. Neuropsychiatry 3, 176–185.

De Mulder, G., Van Rossem, K., Van Reempts, J., Borgers, M., Verlooy, J.,

2000. Validation of a closed head injury model for use in long-term

studies. Acta Neurochir., Suppl. 76, 409–413.

Elliott, K.A., Jasper, H., 1949. Measurement of experimentally induced

brain swelling and shrinkage. Am. J. Physiol. 157, 122–129.

Faden, A.I., 1993. Experimental neurobiology of central nervous system

trauma. Crit. Rev. Neurobiol. 7, 175–186.

Faden, A.I., 2001. Neuroprotection and traumatic brain injury: the search

continues. Arch. Neurol. 58, 1553–1555.

Faden, A.I., 2002. Neuroprotection and traumatic brain injury: theoretical

option or realistic proposition. Curr. Opin. Neurol. 15, 707–712.

Faden, A.I., Fox, G.B., Di, X., Knoblach, S.M., Cernak, I., Mullins, P.,

Nikolaeva, M., Kozikowski, A.P., 2003a. Neuroprotective and noo-

tropic actions of a novel cyclized dipeptide after controlled cortical

impact injury in mice. J. Cereb. Blood Flow Metab. 23, 355–363.

Faden, A.I., Knoblach, S.M., Cernak, I., Fan, L., Vink, R., Araldi, G.L.,

Fricke, S.T., Roth, B.L., Kozikowski, A.P., 2003b. Novel diketopiper-

azine enhances motor and cognitive recovery after traumatic brain in-

jury in rats and shows neuroprotection in vitro and in vivo. J. Cereb.

Blood Flow Metab. 23, 342–354.

Foda, M.A., Marmarou, A., 1994. A new model of diffuse brain injury in

rats: Part II. Morphological characterization. J. Neurosurg. 80, 301–313.

Garnett, M.R., Corkill, R.G., Blamire, A.M., Rajagopalan, B., Manners,

D.N., Young, J.D., Styles, P., Cadoux-Hudson, T.A., 2001. Altered

cellular metabolism following traumatic brain injury: a magnetic reso-

nance spectroscopy study. J. Neurotrauma 18, 231–240.

Gennarelli, T.A., 1994. Animate models of human head injury. J. Neuro-

trauma 11, 357–368.

Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I., Thompson,

C.J., Marcincin, R.P., 1982. Diffuse axonal injury and traumatic coma

in the primate. Ann. Neurol. 12, 564–574.

Goldsmith, W., 2001. The state of head injury biomechanics: past, present,

and future: Part 1. Crit. Rev. Biomed. Eng. 29, 441–600.

Gupta, R.K., Benovic, J.L., Rose, Z.B., 1978. The determination of the free

magnesium level in the human red blood cell by 31P NMR. J. Biol.

Chem. 253, 6172–6176.

Hall, E.D., 1995. The mouse head injury model: utility in the discovery

of acute cerebroprotective agents. In: Ohnishi, S.T. (Ed.), Central

Nervous System Trauma Research Techniques. CRC Press, Boca

Raton, pp. 213–233.

Hamm, R.J., Lyeth, B.G., Jenkins, L.W., O’Dell, D.M., Pike, B.R., 1993.

Selective cognitive impairment following traumatic brain injury in rats.

Behav. Brain Res. 59, 169–173.

Hanstock, C.C., Faden, A.I., Bendall, M.R., Vink, R., 1994. Diffusion-

weighted imaging differentiates ischemic tissue from traumatized tissue.

Stroke 25, 843–848.

Harris, C., DiRusso, S., Sullivan, T., Benzil, D.L., 2003. Mortality risk after

head injury increases at 30 years. J. Am. Coll. Surg. 197, 711–716.

Heath, D.L., Vink, R., 1995. Impact acceleration-induced severe diffuse

axonal injury in rats: characterization of phosphate metabolism and

neurologic outcome. J. Neurotrauma 12, 1027–1034.

Heath, D.L., Vink, R., 1996. Traumatic brain axonal injury produces sus-

tained decline in intracellular free magnesium concentration. Brain Res.

738, 150–153.

Ito, J., Marmarou, A., Barzo, P., Fatouros, P., Corwin, F., 1996. Character-

ization of edema by diffusion-weighted imaging in experimental trau-

matic brain injury. J. Neurosurg. 84, 97–103.

Kita, H., Marmarou, A., 1994. The cause of acute brain swelling after the

closed head injury in rats. Acta Neurochir. (Suppl. 60), 452–455.

Klatzo, I., 1987a. Blood–brain barrier and ischaemic brain oedema.

Z. Kardiol. 76 (Suppl. 4), 67–69.

Klatzo, I., 1987b. Pathophysiological aspects of brain edema. Acta Neuro-

pathol. (Berl.) 72, 236–239.

Knoblach, S.M., Nikolaeva, M., Huang, X., Fan, L., Krajewski, S., Reed,

J.C., Faden, A.I., 2002. Multiple caspases are activated after traumatic

brain injury: evidence for involvement in functional outcome. J. Neuro-

trauma 19, 1155–1170.

