Detection of traumatic axonal injury with diffusion tensor imaging in amouse model of traumatic brain injury
C.L. Mac Donald a, K. Dikranian b, S.K. Song c, P.V. Bayly a,d,D.M. Holtzman b,e,f,g, D.L. Brody e,g,⁎
a Department of Biomedical Engineering, Washington University, One Brookings Drive, Campus Box 1097, St. Louis, MO 63110, USAb Department of Anatomy and Neurobiology, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis, MO 63110, USA
c Department of Radiology, Washington University School of Medicine, Campus Box 8227, 4525 Scott Avenue, Room 2313, St. Louis, MO 63110, USAd Department of Mechanical Engineering and Aerospace, Washington University, One Brookings Drive, Campus Box 1185, St. Louis, MO 63110, USA
e Department of Neurology, Washington University School of Medicine, 660 S. Euclid Avenue, Campus Box 8111, St. Louis, MO 63110, USAf Department of Molecular Biology and Pharmacology, Washington University, 660 S. Euclid Avenue, Campus Box 8111, St. Louis, MO 63110, USA
g Hope Center for Neurological Disorders, Washington University, 660 S. Euclid, Campus Box 8111, St. Louis, MO 63110, USA
Received 8 September 2006; revised 22 January 2007; accepted 23 January 2007Available online 12 February 2007
Traumatic brain injury (TBI) is a major cause of acquiredbrain injury in both children and adults. There are an estimated1.5 million new cases a year in the US and currently 5.3 million
⁎ Corresponding author. Department of Neurology, Washington University,660 S. Euclid, Campus Box 8111, St. Louis, MO 63110, USA. Fax: +1 314 3622244.
Americans (2% of the US population) live with disabilitiesresulting from TBI (Thurman et al., 1999). TBI occurs whenacute, external physical forces cause injuries to the brainstructure that result in impaired brain function. Commonmechanisms of this injury include motor vehicle accidents,falls, assault, or blast injury (Adams et al., 1984; Thurman et al.,1999; Warden, 2006).
Traumatic axonal injury (TAI), also referred to as diffuseaxonal injury (DAI), is thought to be a major contributor tocognitive dysfunction in patients following TBI (Adams, 1982;
117C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
Blumbergs et al., 1994; Gennarelli et al., 1982; Grady et al.,1993; King et al., 2005; Medana and Esiri, 2003; Meythaler etal., 2001; Nevin, 1967; Oppenheimer, 1968; Pilz, 1983;Povlishock et al., 1992; Povlishock and Katz, 2005; Smith etal., 2003a,b; Strich, 1956, 1961).
Axonal injury has been characterized histologically usingsilver stains (Strich, 1961), horse-radish peroxidase uptake (Erband Povlishock, 1988; Povlishock et al., 1983), and, mostcommonly in recent years, neurofilament (Christman et al.,1994; Grady et al., 1993; Maxwell et al., 1997; Yaghmai andPovlishock, 1992) or amyloid beta precursor protein (APP)(Blumbergs et al., 1994; Gentleman et al., 1993; Sherriff et al.,1994; Stone et al., 2000) immunohistochemistry. Previousstudies have shown that the histological characterization ofaxonal injury by neurofilament and APP immunostaining serveas complementary markers of axonal injury (Marmarou et al.,2005; Stone et al., 2001). APP, which normally traverses thelength of the axon (Koo et al., 1990), accumulates at axonalretraction bulbs and varicosities in response to injury (Gentle-man et al., 1993; Lewen et al., 1995; Sherriff et al., 1994; Smithet al., 1999; Stone et al., 2000, 2001). In addition, it is thoughtthat the structural integrity of the axonal cytoskeleton breaksdown leading to increased neurofilament immunoreactivityfollowing injury (Grady et al., 1993; Maxwell et al., 1997;Povlishock and Katz, 2005; Stone et al., 2001; Yaghmai andPovlishock, 1992).
TAI is difficult to diagnose or quantify ante-mortem, andnew diagnostic methods are needed. Diffusion tensor MRimaging (DTI), has shown promise for detection of axonalinjury (Arfanakis et al., 2002; Huisman et al., 2004; Inglese etal., 2005; Mori and Zhang, 2006; Nakayama et al., 2006;Wieshmann et al., 1999; Wilde et al., 2006), but the sensitivityand specificity of DTI measurements have not been established.DTI is likely to have similar properties in humans andexperimental animals, as shown by comparisons betweenhumans with multiple sclerosis and animal models ofdemyelinating diseases (Bammer et al., 2000; Filippi et al.,2001; Kim et al., 2006; Song et al., 2005). This method hasbeen used to detect white matter pathology in experimentalanimal models of neurological diseases (Gaviria et al., 2006;Gulani et al., 2001; Nair et al., 2005; Ono et al., 1995; Song etal., 2004, 2003, 2005), including traumatic spinal cord injury(Deo et al., 2006; Nevo et al., 2001; Schwartz et al., 2005) byexploiting its sensitivity to white matter orientation. It canprovide information about brain microstructure by quantifyingisotropic and anisotropic water diffusion (Huisman et al., 2003;Neil et al., 2002; Sundgren et al., 2004). Experimentalevidence has shown that water diffusion has a directionalasymmetry (anisotropy) in organized tissues such as musclesor brain white matter (Beaulieu, 2002; Mori and van Zijl,2002). In white matter tracts, where most or all of the axonsare aligned in a parallel fashion, diffusion parallel to the axonsis greater than diffusion perpendicular to the axons. Thisparallel diffusion – termed axial diffusivity – is relativelyunhindered, whereas the perpendicular diffusion – termedradial diffusivity – is restricted by surrounding structures likethe axolemma and myelin sheath. When axonal injury occurs,
diffusion along the axon is expected to decrease; intracellulardiffusion may be hindered by membranes after severed axonsreseal, and extracellular diffusion may be hindered by largeaxon retraction balls that compress the extracellular spacesaround axons. In addition, diffusion might also increase per-pendicular to the axon if the axolemma and/or myelin sheath arecompromised.
In contrast to DTI, conventional diffusion weighted imaging(DWI) uses only diffusivity averaged across all directions. Thistechnique is sensitive to cerebral infarctions, certain infectiousprocesses such as abscesses, and other pathologies that affectthe global diffusion of water. It is commonly used as an imagingtechnique for TBI patients (Huisman et al., 2003) to assess forthe presence of cerebral ischemia (Le Bihan et al., 1986; Warachet al., 1992). It may not be as sensitive as DTI to axonal injury inwhite matter that does not result in infarction (Arfanakis et al.,2002; Huisman et al., 2004). Specifically, elements of thediffusion tensor may change in response to injury in such a waythat average diffusivity changes very little. For example, adecrease in axial diffusivity accompanied by increased radialdiffusivity would result in marked reduction in the anisotropy ofdiffusion, but little change in average diffusivity. This sort ofchange is especially likely to occur in white matter or othertissues with elevated baseline anisotropy due to highlystructured intrinsic directionality. These considerations raisethe prospect of more sensitive axonal injury detection comparedwith traditional methods of imaging, such as CT andconventional MRI (i.e. DWI, T1 and T2 imaging), whichassess mainly hemorrhage and edema (Arfanakis et al., 2002;Gupta et al., 2005; Neil et al., 2002; Pierpaoli et al., 1996; Rugg-Gunn et al., 2001; Wieshmann et al., 1999).
In theory, DT imaging can provide microstructural andarchitectural information about axons, but it is unknownwhether this technique can be used to establish patterns ofaxonal injury in practice (Arfanakis et al., 2002; Le Bihan,2003). Decreased diffusion along the axonal fiber tract andincreased diffusion perpendicular to the fiber tract have beenobserved experimentally (Song et al., 2003). There is reducedanisotropy in some human patients with TBI (Arfanakis et al.,2002; Gupta et al., 2005; Huisman et al., 2004; Inglese et al.,2005; Rugg-Gunn et al., 2001; Wieshmann et al., 1999), butthere have been no studies to our knowledge that directlycorrelate MRI abnormalities with histologically verified TAI.Thus, we sought to establish the MRI correlates of traumaticaxonal injury in an animal model where direct histologicalevaluation could be performed.
We hypothesized that regions of reduced relative anisotropy(RA) seen in DTI images would correspond to areas ofhistologically verified traumatic axonal injury. While techni-cally challenging to image, mice were chosen due to thepotential for evaluation of the effects of various human genes inthe response to and recovery from TBI (Longhi et al., 2001;Sabo et al., 2000; Teasdale et al., 2005). In support of ourhypothesis, we found statistically significant decreases in RAmainly due to the decreases in axial diffusivity (AD) in regionswith TAI, relative to controls. These regions of TAI were notdetected with conventional MRI.
