Experimental Neurology 275 (2016) 436–449

Contents lists available at ScienceDirect

Experimental Neurology
journal homepage: www.elsevier.com/locate/yexnr

Repetitive mild traumatic brain injury with impact acceleration in the

mouse: Multifocal axonopathy, neuroinflammation, and
neurodegeneration in the visual system
Leyan Xu a,⁎, Judy V. Nguyen b, Mohamed Lehar c, Adarsh Menon a, Elizabeth Rha a, John Arena a, Jiwon Ryu a,
Nicholas Marsh-Armstrong b,d, Christina R. Marmarou e, Vassilis E. Koliatsos a,f,g
a
Division of Neuropathology, Department of Pathology, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA
b
Department of Neuroscience, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA
c
Department of Otolaryngology-HNS, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA
d
Hugo W. Moser Research Institute at Kennedy Krieger, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA
e
Department of Neurosurgery, Virginia Commonwealth University, Richmond, VA 23298, USA
f
Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA
g
Department of Psychiatry and Behavioral Sciences, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA

a r t i c l e i n f o a b s t r a c t

Article history: Repetitive mild traumatic brain injury (mTBI) is implicated in chronic neurological illness. The development of
Received 6 August 2014 animal models of repetitive mTBI in mice is essential for exploring mechanisms of these chronic diseases, includ-
Revised 29 October 2014 ing genetic vulnerability by using transgenic backgrounds. In this study, the rat model of impact acceleration (IA)
Accepted 4 November 2014
was redesigned for the mouse cranium and used in two clinically relevant repetitive mTBI paradigms. We first de-
Available online 20 November 2014
termined, by using increments of weight dropped from 1 m that the 40 g weight was most representative of mTBI
Keywords:
and was not associated with fractures, brain contusions, anoxic–ischemic injury, mortality, or significant neuro-
Traumatic axonal injury logical impairments. Quantitative evaluation of traumatic axonal injury (TAI) in the optic nerve/tract, cerebellum
Diffuse axonal injury and corpus callosum confirmed that weight increase produced a graded injury. We next evaluated two novel re-
Concussion petitive mTBI paradigms (1 time per day or 3 times per day at days 0, 1, 3, and 7) and compared the resulting TAI,
CTE neuronal cell death, and neuroinflammation to single hit mTBI at sub-acute (7 days) and chronic time points (10
Retinal ganglion cell weeks) post-injury. Both single and repetitive mTBI caused TAI in the optic nerve/tract, cerebellum, corticospinal
Optic nerve tract, lateral lemniscus and corpus callosum. Reactive microglia with phagocytic phenotypes were present at in-
Tau
jury sites. Severity of axonal injury corresponded to impact load and frequency in the optic nerve/tract and cere-
bellum. Both single and repeat injury protocols were associated with retinal ganglion cell loss and optic nerve
degeneration; these outcomes correlated with impact load and number/frequency. No phosphorylated tau immu-
noreactivity was detected in the brains of animals subjected to repetitive mTBI. Our findings establish a new
model of repetitive mTBI model featured by TAI in discrete CNS tracts, especially the visual system and cerebel-
lum. Injury in retina and optic nerve provides a sensitive measure of severity of mTBI, thus enabling further stud-
ies on mechanisms and experimental therapeutics. Our model can also be useful in exploring mechanisms of
chronic neurological disease caused by repetitive mTBI in wild-type and transgenic mice.
© 2014 Elsevier Inc. All rights reserved.

1. Introduction A recent trend that has attracted much attention is an increasing num-
ber of cases of progressive tauopathy in professional and amateur
There are at least 2–3 million new cases of civilian TBI in the US athletes with careers in collision sports such as football (Omalu et al.,
every year, most of them from motor vehicle accidents (MVA) or falls. 2005, 2006; McKee et al., 2009; Goldstein et al., 2012). These cases

Abbreviations: AD, Alzheimer's disease; ALS, amyotrophic lateral sclerosis; APP, amyloid precursor protein; BBB, blood–brain barrier; CCI, controlled cortical impact injury; CST, corticospinal
tract; CTE, chronic traumatic encephalopathy; CV, Cresyl violet; DAB, 3,3′-diaminobenzidine; DAPI, 4′,6-diamidino-2-phenylindole; FPI, fluid percussion injury; GOS, Glasgow Outcome Scale;
H&E, hematoxylin and eosin; IA, impact acceleration; IHC, immunohistochemistry; mTBI, mild traumatic brain injury; MVA, motor vehicle accidents; NFL, national football league; NSS,
Neurological Severity Score; ON, optic nerve; ONc, optic nerve segment proximal to chiasm; RGCs, retinal ganglion cells; ROIs, regions of interest; TAI, traumatic axonal injury.
⁎ Corresponding author at: The Johns Hopkins University School of Medicine, Division of Neuropathology, Department of Pathology, 720 Rutland Ave., Ross Research Building 558,
Baltimore, MD 21205, USA.
E-mail addresses: lxu9@jhmi.edu (L. Xu), judy.v.nguyen@gmail.com (J.V. Nguyen), mlehar@jhmi.edu (M. Lehar), amen692@gmail.com (A. Menon), lizrha@gmail.com (E. Rha),
jarena2@jhu.edu (J. Arena), jryu4@jhmi.edu (J. Ryu), Marsh-Armstrong@Kennedykrieger.Org (N. Marsh-Armstrong), crmarmar@vcu.edu (C.R. Marmarou), koliat@jhmi.edu
(V.E. Koliatsos).

http://dx.doi.org/10.1016/j.expneurol.2014.11.004
0014-4886/© 2014 Elsevier Inc. All rights reserved.
 L. Xu et al. / Experimental Neurology 275 (2016) 436–449 437