Lewelt, W., Jenkins, L.W., Miller, J.D., 1980. Autoregulation of cerebral

blood flow after experimental fluid percussion injury of the brain.

J. Neurosurg. 53, 500–511.

Li, F.H., Fisher, M., 1996. Diffusion-weighted and perfusion magnetic

resonance imaging and ischemic stroke. Drugs Today 32, 615–627.

Lighthall, J.W., Anderson, T.E., 1994. In vivo models of experimental brain

and spinal cord trauma. In: Salzman, S.K., Faden, A.I. (Eds.), The

Neurobiology of Central Nervous System Trauma. Oxford Univ. Press,

New York, pp. 3–12.

Lin, Y.J., Koretsky, A.P., 1997. Manganese ion enhances T1-weighted MRI

during brain activation: an approach to direct imaging of brain function.

Magn. Reson. Med. 38, 378–388.

Loubinoux, I., Volk, A., Borredon, J., Guirimand, S., Tiffon, B., Seylaz, J.,

Meric, P., 1997. Spreading of vasogenic edema and cytotoxic edema

assessed by quantitative diffusion and T2 magnetic resonance imaging.

Stroke 28, 419–426 (discussion 426–417).

Mahoney, E.J., Biffl, W.L., Harrington, D.T., Cioffi, W.G., 2003. Isolated

brain injury as a cause of hypotension in the blunt trauma patient.

J. Trauma 55, 1065–1069.

Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J., Kita, H.,

Demetriadou, K., 1994. A new model of diffuse brain injury in rats:

Part I. Pathophysiology and biomechanics. J. Neurosurg. 80, 291–300.


Marmarou, A., Fatouros, P.P., Barzo, P., Portella, G., Yoshihara, M., Tsuji,

O., Yamamoto, T., Laine, F., Signoretti, S., Ward, J.D., Bullock, M.R.,

Young, H.F., 2000. Contribution of edema and cerebral blood volume to

traumatic brain swelling in head-injured patients. J. Neurosurg. 93,

183–193.

Morita, H., Ogino, T., Seo, Y., Fujiki, N., Tanaka, K., Takamata, A., Naka-

mura, S., Murakami, M., 2002. Detection of hypothalamic activation by

manganese ion contrasted T(1)-weighted magnetic resonance imaging

in rats. Neurosci. Lett. 326, 101–104.

Narita, K., Kawasaki, F., Kita, H., 1990. Mn and Mg influxes through Ca

channels of motor nerve terminals are prevented by verapamil in frogs.

Brain Res. 510, 289–295.

Nilsson, B., Ponten, U., Voigt, G., 1977. Experimental head injury in the

rat. Part 1: mechanics, pathophysiology, and morphology in an impact

acceleration trauma model. J. Neurosurg. 47, 241–251.

Piper, I.R., Thomson, D., Miller, J.D., 1996. Monitoring weight drop ve-

locity and foam stiffness as an aid to quality control of a rodent model

of impact acceleration neurotrauma. J. Neurosci. Methods 69, 171–174.

Povlishock, J.T., Hayes, R.L., Michel, M.E., McIntosh, T.K., 1994. Work-

shop on animal models of traumatic brain injury. J. Neurotrauma 11,

723–732.

Raghupathi, R., Fernandez, S.C., Murai, H., Trusko, S.P., Scott, R.W.,

Nishioka, W.K., McIntosh, T.K., 1998. BCL-2 overexpression attenu-

ates cortical cell loss after traumatic brain injury in transgenic mice.

J. Cereb. Blood Flow Metab. 18, 1259–1269.

Rozsa, L., Grote, E.H., Egan, P., 1989. Traumatic brain swelling studied

by computerized tomography and densitometry. Neurosurg. Rev. 12,

133–140.

Scheff, S.W., Baldwin, S.A., Brown, R.W., Kraemer, P.J., 1997. Morris

water maze deficits in rats following traumatic brain injury: lateral

controlled cortical impact. J. Neurotrauma 14, 615–627.

Schmidt, R.H., Scholten, K.J., Maughan, P.H., 2000. Cognitive impairment

and synaptosomal choline uptake in rats following impact acceleration

injury. J. Neurotrauma 17, 1129–1139.

Shapira, Y., Shohami, E., Sidi, A., Soffer, D., Freeman, S., Cotev, S., 1988.

Experimental closed head injury in rats: mechanical, pathophysiologic,

and neurologic properties. Crit. Care Med. 16, 258–265.

Shibuya, I., Douglas, W.W., 1993. Indications from Mn-quenching of Fura-

2 fluorescence in melanotrophs that dopamine and baclofen close Ca

channels that are spontaneously open but not those opened by high

[K+]O; and that Cd preferentially blocks the latter. Cell Calcium 14,

33–44.

Siesjo, B.K., 1981. Cell damage in the brain: a speculative synthesis.

J. Cereb. Blood Flow Metab. 1, 155–185.

Smith, D.H., Chen, X.H., Xu, B.N., McIntosh, T.K., Gennarelli, T.A.,

Meaney, D.F., 1997. Characterization of diffuse axonal pathology and

selective hippocampal damage following inertial brain trauma in the

pig. J. Neuropathol. Exp. Neurol. 56, 822–834.