118 C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
Materials and methods
Experimental design
For these experiments, male and female, 8–10 week oldB6SJL F1 mice (N=10, Jackson Labs, Bar Harbor, Maine) wereutilized. Mice were imaged before injury using DTI so that eachmouse could serve as its own control. Following the controlimage acquisition, a moderately severe controlled corticalimpact TBI was performed under general anesthesia. Mice werere-imaged 4–6 h and/or 24 h after injury using the same imagingparameters as in the pre-injury imaging session. Immediatelyafter post-TBI imaging was completed at each time point,groups of mice were sacrificed and their brains examined usingAPP and neurofilament immunohistochemistry. An additionaluninjured control group was sacrificed for histological compar-ison (N=5). All procedures involving animals were approvedby Washington University's Animal Studies Committee and areconsistent with NIH guidelines for the care and use of animals.
Diffusion tensor imaging (DTI)
Diffusion can be evaluated by measuring the signal intensityattenuation (I) as a function of the gradient factor or b-value(Beaulieu, 2002).
I ¼ I0e�bD ð1Þ
In this equation, b is a ffunction of (i) the magnetogyric ratioof the nucleus of interest, (ii) gradient strength and (iii) timingparameters of the diffusion-sensitizing gradients. The value D isthe apparent diffusion coefficient in the direction of thediffusion sensitizing gradient. Gradients were applied in sixdirections in order to measure the six independent elements ofthe diffusion tensor. This allows for the calculation of thefollowing DTI parameters; axial diffusivity, radial diffusivityand relative anisotropy. Axial diffusivity (AD) denotes theextent of diffusion in the direction of maximal diffusivity. Inwhite matter this direction is typically parallel to the orientationof the axons and is the first eigenvalue in the diffusion tensor.
AD ¼ k1 ð2ÞRadial diffusivity (RD) is defined as the extent of diffusion
perpendicular to the direction of maximal diffusivity. This isassumed to include diffusion through the axolemma and myelin,as well as through intracellular and extracellular spaceperpendicular to the predominant orientation of the axons. RDis calculated from the average of the second and thirdeigenvalues in the diffusion tensor.
RD ¼ ðk2 þ k3Þ2
ð3Þ
The apparent diffusion coefficient (ADC), or trace is anoverall measure of diffusion and is the mean diffusivity of thethree principle eigenvalues.
hDi ¼ ðk1 þ k2 þ k3Þ3
ð4Þ
Relative anisotropy (RA) is a measure of the directionalasymmetry of diffusion. It is defined as the standard deviation ofthe eigenvalues normalized by the mean of the diffusivity ⟨D⟩.
To acquire DTI images, mice were anesthetized withisoflurane (5% induction and 0.5–1% maintenance). Followingthe methods of Song et al. (2003) they were placed in a 4.7Tscanner (Oxford Instruments 200/330) with an actively shieldedgradient coil (180 mT/m, 400 μs rise time), which interfaceswith a Varian UNITY-INOVA console controlled by SunMicrosystems Ultra-60 Sparc workstation. For imaging prior toinjury, anesthesia was maintained at 1% isoflurane via a nosecone. For imaging post-injury, mice were re-anesthetized with alow dose of isoflurane (0.5–0.7%). A lower dose of isofluranewas used to re-anesthetize because the animals were moresensitive to the anesthesia post-injury. The animals were placedin an MR-compatible stereotaxic frame in the scanner.Respiration was monitored with two fiber optic cables thatdetected the chest motion of the animal. The signal was fed intoan oscilloscope that displayed the chest motion for monitoringthroughout the experiment (Garbow et al., 2004). Bodytemperature was maintained by circulating warm water intubing surrounding the animal. Anatomical landmarks wereused to center and align the brain in each image plane. Identicalmarkers were used both before and after injury, in order toobtain anatomically matched coronal image slices at both timepoints.
A multi-slice, spin-echo imaging sequence, modified toinclude the Stejskal-Tanner diffusion sensitizing gradient pair,was used to acquire the diffusion weighted images (Song et al.,2004). These images were acquired with a repetition period(TR) of 3 s, spin-echo time (TE) 43 ms, time betweenapplication of gradient pulses (Δ) 25 ms, diffusion gradientduration (δ) 10 ms, slice thickness 0.5 mm, field-of-view2.0 cm, data matrix 128×128 (zero filled to 256×256).Diffusion sensitizing gradients were then applied along sixdirections: [Gx,Gy,Gz]=[1,1,0], [1,0,1], [0,1,1], [−1,1,0], [0,−1,1], and [1,0,−1]. The two diffusion sensitizing factors or bvalues used for acquisition of the diffusion weighted series were0 and 764 s/mm2. This resulted in a voxel size of78 μm×78 μm×0.5 mm after zero fill. This voxel size waschosen to obtain high resolution images allowing visualizationof the boundaries between gray and white matter on coronalslices (Figs. 2 and 3). Total imaging time was ∼3 h per animal.
Experimental traumatic brain injury
After the initial DTI scan, mice received 5% isoflurane in air(inhalation to effect, administered in an induction chamber forapproximately 60 s), and were placed in a stereotactic frame(MyNeuroLab, St. Louis, MO). Anesthesia was maintainedwith 2.5% isoflurane via a nose cone. Eye lubricant was applied
119C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
to both eyes to decrease potential for drying during the surgery.The heads of the mice were shaved and the surgical area wasswabbed with betadine. A midline incision was made and thescalp was reflected to expose the skull. A 5 mm left lateralcraniotomy was performed using a motorized drill mounted tothe stereotactic arm. A low voltage DC circuit touch detector(custom built by the Washington University Electronics Shop)was used to determine when the tip first contacted the dura. Theimpactor tip was then mechanically elevated away from thebrain, and the desired impact depth was digitally set. Acontrolled cortical impact injury was produced with anintegrated electromagnetic impact accessory (Brody et al.,2006; Kessons et al., 2005). The impact was centered at 2.7 mmlateral from midline and 3 mm anterior from lambda. Theimpactor tip diameter was 3 mm and the depth and velocitywere 2.5 mm and 5 m/s respectively. This produces amoderately severe injury with pronounced behavioral deficits,but virtually no mortality (Brody et al., 2006). After injury, aplastic skull cap was secured over the impact site with Vetbondadhesive and the skin incision was sutured closed. The closedsuture area was swabbed with vetropolycin, an antibioticointment, and the animals were allowed to fully recover.
Fig. 1. Axonal injury following experimental controlled cortical impact TBI. APP imwhite matter. (B,E) At 24 h post-injury there were numerous APP stained axonal vaepicenter of injury, there is significant tissue loss and APP stained varicosities in the r“beads on a string” appearance of injured axons running parallel to the plane of the seno neurofilament light chain immunostaining in corpus callosum. (H) Neurofilamentwhite matter surrounding the epicenter, areas of neurofilament staining were also pr
This impact depth was chosen based on previous experi-ments conducted in our laboratory which characterized theelectromagnetic impactor, touch detector, and stereotaxicsystem used for these experiments. The 2.5 mm impacts createdlesions that were similar histologically to those produced by1.5 mm impacts using a pneumatic impactor system. Specifi-cally, differences between methods arise because of reducedmechanical overshoot, earlier contact with the cortical surfaceduring alignment of the impactor tip, and slight protrusion of thecortex due to the use of ring head holders (Brody et al., Journalof Neurotrauma, in press). This injury produces a contusion inthe cortex and hippocampus directly underlying the impactor tip(Fig. 1C), and pericontusional white matter injury in surround-ing regions of corpus callosum and external capsule (Figs. 1, 2and 6).
Histology
After images were acquired post-trauma, animals wereimmediately sacrificed for histology. Animals received anoverdose of pentobarbital and were perfused transcardiacallywith ice-cold phosphated-buffered-saline+0.3% heparin. After
munohistochemistry (A–F). (A,D) Uninjured tissue shows no APP staining inricosities in the corpus callosum rostral to the epicenter of injury. (C,F) At theemaining white matter at the edges of the contusion. Many had the characteristicction. Neurofilament immunohistochemistry (G–I). (G) Uninjured tissue showsstaining rostral to the epicenter within the corpus callosum. (I) In the remainingesent.
Fig. 2. Definition of regions of interest containing histologically verified axonal injury. A region of interest within the corpus callosum and external capsule containingAPP stained axons was described. Boundaries were noted for these regions and these boundaries were used for quantitative analysis of the DTI image sets. In theexample shown, the midline was used as one boundary and a horizontal line extending laterally from the inferior edge of the fimbria was used as the other boundary.Similar boundaries for nine rostral to caudal sections were applied to the DTI data sets for quantitative analysis (vhc: ventral hippocampal commissure, fi: fimbria).Center panel adopted from Franklin and Paxinos (1997).