have been linked to repetitive concussions and appear to be identical to

cases of dementia pugilistica (Martland, 1928; Millspaugh, 1937;
Corsellis et al., 1973), with which they are classified under the rubric
of chronic traumatic encephalopathy (CTE) (Goldstein et al., 2012).
Repetitive mild blast TBI from exposure to explosive munitions is the
signature injury in the Iraq and Afghanistan war theaters (Warden,
2006) and the Defense and Veterans Brain Injury Center has recently
developed return-to-duty guidelines to prevent further TBI incidents
in soldiers with concussive histories (Barth, 2011). The public concern
over repeat concussions is growing not only because of the risk among
NFL professionals and active-duty soldiers or veterans, but also because
of the exposure of millions of non-professional athletes playing football,
soccer and other contact sports (Langlois et al., 2006).
The development of animal models that approximate human con-
cussion scenarios can provide a much-needed proof of concept by
linking repetitive injury to adverse long-term effects, including neuro-
degeneration. Furthermore, animal models of repetitive TBI can be ap-
plied to transgenic rodents to work out molecular mechanisms that
may help identify subjects with genetic predispositions to traumatic
tauopathy. Such genetic backgrounds may include pathogenic tau mu-
tations (Ballatore et al., 2007) or H1 tau haplotypes (Ferrari et al.,
2011; Vandrovcova et al., 2010). In addition, animal models can serve
as vehicles for therapeutic targeting and experimental therapeutics.
In this paper, we modified a well-characterized rat model of closed,
non-contusive head injury, i.e. the impact acceleration (IA) model
(Marmarou et al., 1994; Beaumont et al., 1999), for repetitive use in
mice at a mild impact level. Although the IA model can produce a grad-
ed, widespread injury involving neurons, astrocytes, axons, and the mi-
crovasculature, it does not cause focal damage regardless of injury
severity. We believe that the mild, repetitive, non-contusive IA injury
by this IA mouse model that approximates conditions encountered in
repetitive concussion in contact sports, is instructive for repetitive
blast TBI occurring in combat, and may provide general insights for
situations featured by repetitive TBI such as epilepsy, self-injurious be-
haviors, and child and domestic abuse. Our findings are consistent
with the view that repetitive TBI causes cumulative axonopathy that Fig. 1. A simplified sketch of our mouse IA model that was adapted from the original rat
can lead to degeneration of axotomized CNS neurons. Marmarou model.