Smith, D.H., Cecil, K.M., Meaney, D.F., Chen, X.H., McIntosh, T.K.,

Gennarelli, T.A., Lenkinski, R.E., 1998. Magnetic resonance spec-

troscopy of diffuse brain trauma in the pig. J. Neurotrauma 15,

665–674.

Stahel, P.F., Shohami, E., Younis, F.M., Kariya, K., Otto, V.I., Lenzlinger,

P.M., Grosjean, M.B., Eugster, H.P., Trentz, O., Kossmann, T., Mor-

ganti-Kossmann, M.C., 2000. Experimental closed head injury: anal-

ysis of neurological outcome, blood – brain barrier dysfunction,

intracranial neutrophil infiltration, and neuronal cell death in mice

deficient in genes for pro-inflammatory cytokines. J. Cereb. Blood

Flow Metab. 20, 369–380.

Tanno, H., Nockels, R.P., Pitts, L.H., Noble, L.J., 1992. Breakdown of the

blood–brain barrier after fluid percussive brain injury in the rat: Part 1.

Distribution and time course of protein extravasation. J. Neurotrauma 9,

21–32.

Teasdale, G., Jennett, B., 1976. Assessment and prognosis of coma after

head injury. Acta Neurochir. (Wien.) 34, 45–55.

Teasdale, G.M., Maas, A., Iannotti, F., Ohman, J., Unterberg, A., 1999.

Challenges in translating the efficacy of neuroprotective agents in ex-

perimental models into knowledge of clinical benefits in head injured

patients. Acta Neurochir., Suppl. (Wien.) 73, 111–116.

Uyama, O., Okamura, N., Yanase, M., Narita, M., Kawabata, K., Sugita,

M., 1988. Quantitative evaluation of vascular permeability in the gerbil

brain after transient ischemia using Evans blue fluorescence. J. Cereb.

Blood Flow Metab. 8, 282–284.

Van den Heuvel, C., Blumbergs, P.C., Finnie, J.W., Manavis, J., Jones,

N.R., Reilly, P.L., Pereira, R.A., 1999. Upregulation of amyloid pre-

cursor protein messenger RNA in response to traumatic brain injury:

an ovine head impact model. Exp. Neurol. 159, 441–450.

Veech, R.L., Lawson, J.W., Cornell, N.W., Krebs, H.A., 1979. Cytosolic

phosphorylation potential. J. Biol. Chem. 254, 6538–6547.

Vink, R., Cernak, I., 2000. Regulation of intracellular free magnesium in

central nervous system injury. Front. Biosci. 5, D656–D665.

Vink, R., Cernak, I., 2003. Brain injury, investigating metabolic aspects

using magnetic resonance I. In: Adelman, G., Smith, B. (Eds.), Ency-

clopedia of Neuroscience. Elsevier, Amsterdam. CD-ROM.

Vink, R., Nimmo, A.J., 2002. Novel therapies in development for the

treatment of traumatic brain injury. Expert Opin. Investig. Drugs 11,

1375–1386.

Vink, R., Faden, A.I., McIntosh, T.K., 1988. Changes in cellular bioener-

getic state following graded traumatic brain injury in rats: determination

by phosphorus 31 magnetic resonance spectroscopy. J. Neurotrauma 5,

315–330.

Vink, R., Golding, E.M., Headrick, J.P., 1994. Bioenergetic analysis of

oxidative metabolism following traumatic brain injury in rats. J. Neuro-

trauma 11, 265–274.

Vink, R., Heath, D.L., McIntosh, T.K., 1996. Acute and prolonged alter-

ations in brain free magnesium following fluid percussion-induced

brain trauma in rats. J. Neurochem. 66, 2477–2483.

Vink, R., Golding, E.M., Williams, J.P., McIntosh, T.K., 1997. Blood glu-

cose concentration does not affect outcome in brain trauma: a 31P MRS

study. J Cereb. Blood Flow Metab. 17, 50–53.

Yakovlev, A.G., Knoblach, S.M., Fan, L., Fox, G.B., Goodnight, R., Faden,

A.I., 1997. Activation of CPP32-like caspases contributes to neuronal

apoptosis and neurological dysfunction after traumatic brain injury.

J. Neurosci. 17, 7415–7424.

Yuan, X.Q., Wade, C.E., Clifford, C.B., 1991. Immediate hypertensive

response to fluid percussion brain injury may be related to intracerebral

hemorrhage and hypothalamic damage. J. Neurotrauma 8, 219–228.

Zohar, O., Schreiber, S., Getslev, V., Schwartz, J.P., Mullins, P.G., Pick,

C.G., 2003. Closed-head minimal traumatic brain injury produces long-

term cognitive deficits in mice. Neuroscience 118, 949–955.

The pathobiology of moderate diffuse traumatic brain injury as identified using a new experimental...

Documents

Transcript of The pathobiology of moderate diffuse traumatic brain injury as identified using a new experimental...