120 C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
sacrifice, brain tissue was fixed in 4% paraformaldehyde for24 h and then equilibrated in 30% sucrose. Serial coronal slices50 μm thick were cut on a freezing microtome. Every sixthsection was used for histology. Floating tissue sections werewashed 3× in Tris-buffered saline (TBS) for 5 min each and wasthen permeablized in TBS-X (containing 0.025% Triton X) for15 min. Sections were incubated with 0.3% H2O2 in TBS for10 min at room temperature to block endogenous peroxidase.Following the incubation, sections were rinsed in TBS 3× for5 min each, and then the tissue was blocked with 3% normalgoat serum in TBS-X for 30 min at room temperature. It wasfurther blocked with 1% goat serum in TBS-X and incubatedwith either polyclonal rabbit anti-beta-APP (Invitrogen, Carls-bad, CA), in a 1:500 dilution or monoclonal mouse anti-neurofilament 68 clone NR4 (Sigma, St. Louis, MO) in a 1:100dilution overnight at 4 °C. The protocol was optimized usingmultiple dilutions, two different APP antibodies, 3 different
neurofilament antibodies and several antigen retrieval techni-ques. The following day, sections were washed 3× in TBS for5 min and incubated with a biotinylated secondary goat anti-Rabbit antibody (beta-APP) or goat anti-mouse IgG (neurofila-ment) for 1 h at room temperature in a 1:1000 dilution (VectorLaboratories, Burlingame, CA). Following the application ofthe secondary antibody, the sections were washed 3× in TBS for5 min each, incubated with ABC Elite (Vector Laboratories,Burlingame, CA) at a 1:400 dilution in TBS for 1 h at roomtemperature, then washed with TBS 3× for 5 min and developedwith 3,3′-Diaminobenzidine (DAB) tablets (Sigma, St. Louis,MO). A final series of three TBS washes were preformed for5 min each. The tissue was mounted on Superfrost-Plusmicroscope slides (Fisher, Houston, TX) and allowed to dry.Once dry, the slides were dipped for 1 min each in 50–70–95–95–100% ethanol solutions, followed by two treatments for4 min each in Xylene. Finally the slides were coverslipped with
121C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
Cytoseal 60 (Richard-Allan Scientific, Kalamazoo, MI) andallowed to dry. Positive APP staining or positive neurofilamentstaining was visualized with a light microscope (Nikon EclipseE800) and digital images were acquired for comparison withDTI.
Histological definition of a region of interest (ROI) containingaxonal injury in corpus callosum and external capsule
The anatomical extent of axonal injury in the corpuscallosum and external capsule was described based on theAPP stained histological slides. The regions containing axonalinjury were not well captured by standard anatomical defini-tions, so specific ROI boundaries were determined for eachcoronal slice. Specifically, the ROI was defined as follows: Inthe most rostral regions displaying a complete corpus callosum(Bregma +1.10 mm to −0.34 mm) (Franklin and Paxinos,1997), the boundaries were the midline and the lateral edge ofthe cingulum. Moving more caudally (Bregma −0.34 mm to−1.06 mm), the ROI included the corpus callosum and externalcapsule from midline to a boundary defined by a horizontal lineextending laterally from the bottom of the fimbria until itintersected with the external capsule (see Fig. 2). The ROI in themost caudal slices (Bregma −1.06 mm to −3.08 mm) includedthe corpus callosum and external capsule from midline to aboundary defined by a similar horizontal line extending laterallyfrom the lateral-inferior edge of the hippocampus. We verifiedthat all chosen anatomical landmarks were clearly visible onMR images, and all determinations of the ROI were made byblinded observers.
Analysis of DTI images using anatomically defined regions ofinterest
Analysis was performed with ImageJ software (NIH). TheROI was traced on both the ipsilateral and contralateral sidethrough nine coronal slices for each control and post-TBI image
Fig. 3. MRI signal characteristics in control and trauma groups. Grey scale imaganisotropy, greater diffusivity, or longer relaxation time). The region of interest ingradient of signal changes within the ROI on post-TBI images. Radial diffusivity, tracis similar to that of the control images. Examples shown for illustrative purposes, no1997).
set. For the post-TBI image sets, the ROI was retraced using thesame anatomical landmark-based rules as used for the controlimages. These post-TBI ROI's were not necessarily identical inshape to the control ROI's, as the anatomical landmarks andboundaries between gray and white matter were often slightlydistorted by the injury (Figs. 2, 3, and 8). ROI tracing wasperformed using the T2 images. The ImageJ software bindingfeature allowed for the simultaneous replication of the tracedROI to the other imaging modalities (RA, AD, RD, and TR). Atthe epicenter of the injury, there was disruption of cortex, whitematter tracts, and hippocampus; signal abnormality wasapparent on all of the imaging modalities. The epicenter wasomitted from the ROI and only the remaining visible whitematter was traced. The ImageJ program returned the averagesignal intensity for the traced sub-region of interest on eachslice. The average signal intensity for the complete ROI wasobtained by calculating the average across all slices, weightedby the fraction of total pixels in each slice.
A portion of the images were traced again 10 days after theinitial tracing was performed to determine intra-observeragreement. All parameters were compared (i.e. RA, AD, RD,trace, T2) between the first and second tracings. Thequantitative data from of each set of tracings of the same regionexhibited a 98% correlation (r2 =0.979). It was concluded thatthe ROI tracing method did not add significant variance.
Stereological quantification of axonal injury as defined by APPimmunostaining
A stereological method was employed to quantify thenumbers of APP stained axons per cubic mm in the regions ofinterest. This method utilizes the techniques of unbiasedestimation of the numbers of a particular type of discreteobject, in this case APP-stained, injured axons, in a definedthree dimensional volume (Sterio, 1984). This analysis wasperformed using StereoInvestigator 6.0 (Microbrightfield) and aNikon Eclipse E800 microscope. Gunderson coefficient of error
es of signal intensity: lighter shading indicates elevated signal (i.e. increasedeach panel is outlined in red. Relative anisotropy and axial diffusivity show ae, and T2 images show homogeneous signal throughout the region of interest thatt necessarily from the same mouse (Bregma +0.26 mm, Franklin and Paxinos,
122 C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
(m=1) was maintained below 0.05 for each set. Counts wereperformed by a blinded observer. The optical fractionatortechnique was employed to count a systematic random sampleof positively stained axons over the entire rostral to caudalextent of the regions of interest. For each coronal slice fromeach animal, the ROI was outlined at low power (4×) followedby systematic counts of individual injured axons at high power(60×: oil immersion) over sites within the ROI randomly chosenby the StereoInvestigator software. A 40×40 μm countingframe was used (Fig. 5A) and a 15 μm deep region wassampled. Injured axons were defined by the presence of APPstained varicosities that were 8 μm in diameter or greater andwere in focus within the 15 μm deep counting region, beyondthe 2.5 μm guard zone. Red blood cells and other non-axonalstained structures were not counted.
The volumes of the complete regions of interest werecalculated using the Cavalieri principle; the area of the region ofinterest in each sampled slice was multiplied by the distancebetween sampled slices and the resultant volumes weresummed. The total volume of the sampled regions wascalculated by multiplying the area of each counting frame(1600 μm2) by the thickness of the sampled region (15 μm), andthen multiplying by the number of counting frames sampled,which varied from animal to animal. To estimate the totalnumber of injured axons within the complete region of interest,the number of injured axons counted was multiplied by thevolume of the region of interest and divided by the volumesampled. To estimate the number of injured axons per cubicmm, the estimate of total injured axons was divided by thevolume of the sampled region.
MRI-based definition of secondary regions of interest
A second set of regions of interest were chosen to determinewhether DTI changes could predict regions of axonal injury, asdefined by APP and neurofilament immunostaining. Thehippocampal commissure and anterior commissure were chosenfor this purpose. The boundaries of these ROI's were based on astandard atlas (Bregma +0.38 mm for anterior commissure, andBregma −0.70 mm for hippocampal commissure; Franklin andPaxinos, 1997).
Statistical methods
All data was analyzed using Statistica 6.0 (StatSoft).Quantitative results from corresponding ROI's were comparedbetween the control and post-trauma image sets. Paired T-testswere used as there was no evidence for deviation from thenormal distribution (Shapiro–Wilks W-test) for any of theparameters. The 4–6 h post-trauma and 24 h post-trauma datawere pooled as there were no significant differences in any ofthe MRI or histological parameters. Prespecified hypotheseswere that RA and AD would decrease after TBI, so one-sidedpaired T-tests were used for these two parameters. For RD,Trace, and T2, there were no prespecified hypotheses about thedirection of change, so two-sided, paired T-tests were employed.The threshold for statistical significance was set to p<0.05
without correction for multiple comparisons. For comparison ofthe extent of the axonal injury from subregion to subregion, aone-way ANOVAwas used followed by a Fisher post-hoc test tocorrect for multiple comparisons.