2. Materials and methods

2.1. Experimental animals and surgical procedures
The subjects of these experiments were 5–6 weeks old C57BL6/J
mice (male, n = 117; Charles River Laboratories, Wilmington, MA).
The animals received humane care in compliance with the Guide for
the Care and Use of Laboratory Animals, 8th Ed. (National Academy
Press, Washington, DC, 2011) and all animal care, surgical and post-
operative procedures were approved by the Animal Care and Use
Committee of The Johns Hopkins Medical Institutions. The animals
were housed in the vivarium with 12 h light/12 h dark cycles and
given access to pellet food and water ad libitum.
The animals were exposed to a modified version of the IA method of
TBI initially described for rat by Marmarou et al. (1994) (Fig. 1). The
Plexiglass tube was 1 m high with an inner diameter of 10 mm,
designed to clear a column of 10 g, 3.6 mm-diameter brass cylinder
weights. The diameter of the steel disk fixed on the mouse skull was
set at 3.2 mm, designed to ride over bregma and lambda sutures
when affixed with cyanoacrylate to the skull, and thickness was set at
1 mm. As described in original rat IA model (Foda and Marmarou,
1994; Marmarou et al., 1994), the main purpose of the stainless disk
was to prevent fracture. In this position, acceleration was confined to
the sagittal plane and caused both translational and rotational displace-
ment. The foam bed (4–0 spring constant foam, Foam to Size Inc.,
Ashland, VA) was used as initially described (Foda and Marmarou,
1994; Marmarou et al., 1994), but cut at smaller dimensions of 20 × 8
× 6 cm to accommodate the mouse body placed in the prone position.
All surgical procedures were carried out with gas anesthesia
(isoflurane:oxygen:nitrous oxide = 1:33:66) and under aseptic condi-
tions. Briefly, the steel disk was glued on the saline washed, air dried
cranium between bregma and lambda under microscopic guidance
(Fig. 1). The mouse was placed prone on the foam bed under the hollow
Plexiglass tube, and secured with strapping tape. The injury was in-
duced by dropping the column of brass weights through the Plexiglass
tube from a distance of 1 m onto the disk (Fig. 1). The foam bed was
moved quickly right after the impact to avoid secondary rebound injury.
The mouse was then placed on warm pad for recovery. At the same
time, self-righting time was recorded. Unlike the rat IA model, mechan-
ical ventilation was not applied, in order to imitate real-life concussion
conditions. After righting itself, the mouse was re-anesthetized, the
steel disk removed, and the skull checked under the surgical micro-
scope for skull fractures. The scalp incision was then closed with surgi-
cal staples, and the animals were returned to cage for full recovery.
Animals with skull fractures were excluded from further study. Sham
animals were subjected to the same procedures minus the weight
drop maneuver. A list of experimental groups and numbers of animals
is provided in Table 1.
2.2. Grading severity of single TBI regimens
To generate a range of injury severities based on lethality (n = 61),
the impactor height was set at 1 m. Weight was initially set at 10 g and
increased by 10 g increments. Weight increase stopped when mortality
438 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
Table 1
Experimental design and groups of subjects.
Main experiments
A. Grading severity of single IA regimen to define mTBI parameters
B. Ensuring animal wellbeing after each IA hit using parameters
established in (A)
C. Repetitive IA using mTBI parameters derived from (A)—brain
experiments
D. Repetitive IA using mTBI parameters derived from (A)—eye
and ON experiments
The table shows experiments in which quantitative analysis was performed. Some studies, for example: general neuropathology for the assessment of contusions and hippocampal or
Purkinje cell death; establishment of site of indirect impact to the nervous system with BBB studies; neuroinflammation in brain based on IBA1, CD68, and GFAP IHC; and tauopathy
based on phosphorylated tau IHC were not included. All repetitive mTBI experiments in the brain and eye-ON contained 4 groups, i.e., sham, 1 × 10 g, 10 × 40 g and 12 × 40 g
(n = 4–5 each). The same animals were used for multiple experiments.
Abbreviations: mTBI, mild traumatic brain injury; TAI, traumatic axonal injury; NSS, Neurological Severity Score; BBB, blood–brain barrier; RGC, retinal ganglion cell; ON, optic nerve; OT,
optic tract.
reached 50%. The highest weight associated with zero mortality (40 g)
was used for all subsequent single and repeat injury regimens.
2.3. Repetitive mTBI regimens
Using the 40 g− 1 m IA setting associated with zero mortality, we
generated two novel repetitive mild TBI regimens (n = 56). The first
regimen involved 4 hits in one week at days 0, 1, 3 and 7; the second
regimen included 12 hits at days 0, 1, 3 and 7 (3 hits on each of these
days). In the second regimen, to ensure that the animals had enough
time to recover, time intervals were kept at 2 h between first and second
hits and 3 h between second and third hits. A single injury regimen was
used for comparison. In the repetitive regimens, at the date of injury,
neurological deficits were assessed with an abbreviated NSS (Neurolog-
ical Severity Score) system tailored for rodents exactly as described
without any modification (Flierl et al., 2009). In order to evaluate sub-
acute and chronic impacts from repeat injury, the animals were eutha-
nized at day 7 or week 10 after single (for the single TBI regimen) or
final injury (for the repeat TBI regimens) and tissues were processed
as reported in the section below.
2.4. Histology, histochemistry, immunohistochemistry and microscopy
Seven days after final injury, the animals were perfused with freshly
depolymerized, 4% neutral-buffered paraformaldehyde. The brain and
eyes with associated optic nerves (ONs) were dissected and immersed
in the same fixative overnight at 4 °C. All tissues were cryoprotected
and stored at − 80 °C for further processing. Sagittal brain sections
(40 μm) were prepared in series for: general histochemistry,
i.e., hematoxylin and eosin (H&E) and Cresyl violet (CV); Gallyas silver
for axonal injury; immunohistochemistry (IHC) for neuroinflammation
markers, i.e. GFAP, CD68 and IBA1 and for phosphorylated tau markers,
i.e. AT8, PHF1, CP13 and pS422; and Fluoro-Jade C staining combined
with IHC for axonal injury and neuroinflammation. Fluoro-Jade C, a
marker of neuronal degeneration (Schmued et al., 2005; Chidlow
et al., 2009; Pohl et al., 2011), was used to localize axonal injury and
neuronal cell death. For Gallyas silver staining of injured/degenerating
axons and terminals, sections were processed with a commercially
available kit (Neurosilver kit II; FD Neurotechnologies, Ellicott City,
MD) as described (Koliatsos et al., 2011).
IHC utilized immunoperoxidase-DAB or dual-label immunofluores-
cence methods and was performed as previously described (Koliatsos
et al., 2011; Xu et al., 2009). Briefly, sections were first incubated in
the primary antibody overnight at 4 °C; for immunoperoxidase staining,
after incubation with peroxidase-conjugated secondary antibody linked
Survival time Measures (n = animal number)
N/A for mortality, 7 days for TAI Mortality rate, n = 61 total
sham, 10 g, 20 g and 30 g, n = 4
40 g, n = 17
50 g and 60 g, n = 14
TAI severity in ON, corpus callosum and cerebellum, n = 16 total
sham, 20 g, 40 g and 60 g, n = 4 each
N/A Body weight, n = 46 total
NSS, n = 20 total
7 days TAI in OT, corpus callosum, cerebellum, n = 17 total
7 days and 10 weeks Axon counts in ON, n = 16 total
RGC counts, n = 16 total
Microglial activation in retina, n = 16 total
with biotin (1:200; Jackson ImmunoResearch, West Grove, PA), sections
were developed with a 3,3′-diaminobenzidine (DAB) kit (Vectastain
Elite ABC Kit; Vector Laboratories Inc., Burlingame, CA). For immunoflu-
orescence staining, after incubation in secondary antibodies conjugated
with Cy3 or Cy2 (1:200; Jackson ImmunoResearch, West Grove, PA)
for 2–4 h at room temperature, the sections were counterstained with
the fluorescent DNA dye 4′, 6-diamidino-2-phenylindole (DAPI)
and coverslipped with DPX mounting media. Primary antibodies includ-
ed: rabbit anti-GFAP (1:400; Dako, Carpinteria, CA); rat anti-CD68
(1:500; ABD Serotec, Raleigh, NC) and rabbit anti-IBA1 (1:500; Dako,
Carpinteria, CA); mouse anti-AT8 directed against tau pS202 (1:200;
Thermo Scientific Inc., Rockford, IL); mouse anti-PHF1 directed against
tau pS396 and pS404 (1:200); mouse anti-CP13 directed against tau
pS202 and pT205 (1:200); rabbit anti-pS422 directed against phosphor-
ylated Ser-422 tau (1:600; Genetex, Irvine, CA). Antibodies PHF1 and
CP13 were provided by Dr. Peter Davies (Albert Einstein College of
Medicine, Bronx, NY). For combinations of Fluoro-Jade C with IHC,
brain sections were first stained with IHC as described above using a sin-
gle primary antibody. After incubation in the secondary antibody linked
with Cy3, sections were then processed for Fluoro-Jade C (Millipore,
Billerica, MA) staining (Schmued et al., 1997; Bian et al., 2007).
Whole flat-mounted retinas were used to explore axonal injury,
survival of retinal ganglion cells (RGCs) and neuroinflammation. Retinas
were dissected, rinsed with phosphate buffered saline and then dually
labeled with γ-synuclein (1:600; Genetex, Irvine, CA), to identify RGC
cell bodies, and the activated microglia marker CD68.
Stained sections were studied on a Zeiss Axiophot microscope
equipped for epifluorescence (Diagnostic Instruments Inc., Sterling
Heights, MI) or Zeiss LSM 510 inverted confocal microscope (Carl
Zeiss Inc., Oberkochen, Germany). Confocal microscopic images were
captured with pinhole set at 0.8 μm. Three-dimensional reconstruction
by Z-stack scanning through regions of interest (ROIs) was acquired
with LSM software. Adobe Photoshop 7.0 software (Adobe Systems,
San Jose, CA) was used for montaging and image processing.
2.5. Analysis of blood–brain barrier (BBB) disruption
Location and severity of primary traumatic impact to ON were ex-
plored with methods revealing BBB disruption as previously defined
(Wang et al., 2011). Briefly, animals subjected to single injury or two re-
petitive mild TBI regimens were euthanized 4 or 24 h after the sole or
final injury event (n = 3 for each time point). ONs were dissected
away from the eyes and brains, cryoprotected and then sectioned in
the sagittal plane. Serial sections (10 μm) were mounted on gelatin-
coated slides and dually immunolabeled for mouse IgG (serving as
 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
BBB leakage marker) with donkey anti-mouse antibody linked with Cy3
(1:200; Jackson ImmunoResearch, West Grove, PA) and for amyloid
precursor protein (APP) with rabbit anti-APP IgG against the carboxy-
terminus of APP (1:200; Zymed, Camarillo, CA), as per previous Section.
2.6. Quantitation of degenerating axons in the ON
ONs were dissected away from the eyes and brains and then
processed for toluidine blue staining on semithin sections. Briefly,
ONs were treated with a solution containing 4% paraformaldehyde
and 0.2% glutaraldehyde for 24 h. After rinsing in 0.1 M phosphate
buffer (pH 7.3) for 3–10 min, tissues were immersed in 1% osmium te-
troxide for 15 min and stained en bloc with 1% uranyl acetate for 1 h.
Stained tissues were dehydrated in graded concentrations of ethanol,
embedded in Poly/Bed 812 (Polysciences Inc., Warrington, PA) in