Results
TAI was detectable following CCI injury in the corpuscallosum of adult mice (Fig. 1). TAI was identified using APPand neurofilament immunohistochemistry. No APP staining inwhite matter was observed in uninjured animals (Figs. 1A, D).No staining was observed when the primary antibody wasomitted (not shown). We examined the morphology of the APPstained structures under high magnification (Figs. 1E, F).Oblong and round structures were noted, corresponding to theaccumulation of APP in axonal varicosities. These structureswere confirmed in all areas of positive staining with APPantibody in the damaged white matter. These findings wereconsistent with previous descriptions of traumatic axonal injury(Gentleman et al., 1993; Lewen et al., 1995; Sherriff et al.,1994). CCI produced a large area of necrosis and tissue loss thatwas readily apparent histologically in mice sacrificed 4–6 h and24 h after injury (Fig. 1C). In the remaining white matter tracts,adjacent to the epicenter of the necrosis, APP staining wasclearly present (Fig. 1F). A gradient of axonal injury wasobserved, indicated by positively stained injured axons thatwere most frequent directly adjacent to the epicenter, anddecreased in numbers with increasing distance away from theepicenter. White matter tracts distal to the site of injury did notshow APP staining (not shown).
For further verification of axonal injury, adjacent sections werestained with the anti-neurofilament light chain antibody, NR4(Figs. 1G, H, I). As with APP staining, the neurofilament stainingwasmost dense adjacent to the epicenter and then tapered off withdistance away from the epicenter. Positive staining for NR4 wasnoted in a similar distribution to APP immunostaining (Figs. 1H,I). There typically was no staining contralateral to the injury site orin uninjured tissue. These findings are consistent with previousdescriptions of axonal injury as defined by neurofilament lightchain immunostaining (Christman et al., 1994; Grady et al., 1993;Marmarou and Povlishock, 2006; Marmarou et al., 2005;Yaghmai and Povlishock, 1992).
The spatial extent of the axonal injury in the corpus callosumand external capsule was described using anatomical para-meters, which would then be used to define the ROI's for DTIanalysis (Fig. 2). The anatomical regions of interest containingTAI were defined in such a way as to capture approximately90% of the corpus callosum and external capsule containingconsistently identifiable APP stained axonal varicosities. In theexample shown in Fig. 2, the midline was used as one boundaryand a horizontal line extending laterally from the inferior edgeof the fimbria was used as the other boundary. Similarboundaries for nine rostral to caudal sections were then appliedto the DTI data sets for quantitative analysis. Overall, thepattern of APP staining was consistent between animals,allowing the same boundary rules to be used across all micein the study (see Materials and methods).
123C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
Several DTI (RA, AD, and RD) and conventional MRI(trace, T2 relaxation) parameters were analyzed for each ROI.White matter in the control images had high RA, as would beexpected for anisotropic tissue, whereas there was a reduction inRA in the post-TBI images (Fig. 3). The RA and AD imagesrevealed abnormalities in white matter that were not apparent inthe other image sets (Fig. 3). RA images contained a gradient ofreduced signal, most severe adjacent to the epicenter of theinjury, and normalizing with increased distance away from theepicenter. This was similar to the gradient of axonal injuryobserved histologically. The RD, trace, and T2 images showedno apparent signal change after TBI compared to controlimages. Outside the ROI in the cortex and hippocampus withinthe epicenter of injury, there were clearly visible signalabnormalities on all imaging modalities, corresponding tohistologically defined necrosis and tissue loss (not shown).
Fig. 4. Quantitative analysis of MR imaging parameters. (A) Average values for DTinjured white matter in corpus callosum and external capsule. Differences betweendiffusivity (AD) were statistically significant (p=0.00001 RA, p=0.005 AD, one-si(p=0.874), DWI-trace (p=0.273), and T2 relaxation (p=0.797). Differences betweenRA (p=0.00002) and AD (p=0.000035). Error bars denote standard deviations. (B) Gdecreasing trend in RA for each mouse post-injury. (C) Scatter plots of RA values froROI's ipsilateral to the injury. No differences between ipsilateral and contralateral vasimilar to control, with the exception of data from 2 mice. Upon histological re-exa
Only the remaining white matter outside of the epicenter wasanalyzed, as a primary goal of this study was to determinewhether DTI could detect changes associated with axonalinjury.
Quantitative analysis of the mean values of MRI parameterswithin ROI was performed (Fig. 4). The relative anisotropy(p=0.00001) and the axial diffusivity (p=0.005) were sig-nificantly reduced after injury compared to control (1-sided,paired T-tests). In contrast, conventional MRI imaging (i.e.trace, T2 relaxation) showed no statistically significantdifference between the control and post-injury images in theROI containing axonal injury. There were no significantchanges in radial diffusivity following TBI (Fig. 4A). Foreach mouse scanned before and after TBI, there was a consistentdecrease in RA following trauma (Fig. 4B). A scatter plot of thedata for each mouse shows the separation of the RA values
I and conventional MRI parameters within the region of interest encompassingipsilateral TBI and ipsilateral control for relative anisotropy (RA) and axial
ded, paired T-tests) whereas there were no difference for radial diffusivity (RD)the ipsilateral and contralateral sides of the TBI groups were also significant forraph of RAvalues acquired for mice scanned before and after TBI. There was am each mouse. There was no overlap of RA values between injured and controllues were seen in controls. RA in the contralateral ROI's after TBI overall wasmination these two brains were found to have some contralateral axonal injury.
Fig. 5. Quantitative relationship between DTI signal changes and histologically defined axonal injury severity. (A) Counting frame used for stereological estimation ofthe numbers of APP stained, injured axons. APP stained axonal varicosities (marked in yellow) within the computer-generated, systematic random sampling zoneswere counted. Positively stained injured axonal varicosities were not counted if they touched the red boundary and were counted if they were centered within the greenboundary. Faintly stained punctae less than 8 μm in diameter were not counted. (B) Correlation of normalized relative anisotropy with the estimated numbers of APP-stained, injured axons per mm3. Each symbol represents 1 mouse. The values of RA acquired following trauma were normalized by dividing by the mean ipsilateral RAvalue acquired during the control scans. Estimates of the number of APP-stained axons per mm3 were obtained by dividing the total number of counted APP-stainedaxonal varicosities by the total volume of the counted sampling zones. A strong correlation was found between the change in relative anisotropy and the severity ofaxonal injury as defined by APP immunostaining. Dashed lines represent the 95% confidence band.
124 C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
ipsilateral to the site of injury in comparison to the contralateralside and to control (Fig. 4C). There was no overlap of valuesbetween control and post-TBI data sets ipsilaterally. In two mice
Fig. 6. Spatial gradient of DTI signal change and histologically-defined axonal injury.callosum (right; subregions outlined in red) and equivalent subregions of uninjured coshows a marked decrease in RA signal relative to the contralateral side in several subthough still elevated, RA values more laterally. Middle: APP stained histological semagnification (60×) images demonstrating the spatial gradient in the numbers of AP
there was a reduction in RA in the contralateral ROI followingTBI. On careful re-examination of the histology from these twobrains; APP immunostaining was apparent on the contralateral
Top: DTI image showing a spatial gradient of RA signal change in injured corpusntralateral corpus callosum (left; subregions outlined in blue). The ipsilateral sideregions. The contralateral side shows very high RA values medially and lower,ction from approximately equivalent region as the DTI image. Bottom: HigherP stained axonal varicosities.
Fig. 7. Quantitative analysis of the spatial gradients of relative anisotropy signalchanges and histologically-defined axonal injury. (A) Relative anisotropy as afunction of subregions, as defined in Fig. 6. There was a significant reduction inrelative anisotropy ipsilaterally in subregions two and three when compared tothe contralateral subregions (Student's t-test, p<0.05). (B) Number of APPstained axons per mm3 on the ipsilateral side as a function of subregion. Therewere significant differences between subregions, specifically; subregions two,three, and four were found to be significantly different than subregions one andfive (p<0.05) (one-way ANOVA p=0.00008, F4,40=7.97 followed by Fisherpost-hoc analysis). (C) Correlation between normalized relative anisotropy andnumbers of APP stained axons per mm3 was significant (p=0.0021). Dashedlines represent 95% confidence band.
125C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
side, whereas such staining was absent in the other brains. Therewere no significant differences between the ipsilateral andcontralateral sides of the control data sets. These resultsprovided initial support for the hypothesis that DTI may becapable of detecting histologically-verified axonal injury, andthat this technique may be more sensitive for this purpose thanconventional MRI.
Next, the changes in DTI parameters were correlated withthe number of APP immunostained axons per cubic mm withinthe ROI for each mouse. Changes in RA and AD wereexpressed using normalized relative anisotropy and normalizedaxial diffusivity. These were calculated by dividing RA or ADin the injured ROI by the mean RA or AD in the ipsilateralROI from the control group. There was a strong correlationbetween changes in RA and the numbers of APP stained axonsper cubic mm (r2 =0.7895, p=0.0014; Fig. 5). The data wastightly clustered around the regression line, such that the 95%confidence bands were narrow. The correlation betweennormalized AD and APP stained axons per cubic mm wasalso significant (r2 =0.551, p=0.03; not shown) although notas strong.
Given this result, we further hypothesized that there wouldbe a correlation between the extent of RA signal change and thenumber of APP stained axons per cubic mm within subregionsspanning the spatial gradient of observed injury. To test thishypothesis, the corpus callosum in a single coronal section atthe level of the early rostral presentation of the septofimbrialnucleus was subsectioned into 5 similarly sized regions (Fig. 6).APP stained axons were counted separately in each subregionipsilaterally and contralaterally. These subregions were appliedto the DTI data set for analysis (Fig. 6).
Quantitative analysis of the subregions (Fig. 7) revealed thattwo of the central subregions showed significant decreases inRA when compared to the equivalent subregions on thecontralateral side (Fig. 7A). Stereological counts of thenumbers of APP stained axon per cubic mm showedsignificant increases in subregions two, three, and four (Fig.7B, one-way ANOVA p=0.00008, F4,40=7.97). In this coronalslice, there was essentially no contralateral injury or RA signalabnormality, even in the two brains with decreased RA andaxonal injury in the complete ROI noted above (Fig. 4B). Thecorrelation between the normalized RA values and the numbersof APP stained axons per cubic mm in these subregions wasmodest but statistically significant (p=0.0021). The correlationwas not as strong as that seen using the entire ROI(r2 =0.7895, Fig. 5), possibly due to imperfect anatomicalcoregistration of the data sets at the level of the individualsubregions (see Discussion). Overall, these quantitativecorrelations strengthen the evidence that DTI is capable ofdetecting axonal injury.
In the previous analyses, we used regions of interest definedby histologically-verified axonal injury and investigated thecorresponding MRI signal characteristics. We next performedthe converse analysis and tested the ability of DTI signalabnormalities to predict the presence of histologically-definedaxonal injury (Figs. 8 and 9). We first analyzed the hippocampalcommissure, as there were clear abnormalities on DTI following
Fig. 8. DTI-based prediction of axonal injury in the hippocampal commissure. (A) DTI images demonstrating reduced RA in the hippocampal commissure followingtrauma. As predicted, anatomically equivalent areas of APP stained tissue show strong evidence of axonal injury in this region. (B) Quantitative analysis of MRI signalcharacteristics. There was a significant decrease in relative anisotropy following injury in both the ipsilateral and contralateral hippocampal commissure relative tocontrol. Axial diffusivity was reduced more prominently in the ipsilateral hippocampal commissure than on the contralateral side. Additionally there was a significantincrease in radial diffusivity on both the ipsilateral and contralateral sides relative to control. In contrast, conventional imaging methods (i.e. trace, T2 relaxation)showed no significant changes.
126 C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
TBI in this region (Fig. 8A). Specifically, there were significantdecreases in RA and AD, with a slight increase in RD in theipsilateral hippocampal commissure following TBI comparedto control (Fig. 8B). There was also a significant decrease inRA and a significant increase in RD in the contralateralhippocampal commissure. Additionally, there was a statisti-cally significant decrease in RA ipsilaterally compared tocontralaterally in the injured group. There were no significantdifferences between ipsilateral and contralateral sides of thecontrol group (not shown). Based on these findings and on our
previous results, we predicted that there would be axonalinjury to both the ipsilateral and contralateral sides of thehippocampal commissure with greater damage on the ipsilat-eral side. This prediction was confirmed histologically; uponevaluation of the hippocampal commissure region there wasdense APP staining ipsilaterally, which extended into thecontralateral side where it was less dense (Fig. 8A, bottomright panel) as well as neurofilament staining (not shown). NoAPP or neurofilament staining was observed in the uninjuredhippocampal commissure.
Fig. 9. DTI-based prediction of normal histology in the anterior commissure following TBI. (A) DTI images showing RA in the anterior commissure. The APP stainedregions show no positive staining that would indicate injury (4×; inset 10×). (B) There were no significant changes in any of the DTI or conventional MRI parametersin the anterior commissure following trauma.
127C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
We also analyzed the anterior commissure, which in contrastappeared normal on DTI after injury (Fig. 9). In this region,there were no significant changes when comparing bothipsilateral and contralateral sides of the post-TBI and controlscans (Fig. 9). Based on this result, we predicted that therewould be no axonal injury in the anterior commissure on boththe ipsilateral and contralateral sides following trauma. Againthis prediction was confirmed, as there were no positively
stained axons using APP or neurofilament immunohistochem-istry in the anterior commissure in either the trauma or controlgroups (Fig. 9). Thus, DTI signal changes could be used topredict areas of axonal injury. In contrast, conventional MRIrevealed similar characteristics in injured and uninjuredhippocampal commissure and did not distinguish an injuredregion like the hippocampal commissure from an uninjured one,like the anterior commissure.
128 C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
Discussion
In summary, we found that histologically-verified traumaticaxonal injury could be detected using diffusion tensor imagingin a mouse model of TBI. Likewise, DTI could also be used topredict areas of axonal injury. Two DTI parameters, relativeanisotropy and axial diffusivity, showed statistically significantdecreases after trauma in areas of injured white matter. Of theparameters studied, changes in relative anisotropy correlatedbest with the density of APP stained injured axons. Parametersassociated with conventional MR imaging showed no sig-nificant changes in the same regions; DWI-trace and T2relaxation were sensitive to effects related to necrosis at theepicenter of the lesion, but did not reflect the more subtlechanges associated with axonal injury. This indicates that DTI islikely to be more sensitive to TAI than conventional MRI.
Axonal injury was confirmed in the peri-contusional regionsfollowing controlled cortical impact in the mouse as evidencedby APP and neurofilament immunostaining. This has also beendescribed recently by others using de Olmos silver stainingfollowing injury (Hall et al., 2005). This description of axonalinjury following controlled cortical impact in the mouse isimportant on its own, as the use of mice may make it possible tostudy the effects of relevant human genes on axonal injuryseverity and recovery from axonal injury using transgenic andknockout animals (Longhi et al., 2001; Sabo et al., 2000).
The experimental traumatic brain injuries used in the currentstudy are in many respects different from those seen in humanclinical TBI. The controlled cortical impact model produces acentral contusion and surrounding pericontusional axonalinjury, whereas much of the traumatic axonal injury in humansoccurs in a scattered, multifocal distribution. The axonal injuryin our mouse model may be most directly analogous to the well-described subset of peri-contusional white matter lesions seen inhuman TBI patients (Cervos-Navarro and Lafuente, 1991;Strich, 1961). While the mechanisms of injury differ (directcortical impact in the mice vs. deformation of the brain withinthe skull in humans), the characteristics of the pericontusionalwhite matter injury in mice and humans are quite similar overall(Cervos-Navarro and Lafuente, 1991; Strich, 1961). In addition,the histological features of these pericontusional injuries arecomparable to those seen in white matter that is not immediatelyadjacent to contusions (Cervos-Navarro and Lafuente, 1991),and the signal abnormalities detected using DTI in our study areconcordant with those seen in such injuries (Arfanakis et al.,2002; Huisman et al., 2004; Inglese et al., 2005). This suggeststhat our findings may be applicable to both peri-contusional andnon peri-contusional white matter injuries. Of course, thisshould not be assumed to be true, and could be testedempirically. Specifically, the DTI signal characteristics ofwhite matter lesions that are adjacent to contusions could becompared with the DTI characteristics of white matter lesionsdistant from contusion injuries in human TBI patients.Likewise, the DTI signal characteristics of non-contusionaltraumatic white matter injury in animal models (Gennarelli etal., 1982; Gennarelli et al., 1989; Smith et al., 1999) could beassessed and correlated with histological findings.
The recent development of high-resolution DTI imagingtechniques makes work in animals with small white mattertracts, such as mice, technically feasible. The structure of thecorpus callosum and external capsule in mice is such that highresolution images within each coronal plane are needed toclearly visualize the boundaries between gray and white matter.In contrast, these white matter regions are anatomicallycontinuous and relatively homogenous throughout the ante-rior–posterior extent of the forebrain, so the same level ofresolution is not needed between coronal slices. There may besome partial volume effects due to the thickness of the slicesused (0.5 mm), and this does represent a limitation of the currentstudy. However, to keep imaging time within the range that canbe tolerated by an anesthetized animal while obtaining thenecessary high resolution coronal images, a slice thickness of0.5 mm was determined to be optimal.