BEEM® capsules and polymerized at 60 °C for 72 h. Semithin sections
(1 μm) were cut transversely from segments of ONs proximal to the
optic chiasm. Next, sections were stained with 1% toluidine blue.
Myelinated axons were counted under 100× magnification by investi-
gators blinded to experimental groups with stereological methods
(the optical fractionator probe) using Stereo Investigator® software
(Microbrightfield Inc., Williston, VT) as described (Turgut et al., 2010).
The counting frame was set at 7 × 7 μm and sampling grid was
60 × 60 μm. The final number of axons was automatically estimated
by the software and used for analysis.
2.7. Quantitation of silver staining
The intensity of silver labeling of injured axons within regions of in-
terest (ROIs) in the brain and ON was studied with unbiased stereolog-
ical methodologies, i.e. area fraction fractionator (Hartman et al., 2005;
Jefferson et al., 2011) using Stereo Investigator® software. This method
obviates irrelevant variance caused by some technical factors present in
densitometric analysis such as overstaining, granular-particle back-
ground and image processing. The result is presented as area fraction
of silver (+) linear configurations in counted frames through the ROI.
In the present experiments, ROIs included optic tract (next to optic
chiasm), corpus callosum (same sections as optic tract) and cerebellar
lobule 9 (starting from section proximal to the lateral edge of lobule
9). Four serial sagittal sections (every 8 in series of 40 μm thickness
transverse sections) were selected through the ROIs and subjected to
stereological analysis. Counting frame was set at 50 × 50 μm and sam-
pling grid and grid spacing were 200 × 200 μm and 10 μm, respectively.
The final number of axons was automatically estimated by the software
and entered for analysis.
2.8. Retinal ganglion cell counts in whole-mount retina
Whole flat-mounted retinas dually immunolabeled for γ-synuclein
and CD68 were imaged with a 0.4 numerical aperture Zeiss 20× objec-
tive on a Zeiss 200 M inverted microscope (Carl Zeiss, Oberkochen,
Germany) equipped with a Linear Encoder Flat Top Zeiss 200 Ludl mo-
torized stage (Ludl, Hawthorne, NY), Roper Scientific Coolsnap FX digi-
tal camera (Photometrics, Tucson, AZ), and Excite fluorescence source
(EXFO Burleigh, Victor, NY) (Soto et al., 2008). After acquired images
were adjusted for right contrast and brightness using Adobe Photoshop
7.0, ImageJ (National Institutes of Health, Bethesda, MD) was used for
counting surviving γ-synuclein (+) RGCs and CD68 (+) microglia.
Cells were counted in 100 × 100 μm fields evenly assigned in the
peripheral, intermediate and central zone of retina. At least 10 fields
in the peripheral and intermediate zones and 8 fields in the central
zone were utilized. Cell density was then calculated by dividing total
cell numbers with total area surveyed.
439
2.9. Statistical methods
One-way analysis of variance (STATISTICA 8.0; StatSoft, Tulsa, OK)
was used to compare among groups of animals based on silver staining,
ON axon count and RGC count data. Comparisons were performed
across single impact and sham groups in assessing the role of cylinder
weight on injury severity and across repeat injury groups and sham
animals in all repeat-injury experiments. Tukey's post hoc test was
applied to reveal the significant main effects or interactions.
3. Results
3.1. General characterization of mouse IA model with single and repetitive
mild impacts
As laid out in Materials and methods, the weight drop height was set
at 1 m. Weight of impactor was tested starting at 10 g and further loaded
at 10 g increments. There was no fatality until weight reached 50 g. With
the 50 g− 1 m impact, mortality was 18.2% and with the 60 g− 1 m
impact, mortality was increased up to 40%. Based on these initial mortality
data, the 40 g−1 m impact was used in all subsequent repetitive injury
paradigms.
Change of body weight was used as an index to determine the opti-
mal between-impact time intervals in the repetitive injury experiments.
In an initial pilot experiment using a consecutive 4-day schedule, the
40 g− 1 m impact one or three times per day (12× or 4 × altogether)
arrested the physiological increase in body weight after the 2nd day
and/or 3rd day injury, respectively (Fig. S1). When the injury schedule
included time intervals between impacts, i.e. impacts at days 0, 1, 3
and 7, no significant disruption of body weight curve or enduring neu-
rological deficits were observed irrespective of impact burden (Fig. 2).
No NSS increase was found on any scheduled day in the two repetitive
injury regimens (4× and 12×).
Our single or repetitive IA model did not cause focal contusive injury
under the cranial disk or in other brain locations; in addition, our model
did not cause anoxic/ischemic injury in hippocampal CA1 or overt
reduction in numbers of cerebellar Purkinje cells (Fig. 3A–C). Fluoro-
Jade C staining was negative for cell bodies in all these locations in single
or repetitive regimens (data not shown). Based on the absence of frac-
tures, zero mortality, no significant changes in body weight and no
major neurological deficits as well as rapid neurological recovery after
each injury event, the single 40 g− 1 m impact was designated here
as “mild” injury. This designation conforms to the original IA model
description in which 450 g− 1 m was used to generate “mild” injury
(Foda and Marmarou, 1994; Marmarou et al., 1994; Beaumont et al.,
1999). As shown in the following sections, this setting (40 g− 1 m)
causes traumatic axonal injury (TAI) in multiple locations in mice, but
this pattern is similar to that in other mild TBI mouse models including
Fig. 2. Trends in body weight of animals included in the present study. Body weight curves
with single (1 × 40 g) and repetitive mild (4 × 40 g and 12 × 40 g) injuries illustrate that
animals gain weight at a rate comparable to that of sham subjects.
440 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
Fig. 3. Absence of focal contusion or anoxic/ischemic injury in the brain of mice subjected
to mild IA injury. CV staining 7 days after repetitive mild injury (4 × 40 g) does not show
focal contusion underneath the cranial disk (A) or significant neuronal death in Ammon's
horn (B) or cell death of Purkinje cells (C). Mol, molecular layer; P, Purkinje cell layer;
Gran, granular layer. Scale bars: A, 500 μm; B and C, 100 μm.
fluid percussion injury (FPI) (Greer et al., 2011; Wang et al., 2011), blast
injury (Koliatsos et al., 2011), electromagnetic coil-based impact
(Mouzon et al., 2012) and controlled cortical impact injury (CCI)
(Shitaka et al., 2011).
3.2. TAI in the brains of mice subjected to IA with single and repetitive mild
impacts
Axonal injury in the brain of mice subjected to IA with single and re-
petitive mild impacts was investigated with Gallyas silver and Fluoro-
Jade C staining. With these methods, our mouse IA model showed selec-
tive TAI in the ON/optic tract (Fig. 4A), lower corticospinal tract (Fig. 4B)
and cerebellar white matter/cerebellar peduncles (Fig. 4C); fiber abnor-
malities were also noted in the corpus callosum (Fig. 4D), internal cap-
sule and lateral and medial lemnisci. Injuries in ON/optic tract and lower
corticospinal tract were especially prominent. TAI in cerebellum was
restricted to lower folia close to the medulla.
Intensity of axonal degeneration depended on severity of injury as
determined by the weight of the impactor (in single hit scenarios)
(Fig. 4E and F) or number of hits (in multiple-hit scenarios) (Fig. 5).
The severity of TAI at various levels of impact burden in single-hit regi-
mens (sham, 20 g, 40 g and 60 g) and repeat-hit regimens (4 × 40 g
and 12 × 40 g) was assessed using Gallyas silver staining in three
regions, i.e. optic tract, the white matter of cerebellar lobule 9 and
corpus callosum (Fig. 4A–D). These regions are representative of severe
(optic tract), intermediate (lobule 9) and mild (corpus callosum) TAI
due to IA (Figs. 4 and 5). In the case of optic tract, the severity of TAI
appears to worsen as impactor weight increases from 20 g to 60 g in
the single hit group (Fig. 4E). Area fraction fractionator data confirm
this trend (Fig. 4F, left panel), although differences between the 40 g
and 60 g animals are not statistically significant. Severity of TAI in cere-
bellar lobule 9 demonstrates a similar trend, but the significant differ-
ences in this case are between sham and both 40 g and 60 g animals
(Fig. 4F, center panel). In the case of corpus callosum, only the 60 g im-
pact causes significant TAI (Fig. 4F, right panel). Besides showing a rela-
tionship between IA burden and TAI in our mouse model, the above
trends also indicate the differential vulnerability of axons in various
white matter tracts/regions.
In the case of mild repetitive TBI, both the low- and high-frequency
regimens (4 × 40 g and 12 × 40 g, respectively) produced significant TAI
in optic tract compared with the single injury regimen (1 × 40 g) and
sham, although there was no difference in the TAI outcome between
the two repetitive injury groups (Fig. 5A). The same pattern was seen
in the corpus callosum (Fig. 5C). In cerebellar lobule 9, both repetitive
injury regimens cause TAI, but they are not different when compared
with the single injury group (1 × 40 g) (Fig. 5B). These results imply
that, at some level, at least using the regimens and histochemical proce-
dures of this paper and in certain white matter regions, the cumulative
effect of repetitive injury becomes stabilized.
3.3. Neuroinflammatory changes in sites of TAI in the mouse repetitive mild
IA model
With repetitive mild IA injury, we found extensive microglial infil-
tration/activation in TAI sites, especially optic tract including superior
colliculus, corticospinal tract, corpus callosum and part of cerebellum
white matter (Fig. 6). Microglial activation was ascertained by the pres-
ence of deramified, large, often clustered IBA1 (+) cells, many of which
expressed the lysosomal phagocytic marker CD68 (Fig. 6A and B).
Astrocytic response is also increased based on the density of GFAP (+)
processes (Fig. 6C and D). In several TAI sites including ON/optic tract,
corticospinal tract and cerebellum areas, there was co-distribution of in-
jured Fluoro-Jade C (+) axons and phagocytic CD68 (+) microglia
(Fig. 6E–H).
3.4. Degenerative changes in the visual system with single or repetitive mild
TBI
The initial impact of TBI to the ON was assessed at 4 or 24 h post-IA
injury with mouse IgG and APP immunofluorescence on same sections.
Four hours after a single impact (1 × 40 g), we found that a sizeable seg-
ment of ON closer to the eye was labeled by anti-mouse IgG antibody, a
sign of BBB disruption due to the impact; this segment also showed
dense axonal swellings and bulbs stained with APP antibodies
(Fig. 7A). With a repetitive injury regimen (4 × 40 g), IgG immunoreac-
tivity was weaker than the single injury and so was the presence of APP
(+) axonal swellings or bulbs (Fig. 7C). Compared to 4 h, ONs at 24 h
post-single injury or after the last hit of repetitive injury regimens
showed restricted IgG immunoreactivity and fewer axonal abnormalities
(Fig. 7B and D). Furthermore, in animals with repetitive injury regimens,
axon bulbs presented mainly at the periphery of ONs (Fig. 7C and D).
 L. Xu et al. / Experimental Neurology 275 (2016) 436–449 441