The extent of injury was quantified using stereologicaltechniques. These techniques allowed for the unbiasedsystematic random sampling of regions of interest to quantifythe extent of axonal injury as evidenced by APP immunostain-ing. Although robust in its ability to determine the density ofAPP stained axonal varicosities in an unbiased fashion, there isa possibility that over-counting of injured axons may occur as asingle injured axon may have multiple APP stained varicos-ities. In addition, under-counting may occur as some injuredaxons may not stain with APP (Marmarou et al., 2005; Stone etal., 2001). Thus, the quantitative estimates of axonal injuryshould be regarded as approximate. Advanced methods usingmultiple markers of axonal injury and explicit corrections fordouble-counting are under development, but are beyond thescope of this current manuscript. Despite these limitations,stereological quantitative methods provide marked advantagesover semi-quantitative methods, such as thresholding, regionscoring systems, and injured axon counts performed in a non-systematic fashion. There was a strong correlation betweenchanges in relative anisotropy and the density of APP stainedaxons. It will be interesting to determine whether thiscorrelation improves even further if multiple markers of axonalinjury are used.
Axonal injury was observed to occur in a spatially gradedpattern; its severity appears to fall off with distance from theepicenter of the focal injury in this model. Therefore, no singleanatomical structure included all of the axonal injury, andinjured regions included some normal white matter. We usedconsistent anatomical landmarks surrounding the region ofinjury to compare changes in DTI parameters and conventionalMR parameters. This technique allowed analysis of the MRIsignal characteristics arising from a well-defined regioncontaining histologically-verified TAI. As techniques for theco-registration of MRI and histological data sets improve, itmay be possible to more precisely compare the spatialdistribution of histologically defined TAI to spatial patterns ofabnormal DTI or MRI data. Such techniques include transfor-mation into standardized space (Nowinski et al., 2006), ex vivoimaging of tissue slices (Bo et al., 2004; De Groot et al., 2001;Schmierer et al., 2003), and the use of fiducial marker systems(Susil et al., 2006).
129C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
The observed decreases in relative anisotropy were due toeither isolated reductions in AD (Eq. (5); Fig. 4) or thecombination of reduced AD and increased RD (Fig. 8). It hasbeen hypothesized that reduced AD reflects axonal injury andincreased RD reflects myelin injury (Song et al., 2003). Thepurpose of the current study was not to test this hypothesis, asthis model of experimental TBI produces a complex injury.Instead, our findings provide evidence that DTI showsabnormal signal in areas with histologically confirmed axonalinjury. A limitation of the current study is that we did notexhaustively quantify other pathologies, such as demyelination,gliosis, and inflammatory responses that may be occurring inaddition to axonal injury.
Our findings provide support for the hypothesis thattraumatic axonal injury can be associated with changes inDTI signal characteristics. They do not fully address all of theissues of timing, sensitivity and specificity that are required todetermine the utility of this method as a novel diagnostictechnique for clinical use. In the short term, more experimentaltime points are needed to describe the evolution of TAI and theability of DTI to accurately track this progression. A greaterrange of injury severities is needed to establish the limits ofsensitivity of this technique. A more precise co-registrationmethod would improve the quantitative comparison betweenMRI images and histology. Evaluation of demyelination andedema may help address the histological correlates of anisolated reduction in AD versus a decrease in AD with aconcomitant increase in RD. It will also help address thespecificity of a reduction in relative anisotropy without changesin conventional MRI parameters. For example, in a recent DTimaging study reduced RA associated with elevated T2 signal(suggestive of edema) was associated with unaltered whitematter tracts assessed by fiber tracking (Ducreux et al., 2005). Incontrast, reduced RA and normal T2 signal was associated withdisrupted white matter tracts. However, neither finding wasconfirmed histologically. In the current study, there were no T2abnormalities suggestive of edema in the regions of interest. Infuture work, it will be important to assess the extent of edemahistologically. Imaging studies in other animals in-vivo and inpost-mortem human tissue samples will be needed to assess thegenerality of these findings.
In the long term, two groups of patients could benefit mostfrom improved non-invasive assessment of TAI. First, TBIpatients with low initial Glasgow Coma Score ratings couldbenefit from a technique that allows accurate distinctionbetween reduced neurological function due to brain injury asopposed to concomitant medical issues such as the effects ofalcohol or sedative drugs, infection, metabolic disturbances, orextracranial injuries. Second, mild TBI patients with ambiguousclinical histories and normal conventional imaging results couldbenefit from an objective test that could verify the presence ofaxonal injury for the purposes of rehabilitation, prognosis, andclarification of the nature of the injury to third parties (i.e.insurance companies, workman's compensation, legal proceed-ings, etc.).
The importance of noninvasive methods to detect andquantify axonal injury is underscored by ongoing trials of
neuroprotectants such as cyclosporin A (Empey et al., 2006;Mazzeo et al., 2006) and hypothermia (Adelson et al., 2005)that are aimed at least in part at reducing axonal injury (Buki etal., 1999; Koizumi and Povlishock, 1998; Okonkwo et al.,1999; Scheff and Sullivan, 1999). Stratification of patientsbased on the presence or severity of axonal injury couldsignificantly improve the design of such trials. Furthermore,changes in imaging characteristics reflective of axonal injurycould be used as a pharmacodynamic biomarker to develop andmonitor such therapeutics.
In conclusion, DTI has significant potential as a tool to assistin the clinical and experimental evaluation of traumatic braininjury. Improved assessment techniques may speed the devel-opment of effective therapeutic strategies, aid in the clinicalmanagement of patients, and allow more accurate prognosticstatements to be made at earlier time points following injury.However, further experimental work is clearly required to fullydefine the utility of this approach.
Acknowledgments
The study was funded in part by NIH RO1 NS047592 (SKS),NIH NS049237 (DLB), NIH P01 NS032636 (PVB), NIH R21NS45237 (PVB) and a Burroughs Wellcome Career Award inthe Biomedical Sciences (DLB). We would like to thank Dr. JeffNeil, Dr. Kurt Thoroughman, and Dr. Joong Hee Kim forinsightful discussions regarding the manuscript and Mrs. MaiaParsadanian for immunocytochemistry instruction.
References
Adams, J.H., 1982. Diffuse axonal injury in non-missile head injury. Injury 13,444–445.
Adams, J.H., Doyle, D., Graham, D.I., Lawrence, A.E., McLellan, D.R., 1984.Diffuse axonal injury in head injuries caused by a fall. Lancet 2, 1420–1422.
Adelson, P.D., Ragheb, J., Kanev, P., Brockmeyer, D., Beers, S.R., Brown, S.D.,Cassidy, L.D., Chang, Y., Levin, H., 2005. Phase II clinical trial of moderatehypothermia after severe traumatic brain injury in children. Neurosurgery56, 740–754.
Arfanakis, K., Haughton, V.M., Carew, J.D., Rogers, B.P., Dempsey, R.J.,Meyerand, M.E., 2002. Diffusion tensor MR imaging in diffuse axonalinjury. AJNR Am. J. Neuroradiol. 23, 794–802.
Bammer, R., Augustin, M., Strasser-Fuchs, S., Seifert, T., Kapeller, P.,Stollberger, R., Ebner, F., Hartung, H.P., Fazekas, F., 2000. Magneticresonance diffusion tensor imaging for characterizing diffuse and focal whitematter abnormalities in multiple sclerosis. Magn. Reson. Med. 44, 583–591.
Beaulieu, C., 2002. The basis of anisotropic water diffusion in the nervoussystem — a technical review. NMR Biomed. 15, 435–455.
Blumbergs, P.C., Scott, G., Manavis, J., Wainwright, H., Simpson, D.A.,McLean, A.J., 1994. Staining of amyloid precursor protein to study axonaldamage in mild head injury. Lancet 344, 1055–1056.
Bo, L., Geurts, J.J., Ravid, R., Barkhof, F., 2004. Magnetic resonance imagingas a tool to examine the neuropathology of multiple sclerosis. Neuropathol.Appl. Neurobiol. 30, 106–117.
Brody, D.L., Mac Donald, C.L., Kessons, C.K., Yuede, C., Parsadanian, M.P.,Spinner, M.S., Holtzman, D.M., Bayly, P.V., 2006. An electromagneticcontrolled cortical impact device for precise, graded experimental traumaticbrain injury. J. Neurotrauma 23, 1021.
Buki, A., Koizumi, H., Povlishock, J.T., 1999. Moderate posttraumatichypothermia decreases early calpain-mediated proteolysis and concomitantcytoskeletal compromise in traumatic axonal injury. Exp. Neurol. 159,319–328.
130 C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
De Groot, C.J., Bergers, E., Kamphorst, W., Ravid, R., Polman, C.H., Barkhof,F., van der Valk, P., 2001. Post-mortem MRI-guided sampling of multiplesclerosis brain lesions: increased yield of active demyelinating and (p)reactive lesions. Brain 124, 1635–1645.
Deo, A.A., Grill, R.J., Hasan, K.M., Narayana, P.A., 2006. In vivo serialdiffusion tensor imaging of experimental spinal cord injury. J. Neurosci.Res. 83, 801–810.
Ducreux, D., Huynh, I., Fillard, P., Renoux, J., Petit-Lacour, M.C., Marsot-Dupuch, K., Lasjaunias, P., 2005. Brain MR diffusion tensor imaging andfibre tracking to differentiate between two diffuse axonal injuries.Neuroradiology 47, 604–608.
Erb, D.E., Povlishock, J.T., 1988. Axonal damage in severe traumatic braininjury: an experimental study in cat. Acta Neuropathol. (Berl.) 76, 347–358.
Filippi, M., Cercignani, M., Inglese, M., Horsfield, M.A., Comi, G., 2001.Diffusion tensor magnetic resonance imaging in multiple sclerosis.Neurology 56, 304–311.
Franklin, K.B., Paxinos, G., 1997. The Mouse Brain in Stereotaxic Coordinates.Academic Press, San Diego, CA.
Garbow, J.R., Dugas, J.P., Song, S.K., Conradi, M.S., 2004. A simple, robusthardware device for passive or active respiratory gating in MRI and MRSexperiments. Concepts Magn. Reson., Part B 21B, 40–48.
Gaviria, M., Bonny, J.M., Haton, H., Jean, B., Teigell, M., Renou, J.P., Privat,A., 2006. Time course of acute phase in mouse spinal cord injury monitoredby ex vivo quantitative MRI. Neurobiol. Dis. 22, 694–701.
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 theprimate. Ann. Neurol. 12, 564–574.
Gennarelli, T.A., Thibault, L.E., Tipperman, R., Tomei, G., Sergot, R., Brown,M., Maxwell, W.L., Graham, D.I., Adams, J.H., Irvine, A., et al., 1989.Axonal injury in the optic nerve: a model simulating diffuse axonal injury inthe brain. J. Neurosurg. 71, 244–253.
Gentleman, S.M., Nash, M.J., Sweeting, C.J., Graham, D.I., Roberts, G.W.,1993. Beta-amyloid precursor protein (beta APP) as a marker for axonalinjury after head injury. Neurosci. Lett. 160, 139–144.
Grady, M.S., McLaughlin, M.R., Christman, C.W., Valadka, A.B., Fligner, C.L.,Povlishock, J.T., 1993. The use of antibodies targeted against theneurofilament subunits for the detection of diffuse axonal injury in humans.J. Neuropathol. Exp. Neurol. 52, 143–152.
Gupta, R.K., Saksena, S., Agarwal, A., Hasan, K.M., Husain, M., Gupta, V.,Narayana, P.A., 2005. Diffusion tensor imaging in late posttraumaticepilepsy. Epilepsia 46, 1465–1471.
Hall, E.D., Sullivan, P.G., Gibson, T.R., Pavel, K.M., Thompson, B.M., Scheff,S.W., 2005. Spatial and temporal characteristics of neurodegenerationafter controlled cortical impact in mice: more than a focal brain injury.J. Neurotrauma 22, 252–265.
Huisman, T.A., Sorensen, A.G., Hergan, K., Gonzalez, R.G., Schaefer,P.W., 2003. Diffusion-weighted imaging for the evaluation of diffuseaxonal injury in closed head injury. J. Comput. Assist. Tomogr. 27,5–11.
Huisman, T.A., Schwamm, L.H., Schaefer, P.W., Koroshetz, W.J., Shetty-Alva,N., Ozsunar, Y., Wu, O., Sorensen, A.G., 2004. Diffusion tensor imaging aspotential biomarker of white matter injury in diffuse axonal injury. AJNRAm. J. Neuroradiol. 25, 370–376.
Inglese, M., Makani, S., Johnson, G., Cohen, B.A., Silver, J.A., Gonen, O.,Grossman, R.I., 2005. Diffuse axonal injury in mild traumatic brain injury: adiffusion tensor imaging study. J. Neurosurg. 103, 298–303.
Kessons, C.C., Mac Donald, C.L., Brody, D.L., Bayly, P.V., 2005. Effect of
velocity on depth error in controlled cortical impact studies. J. Neurotrauma22, 1253.
Kim, J.H., Budde, M.D., Liang, H.F., Klein, R.S., Russell, J.H., Cross, A.H.,Song, S.K., 2006. Detecting axon damage in spinal cord from a mousemodel of multiple sclerosis. Neurobiol. Dis. 21, 626–632.
King Jr., J.T., Carlier, P.M., Marion, D.W., 2005. Early Glasgow Outcome Scalescores predict long-term functional outcome in patients with severetraumatic brain injury. J. Neurotrauma 22, 947–954.
Koizumi, H., Povlishock, J.T., 1998. Posttraumatic hypothermia in thetreatment of axonal damage in an animal model of traumatic axonal injury.J. Neurosurg. 89, 303–309.
Koo, E.H., Sisodia, S.S., Archer, D.R., Martin, L.J., Weidemann, A.,Beyreuther, K., Fischer, P., Masters, C.L., Price, D.L., 1990. Precursor ofamyloid protein in Alzheimer disease undergoes fast anterograde axonaltransport. Proc. Natl. Acad. Sci. U. S. A. 87, 1561–1565.
Le Bihan, D., 2003. Looking into the functional architecture of the brain withdiffusion MRI. Nat. Rev., Neurosci. 4, 469–480.
Le Bihan, D., Breton, E., Lallemand, D., Grenier, P., Cabanis, E., Laval-Jeantet,M., 1986. MR imaging of intravoxel incoherent motions: application todiffusion and perfusion in neurologic disorders. Radiology 161, 401–407.
Lewen, A., Li, G.L., Nilsson, P., Olsson, Y., Hillered, L., 1995. Traumatic braininjury in rat produces changes of beta-amyloid precursor proteinimmunoreactivity. NeuroReport 6, 357–360.
Longhi, L., Saatman, K.E., Raghupathi, R., Laurer, H.L., Lenzlinger, P.M.,Riess, P., Neugebauer, E., Trojanowski, J.Q., Lee, V.M., Grady, M.S.,Graham, D.I., McIntosh, T.K., 2001. A review and rationale for the useof genetically engineered animals in the study of traumatic brain injury.J. Cereb. Blood Flow Metab. 21, 1241–1258.
Marmarou, C.R., Povlishock, J.T., 2006. Administration of the immunophilinligand FK506 differentially attenuates neurofilament compaction andimpaired axonal transport in injured axons following diffuse traumaticbrain injury. Exp. Neurol. 197, 353–362.
Marmarou, C.R., Walker, S.A., Davis, C.L., Povlishock, J.T., 2005. Quantitativeanalysis of the relationship between intra-axonal neurofilament compac-tion and impaired axonal transport following diffuse traumatic brain injury.J. Neurotrauma 22, 1066–1080.
Maxwell, W.L., Povlishock, J.T., Graham, D.L., 1997. A mechanistic analysis ofnondisruptive axonal injury: a review. J. Neurotrauma 14, 419–440.
Mazzeo, A.T., Kunene, N.K., Gilman, C.B., Hamm, R.J., Hafez, N., Bullock,M.R., 2006. Severe human traumatic brain injury, but not cyclosporin atreatment, depresses activated T lymphocytes early after injury. J. Neurotrauma23, 962–975.
Medana, I.M., Esiri, M.M., 2003. Axonal damage: a key predictor of outcome inhuman CNS diseases. Brain 126, 515–530.
Mori, S., van Zijl, P.C., 2002. Fiber tracking: principles and strategies—Atechnical review. NMR Biomed. 15, 468–480.
Mori, S., Zhang, J., 2006. Principles of diffusion tensor imaging and itsapplications to basic neuroscience research. Neuron 51, 527–539.
Nair, G., Tanahashi, Y., Low, H.P., Billings-Gagliardi, S., Schwartz, W.J.,Duong, T.Q., 2005. Myelination and long diffusion times alter diffusion-tensor-imaging contrast in myelin-deficient shiverer mice. NeuroImage 28,165–174.