Fig. 4. Selective axonal injury in the white matter of single and repetitive mild injured mice. Gallyas silver staining shows axonal injury (black lines or trails of dots) in select locations
including optic tract (OT, A, arrow heads), cerebellar lobule 9 (Cb9, B), corticospinal tract (CST, C, arrow heads) and corpus callosum (cc, D, arrow heads) after low-frequency repetitive
injury (4 × 40 g). Higher magnification images (insets) are enlarged from corresponding frames in main panels to show details. In the corticospinal tract (C), inset showcases a typical axon
bulb (arrow). Severity of TAI in the optic tract worsens when impactor weight increases from 20 g to 60 g (E). There is no TAI in the optic tract of sham animals (E, upper panel). Panel F
shows the results of quantitation of silver staining in the optic tract (left panel), cerebellar lobule 9 (center panel) and corpus callosum (right panel) of sham and single injured animals
with impactor weights at 20, 40 and 60 g. One-way ANOVA is used for analysis. * p b 0.05. Scale bars: A, 100 μm; B–D, 50 μm.

As mentioned in previous Sections, IA causes widespread axonal in-
jury and intense microglial response in the mouse visual pathway in-
cluding ON, optic tract and superior colliculus. Retinal ganglion cells
(RGCs), i.e. the main projection neurons in ON/optic tract were further
examined for evidence of neuronal degeneration in the animals subject-
ed to mild single and repetitive IA injury. RGC loss after mild single and
repetitive injuries in whole-mount retinas was obvious at 10 weeks
after final injury (Fig. 8A–D). The density of surviving γ-synuclein (+)
RGCs in whole-mount retinas was already decreased in mice subjected
to high-frequency repetitive injury (12×) at 7 days after injury (Fig. 8E),
whereas RGCs in single and low-frequency repetitive (4×) injury regi-
mens did not show significant cell loss at this early time point
(Fig. 8E). At 10 weeks, RGC density was significantly reduced in mice in-
jured with either single or both repetitive injury regimens (Fig. 8F).
Mice injured with repetitive mild regimens had significantly lower
RGC densities than mice subjected to single injury. Although mice
injured with the high-frequency (12 ×) repetitive regimen had lower
RGC densities than animals injured with the low-frequency (4 ×)
regimen, there was no significant difference between the two repetitive
injury groups. RGC loss in the mild single injury group was significant at
10 weeks but not 7 days post final injury (p b 0.05), a difference suggest-
ing that RGC death begins to take place after 7 days.
Axonal counts in the ON were performed at 10 weeks post-injury to
confirm the degeneration of RGCs. Counts were performed on semithin
sections through ON segments proximal to chiasm (ONc) after toluidine
blue staining. In ONc, surviving axon numbers after single and repetitive
injury were significantly lower than after single or sham injury. Fifty to
eighty percent of axons were lost after the repetitive injury regimens
(Fig. 9E). The area affected the most was the center of the nerve that
contained large spaces devoid of axons (Fig. 9A–D). Mice with high-
frequency repetitive injury (12 ×) lost significant number of axons
than low-frequency repetitive injury group (4×) in this area. Some de-
gree of ONc axonal degeneration was also seen with single mild injury
(Fig. 9B).
In the RGC layer of injured mice, there was intense microglial activa-
tion. Using whole-mount retinal preparations stained with markers of
activated microglia, we found that numbers of CD68 (+) microglial
cells were increased 7 days after injury in all injured groups (Fig. 10D).
Confocal 3D reconstruction of CD68 and γ-synuclein stained retina im-
ages demonstrates that the RGC layer was populated by activated mi-
croglia (Fig. 10A-C). Cell counts of CD68 (+) microglia 7 days after
injury confirmed that both single and repetitive injuries caused a signif-
icant increase in the number of activated microglial cells. Although there
were no significant differences among injury groups, these results show
that there is a positive correlation between injury burden and microglia
activation (Fig. 10D). At 10 weeks post injury, there were no significant
differences in numbers of activated microglial cells among sham, single
and repetitive injury groups (Fig. 10E), a pattern suggesting that the
442 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
Fig. 5. Quantitation of Gallyas silver signal. This figure shows quantitation of Gallyas silver
(+) degenerating axonal profiles in the optic tract (OT, A), cerebellar lobule 9 (Cb9,
B) and corpus callosum (cc, C) of sham, single injured (1 × 40 g) and repetitive mild in-
jured (4 × 40 g and 12 × 40 g) animals. One-way ANOVA with Tukey's post hoc test was
used to assess significance of variance. * p b 0.05.
microglial neuroinflammatory response is an early event in the neuro-
degenerative process after the injury.
We also explored the issue of tau hyperphosphorylation in the brains
of mice 10 weeks after high-frequency (12×) repetitive injury with an
extensive panel of phosphorylated tau antibodies. In representative sag-
ittal brain sections, we found no phosphorylated tau immunoreactivity
in cortical or subcortical neurons. Our antibodies did stain neurons in
the brains of tau P301S mice that served as positive controls (Fig. S2).
4. Discussion
4.1. General
Our findings indicate that our mild repetitive TBI mouse model, a
modification of the rat IA paradigm, does not cause either focal contusive
or anoxic–ischemic injury in the mouse brain. Rather, our model pro-
duces selective TAI in certain white matter tracts; this effect is coupled
with local neuroinflammatory responses and is positively associated
with TBI severity. A remarkable finding is the consistent involvement
of the visual system, with degeneration of axons in both ON and optic
tract and loss of RGCs. Visual involvement may be associated with direct
injury related to the concussion. The effect of repetitive mild TBI at sites
of TAI is cumulative but eventually reaches a plateau, although it is un-
known whether the effect is saturated at that level. In summary, here
we have developed and characterized a mouse model of repeat mild
TBI featured by TAI and neuroinflammatory components that are TBI
dose-dependent and accumulate with each repeat TBI event. As such,
this model lends itself to further investigations that have translational
value for sports and military injuries, including CTE.
Animal models are powerful tools to explore cause-and-effect rela-
tionships between traumatic insults and progressive neurodegenera-
tion and, eventually, to serve as vehicles for therapeutic targeting and
experimental therapeutics or for biomarker discovery and validation.
A distinct advantage of mouse TBI models in particular is that they can
be used to study the role of generic vulnerability in the outcome of
TBI. For example, transgenic mice harboring dementia-associated iso-
forms of ApoE, especially ApoE4, or pro-aggregation forms of tau may
be extremely useful to disclose the role of genetic vulnerability to CTE
after exposure to mild repetitive TBI (Strittmatter et al., 1993; Bu,
2009; McKee et al., 2009; Wang et al., 2007).
Some TBI animal models have been used in mild repetitive scenarios,
for example a modified CCI model (Laurer et al., 2001; Uryu et al., 2002;
Conte et al., 2004; Longhi et al., 2005; Shitaka et al., 2011), or the weight
drop model of Hall (Deford et al., 2002; Creeley et al., 2004). Another
mild repetitive TBI model is an electromagnetic coil-based impact
model modified from the electromagnetic version of the CCI paradigm
(Brody et al., 2007; Mouzon et al., 2012). Blast injury models have also
been scaled down to generate mild-moderate TBI, but these models
have not been used in repetitive injury scenarios because of potential le-
thal injury to vital organs, such as the lung (Cernak et al., 2011; Koliatsos
et al., 2011). Compared to the IA model used here, these mild repetitive
injury scenarios cause contusions or involve minimal to no head move-
ment. Although the IA model has been used to inflict mild injury in the
rat (Fujita et al., 2012; Marmarou et al., 1994), it has not been
established as a model of repetitive TBI in the rat or mouse.
As indicated elsewhere, the mouse model used in this paper was de-
rived from the rat IA model by modifying some key parameters. An im-
portant parameter is the weight–height configuration of the impactor.
Assuming a steady height of 1 m, weights of 50 g or above were associ-
ated with high mortality because of apnea. The 40 g− 1 m setting did
not have any fatalities, as is almost universally the case with human
concussions; therefore, this setting was used in all single and repetitive
mild TBI regimens employed here. Another important variable is the
size of the stainless steel disk used to prevent fractures. Because of the
thinness of the mouse cranium, unless the disk overlapped the bregma
and lambda sutures, i.e. cut at a diameter of 3.2 mm, it could not prevent
fractures. The young age of mice used (5–6 weeks old) was chosen in
order to recapitulate the age of athletes suffering mild TBI from contact
sports (Ojo et al., 2013).
In our repetitive mild TBI regimens, we used the “subacute” injury
schedule, i.e., injury at days 0, 1, 3 and 7. On each one of these days,
the animals were exposed to single or three hits (for a total of 4 or 12
TBI events, respectively). The use of three hits in a single day was
meant to replicate the pattern of multiple concussions in a single athlet-
ic encounter that is common in the world of collision sports. The use of
consecutive-day schedules was discouraged by our initial finding that
such a regimen interfered with normal weight gain, evidence of feeding
or metabolic problems, or both. The “recovery” time intervals used in
the subacute repetitive TBI schedule were 1 day, 2 days and 4 days for
the second, third and fourth hits, respectively. Animals injured in such
a fashion did not show evidence weight loss or neurological deficits
 L. Xu et al. / Experimental Neurology 275 (2016) 436–449 443

Fig. 6. Neuroinflammation in white matter tracts undergoing TAI after IA injury. Neuroinflammation is found in the optic nerve (ON)/optic tract (OT), corticospinal tract (CST) and cere-
bellar lobule (Cb) of repetitive mild injured (4 × 40 g) animals 7 days after final injury and is evidenced by IBA1 (A and B) or GFAP (C and D) staining or dual staining with CD68 and Fluoro-
Jade C (E–H). In the optic tract, repetitive injured (4 × 40 g) animals (B) have more IBA1 (+) microglia than sham animals (A). Some reactive microglial cells acquire phagocytic cytologies
(asterisks in B). Inset in (B) shows an activated microglial cell with phagocytic appearance (arrow) and one resting microglial cell (arrowhead). GFAP staining (C and D) in the optic tract
shows that animals exposed to repetitive mild IA injury generate a strong astrocytic response. CD68 (+) reactive microglia colocalizes with Fluoro-Jade C (+) injured axons in the ON (ON,
E), optic tract (OT, F), corticospinal tract (CST, G) and cerebellar lobule (cb, H) of animals exposed to repetitive mild injury at 7 days post final injury. Reactive microglia is labeled in red and
injured axons in green. Scale bars: A, B and E, 50 μm; C, D, F, G and H, 100 μm.