Nakayama, N., Okumura, A., Shinoda, J., Yasokawa, Y.T., Miwa, K.,Yoshimura, S.I., Iwama, T., 2006. Evidence for white matter disruption intraumatic brain injury without macroscopic lesions. J. Neurol. Neurosurg.Psychiatry 77, 850–855.
Neil, J., Miller, J., Mukherjee, P., Huppi, P.S., 2002. Diffusion tensor imaging ofnormal and injured developing human brain — a technical review. NMRBiomed. 15, 543–552.
Nevin, N.C., 1967. Neuropathological changes in the white matter followinghead injury. J. Neuropathol. Exp. Neurol. 26, 77–84.
Nevo, U., Hauben, E., Yoles, E., Agranov, E., Akselrod, S., Schwartz, M.,Neeman, M., 2001. Diffusion anisotropy MRI for quantitative assessment ofrecovery in injured rat spinal cord. Magn. Reson. Med. 45, 1–9.
Nowinski, W.L., Qian, G., Bhanu Prakash, K.N., Hu, Q., Aziz, A., 2006. Fast
131C.L. Mac Donald et al. / Experimental Neurology 205 (2007) 116–131
Talairach Transformation for magnetic resonance neuroimages. J. Comput.Assist. Tomogr. 30, 629–641.
Okonkwo, D.O., Buki, A., Siman, R., Povlishock, J.T., 1999. Cyclosporin Alimits calcium-induced axonal damage following traumatic brain injury.NeuroReport 10, 353–358.
Ono, J., Harada, K., Takahashi, M., Maeda, M., Ikenaka, K., Sakurai, K., Sakai,N., Kagawa, T., Fritz-Zieroth, B., Nagai, T., et al., 1995. Differentiationbetween dysmyelination and demyelination using magnetic resonancediffusional anisotropy. Brain Res. 671, 141–148.
Oppenheimer, D.R., 1968. Microscopic lesions in the brain following headinjury. J. Neurol. Neurosurg. Psychiatry 31, 299–306.
Pierpaoli, C., Jezzard, P., Basser, P.J., Barnett, A., Di Chiro, G., 1996. Diffusiontensor MR imaging of the human brain. Radiology 201, 637–648.
Pilz, P., 1983. Axonal injury in head injury. Acta Neurochir. Suppl. (Wien.) 32,119–123.
Povlishock, J.T., Katz, D.I., 2005. Update of neuropathology and neuro-logical recovery after traumatic brain injury. J. Head Trauma Rehabil. 20,76–94.
Povlishock, J.T., Becker, D.P., Cheng, C.L., Vaughan, G.W., 1983. Axonalchange in minor head injury. J. Neuropathol. Exp. Neurol. 42, 225–242.
Rugg-Gunn, F.J., Symms, M.R., Barker, G.J., Greenwood, R., Duncan, J.S.,2001. Diffusion imaging shows abnormalities after blunt head trauma whenconventional magnetic resonance imaging is normal. J. Neurol. Neurosurg.Psychiatry 70, 530–533.
Sabo, T., Lomnitski, L., Nyska, A., Beni, S., Maronpot, R.R., Shohami, E.,Roses, A.D., Michaelson, D.M., 2000. Susceptibility of transgenic miceexpressing human apolipoprotein E to closed head injury: the allele E3 isneuroprotective whereas E4 increases fatalities. Neuroscience 101,879–884.
Scheff, S.W., Sullivan, P.G., 1999. Cyclosporin A significantly amelioratescortical damage following experimental traumatic brain injury in rodents.J. Neurotrauma 16, 783–792.
Schmierer, K., Scaravilli, F., Barker, G.J., Gordon, R., MacManus, D.G., Miller,D.H., 2003. Stereotactic co-registration of magnetic resonance imaging andhistopathology in post-mortem multiple sclerosis brain. Neuropathol. Appl.Neurobiol. 29, 596–601.
Schwartz, E.D., Duda, J., Shumsky, J.S., Cooper, E.T., Gee, J., 2005. Spinalcord diffusion tensor imaging and fiber tracking can identify white mattertract disruption and glial scar orientation following lateral funiculotomy.J. Neurotrauma 22, 1388–1398.
Sherriff, F.E., Bridges, L.R., Sivaloganathan, S., 1994. Early detection of axonalinjury after human head trauma using immunocytochemistry for beta-amyloid precursor protein. Acta Neuropathol. (Berl.) 87, 55–62.
Smith, D.H., Chen, X.H., Nonaka, M., Trojanowski, J.Q., Lee, V.M., Saatman,K.E., Leoni, M.J., Xu, B.N., Wolf, J.A., Meaney, D.F., 1999. Accumulationof amyloid beta and tau and the formation of neurofilament inclusionsfollowing diffuse brain injury in the pig. J. Neuropathol. Exp. Neurol. 58,982–992.
Smith, D.H., Chen, X.H., Iwata, A., Graham, D.I., 2003a. Amyloid betaaccumulation in axons after traumatic brain injury in humans. J. Neurosurg.98, 1072–1077.
Smith, D.H., Meaney, D.F., Shull, W.H., 2003b. Diffuse axonal injury in headtrauma. J. Head Trauma Rehabil. 18, 307–316.
Song, S.K., Sun, S.W., Ju, W.K., Lin, S.J., Cross, A.H., Neufeld, A.H., 2003.Diffusion tensor imaging detects and differentiates axon and myelindegeneration in mouse optic nerve after retinal ischemia. NeuroImage 20,1714–1722.
Song, S.K., Kim, J.H., Lin, S.J., Brendza, R.P., Holtzman, D.M., 2004.Diffusion tensor imaging detects age-dependent white matter changes in atransgenic mouse model with amyloid deposition. Neurobiol. Dis. 15,640–647.
Song, S.K., Yoshino, J., Le, T.Q., Lin, S.J., Sun, S.W., Cross, A.H., Armstrong,R.C., 2005. Demyelination increases radial diffusivity in corpus callosum ofmouse brain. NeuroImage 26, 132–140.
Sterio, D.C., 1984. The unbiased estimation of number and sizes of arbitraryparticles using the dissector. J. Microsc. 134, 127–136.
Stone, J.R., Singleton, R.H., Povlishock, J.T., 2000.Antibodies to the C-terminusof the beta-amyloid precursor protein (APP): a site specific marker for thedetection of traumatic axonal injury. Brain Res. 871, 288–302.
Stone, J.R., Singleton, R.H., Povlishock, J.T., 2001. Intra-axonal neurofilamentcompaction does not evoke local axonal swelling in all traumatically injuredaxons. Exp. Neurol. 172, 320–331.
Strich, S.J., 1956. Diffuse degeneration of the cerebral white matter in severedementia following head injury. J. Neurol. Neurosurg. Psychiatry 19,163–185.
Strich, S.J., 1961. Shearing of nerve fibres as cause of brain damage due to headinjury. Lancet 2, 443–448.
Sundgren, P.C., Dong, Q., Gomez-Hassan, D., Mukherji, S.K., Maly, P., Welsh,R., 2004. Diffusion tensor imaging of the brain: review of clinicalapplications. Neuroradiology 46, 339–350.
Susil, R.C., Menard, C., Krieger, A., Coleman, J.A., Camphausen, K., Choyke,P., Fichtinger, G., Whitcomb, L.L., Coleman, C.N., Atalar, E., 2006.Transrectal prostate biopsy and fiducial marker placement in a standard 1.5Tmagnetic resonance imaging scanner. J. Urol. 175, 113–120.
Teasdale, G.M., Murray, G.D., Nicoll, J.A., 2005. The association betweenAPOE epsilon4, age and outcome after head injury: a prospective cohortstudy. Brain 128, 2556–2561.
Thurman, D.J., Alverson, C., Dunn, K.A., Guerrero, J., Sniezek, J.E., 1999.Traumatic brain injury in the United States: a public health perspective.J. Head Trauma Rehabil. 14, 602–615.
Warach, S., Chien, D., Li, W., Ronthal, M., Edelman, R.R., 1992. Fast magneticresonance diffusion-weighted imaging of acute human stroke. Neurology42, 1717–1723.
Warden, D., 2006. Military TBI during the Iraq and Afghanistan wars. J. HeadTrauma Rehabil. 21, 398–402.
Wilde, E.A., Chu, Z., Bigler, E.D., Hunter, J.V., Fearing, M.A., Hanten, G.,Newsome, M.R., Scheibel, R.S., Li, X., Levin, H.S., 2006. Diffusion tensorimaging in the corpus callosum in children after moderate to severetraumatic brain injury. J. Neurotrauma 23, 1412–1426.
Yaghmai, A., Povlishock, J., 1992. Traumatically induced reactive change asvisualized through the use of monoclonal antibodies targeted to neurofila-ment subunits. J. Neuropathol. Exp. Neurol. 51, 158–176.