with abbreviated NSS testing. NSS used in our subjects is a sensitive,
specific and very popular motor and cognitive scale used for the evalu-
ation of TBI severity and recovery in mice (Flierl et al., 2009). The fact
that NSS does not detect motor or behavioral deficits in our mice is con-
sistent with the idea that our regimen causes mild TBI. However, it is
possible that subtle impairments may require more specialized testing;
for example, in the rat IA model, mild spatial learning and memory
deficits were detected with the Morris Water Maze (Beaumont et al.,
1999).
4.2. Traumatic axonal injury
Gallyas silver and Fluoro-Jade C staining indicate the presence of se-
lective axonal injury whose severity is related to the IA burden as
444 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
Fig. 7. Disruption of blood brain- barrier (BBB) in ON of IA-injured animals. BBB disruption
and primary TAI in the ON with single (A and B) and repetitive mild injury (C and D) was
examined with double immunofluorescence for mouse IgG (red) and APP (green). Eyeball
is to the left of panel, optic chiasm on the right (asterisk). Four hours after single mild in-
jury (1 × 40 g) (A), there is disruption of BBB over a large segment of ON close to the eye
(red, arrowhead). There are also axonal swellings and bulbs on the ocular (left) side of the
disrupted BBB front, evidence of primary TAI at the BBB disruption site. Inset illustrates de-
tails of a typical axonal swelling. Twenty four hours after single injury (B), IgG immunore-
activity is diminished (red, arrowhead) and abnormal axon signal is reduced as well
(green, arrows). In the case of repetitive injury (4 × 40 g) (C and D), there is little BBB dis-
ruption (red, arrowhead) and only few retraction balls form at the ocular aspect of seg-
ment with BBB disruption (green, arrows). Scale bars: A–D, 500 μm.
measured by impactor weight and number of IA events (Figs. 4 and 6).
Although the size and positioning of the metal disk could contribute to
this distribution of impact forces and therefore influence the regional
pattern of TAI observed here, similar patterns have been observed
with other blunt (Wang et al., 2011) and blast (Koliatsos et al., 2011)
TBI models. It is possible that additional factors germane to the involved
pathways may determine the regional intensity of TAI.
With the single-injury regimen, our findings indicate a correlation
between severity of axonal injury and weight of impactor, especially
evident in optic tract and cerebellar white matter (area 9Cb) (Fig. 5).
In the corpus callosum, single impacts below 50 g− 1 m do not
cause significant axonal injury; TAI is evident at the level of 60 g−1 m
(severe injury). With the repetitive mild injury regimens (4 × 40 g
and 12 × 40 g), we found a cumulative effect of IA in the optic tract
and corpus callosum (Fig. 5). However, with the silver staining used
here, we did not see a difference in TAI in the optic tract, cerebellum
and corpus callosum between the two regimens, a pattern suggesting
that TAI may plateau at a level between 4 × 40 g and 12 × 40 g. Although
TAI in corticospinal tract (CST) was not quantified, severity of impact
(as represented by impactor weight) and number of hits (the cumulative
factor) seem to also correlate with TAI severity (L. Xu, personal
observations). All these findings indicate a selective vulnerability to in-
jury within white matter tracts in our mouse IA model and also demon-
strate some correlation of TAI severity in these locations with impact
burden and number of IA hits.
4.3. Neuroinflammatory changes
TAI sites in our mild single or repetitive IA model show intense
neuroinflammatory responses involving activated microglia and
astrocytes. Activated microglia are deramified or present with round
cytologies that are typical of phagocytes and express the macrophage/
lysosomal marker CD68. This microglial response is consistent with ob-
servations in other animal models of TBI (Chen et al., 2003; Kelley et al.,
2007; Shitaka et al., 2011) and is also seen in the brains of patients with
TBI histories (Johnson et al., 2013; Smith et al., 2013; Gentleman et al.,
2004). Microglial activation in TBI can be detrimental or beneficial
(Morganti-Kossmann et al., 2002). The first corresponds to the classical
neurotoxic pathway (M1 cells); the latter involves activation in the al-
ternative neuroprotective pathway (M2 phenotype) (Martinez et al.,
2008; Loane and Byrnes, 2010). With progressive injury beyond two
weeks, microglia may transform from M2 to M1 (Hu et al., 2012). In
the early stage of TBI, M2 microglia clears debris from dead neurons
and protects vulnerable neurons by releasing trophic factors or cyto-
kines (Neumann et al., 2009). In some cases, microglial activation after
TBI persists for months or even years (Gentleman et al., 2004; Johnson
et al., 2013), a pattern indicating chronic neuroinflammation. Although
the role of this chronic neuroinflammation in TBI scenarios is not entire-
ly clear, in classical neurodegenerative diseases such amyotrophic later-
al sclerosis (ALS) and Alzheimer's disease (AD) activated (transformed)
microglia is presumed to promote further degeneration (Akiyama et al.,
2000; Liao et al., 2012). Similarly, in the hypoglossal axotomy model,
activated microglia becomes neurotoxic 14 days after injury, a transfor-
mation suppressible by minocycline (Yamada and Jinno, 2013). In
addition, suppression of microglial activation by minocycline has
shown benefits in the animals subjected to weight drop TBI (Siopi
et al., 2011, 2012a,b; Homsi et al., 2010).
The role of microglia in axonal regeneration is complex. Both bene-
ficial and detrimental effects have been reported (McPhail et al., 2004;
Prewitt et al., 1997; Donnelly and Popovich, 2008). In one study, activat-
ed microglia was shown to promote the retraction of injured axons
(Horn et al., 2008). While the role of activated microglia in the outcome
of TAI will need to be clarified with further research, the specific
association of neuroinflammation with TAI sites may serve as the basis
of neuroimaging for the diagnosis of concussion. Based on previous
pathological evidence showing inflammation in the brain of patients
with concussion (Oppenheimer, 1968) and our present findings in ani-
mals, PET ligands marking activated microglia such as [11C](R)-PK11195
or even astrocytes such as [11C]-DED can be used to localize areas of
axonal injury in patients with concussion (Banati, 2002; Carter et al.,
2012).
4.4. Degeneration of the visual system
The visual system was severely and consistently impacted in our
model and the visual pathway was the principal site of TAI along with
the corticospinal tract (Fig. 4). Injury to the visual system include the
retina, ON, optic tract, the lateral geniculate nucleus and superficial
layers of the superior colliculus and is associated with an early primary
effect of IA to the ON. The intense BBB disruption at 4 h after single inju-
ry that subsides at 24 h is consistent with an early effect at the ON level.
A corresponding trend in APP (+) axon abnormalities is also supportive
of this idea. In the rat, BBB is known to open quickly in less than 4 h and
then close after IA injury using parameters corresponding to the ones
used here (Beaumont et al., 2000). The presence of early axonal lesions
in ON segments overlapping these with BBB disruption indicates the site
of impact and is consistent with findings in the fluid percussion model
of TBI (Wang et al., 2011,). The peak of APP immunoreactivity in TAI
varies from location to location (Bramlett et al., 1997; Spain et al.,
2010). The presence of fewer APP (+) axonal abnormalities in the ON
with repetitive injury suggests that the primary traumatic effect of IA
in the ON may have reached a plateau prior to the final hit or that
 L. Xu et al. / Experimental Neurology 275 (2016) 436–449 445

Fig. 8. Retinal ganglion cell (RGC) loss after single and repetitive mild injury. In whole-mount retinas stained with the RGC marker, γ-synuclein (Sncg), repetitive mild injury (4 × 40 g and
12 × 40 g) causes significant RGC loss (C and D) at 10 weeks. Even single injured mice (1 × 40 g) have lower RGCs (B) than sham mice (A). Insets are enlargements of frames in main panels
to show more detail. Densities of surviving RGCs in the retinas of sham, single and repetitive mild injured animals are shown in (E) and (F), 7 days and 10 weeks post final injury,
respectively. Data were analyzed with one-way ANOVA followed by Tukey's post hoc test. * p b 0.05. Scale bars: A–D, 100 μm.

ongoing axonal degeneration leaves fewer residual axons exposed to in-
jury with every subsequent hit.
In the retina, no RGC loss is seen after single mild injury at 7 days, but
cell death is significant at 10 weeks and involves more than a third of
RGCs. Based on the known outcomes of ON axotomy, RGC degeneration
is most likely secondary to ON injury (Dratviman-Storobinsky et al.,
2008; Leung et al., 2008). In the case of repetitive injury, RGC loss is sig-
nificant already at 7 days and approaches 40% of RGCs in the high-
frequency repetitive regimen (12 × 40 g). At 10 weeks post injury,
more than 50% of RGCs degenerate with the 4 × 40 g regimen and 70%
of RGCs die with the 12 × 40 g regimen. In the case of repetitive injury,
severity of RGC degeneration corresponds to severity of TAI in the ON.
Neuroinflammatory response in retinas, primarily evidenced by an in-
creased number of activated microglial cells, is significant at 7 days post
injury but returns to control levels at 10 weeks in all experimental groups
exposed to either single or repetitive injury. These findings suggest that
the bulk of neuroinflammatory response in retinas has ended at 10
weeks, although it is conceivable that low-grade neuroinflammation
(and RGC degeneration) may continue, to some extent, even beyond
this time point.
This paper is a detailed report of RGC degeneration in rodents after
repetitive mild blunt TBI. We have previously presented evidence of
RGC degeneration associated with TAI in the ON-optic tract after
mild–moderate blast injury in mice (Koliatsos et al., 2011). ON injury
was also reported in a rat model of central fluid percussion injury
(Wang et al., 2011), although a subsequent study from this group did
not demonstrate RGC loss (Wang et al., 2013). In another repetitive
mild TBI model with minimal, if any, head movement (Mouzon et al.,
2012), five consecutive insults caused significant loss of brain-specific
homeobox/POU domain protein 3A (+) RGCs, thinning of the inner
retina, and decreased photopic negative response at 10–13 weeks
(Tzekov et al., 2014). These results are consistent with our findings
using the IA model and suggest that injury to visual system is not un-
common in rodent models of mild TBI. Concussion patients frequently
report blurred or double vision, problems with accommodation and
other visual symptoms (Rees, 2003; Bowen, 2003; Erlanger et al.,
446 L. Xu et al. / Experimental Neurology 275 (2016) 436–449

Fig. 9. Axonal degeneration in the ON of IA-injured animals. Axonal pathology in the ON of mice exposed to single and repetitive mild injury was examined with toluidine blue staining of
semithin sections from ON segments proximal to chiasm 10 weeks post-injury. Panels A–D illustrate type and distribution of axonal pathology. Panel E is a bar diagram plotting numbers of
surviving axons in animals subjected to sham, single and repetitive mild injury. In A–D, stained semithin sections depict increasing severity of axonal degeneration from single (B) to
repetitive (C) injury and from low-frequency (C) to high-frequency (D) repetitive injury. Panel A illustrates a sham control for the purpose of comparison. Degenerating axonal profiles
(arrows) are especially abundant at the center of the nerve, whereas the periphery is relatively spared with higher numbers of surviving axons (arrowheads) at each level of injury severity.
Axonal loss leaves spaces and cavities, especially at the center of the nerve (asterisks). Panel E illustrates the progressive nature of degeneration based on injury burden. Axons were
grouped per experimental history; averages were analyzed with one-way ANOVA followed by Tukey's post hoc testing. * p b 0.05. Scale bars: A–D, 10 μm.

1999). In addition, patients with severe TBI including blast injury have
marked visual deficits (Brahm et al., 2009; Hellerstein et al., 1995;
Summers, 2006). Such symptoms probably have complex origin and
there is no evidence that, in humans, there is TAI in visual pathways
after concussion. The involvement of the visual system in rodent models
of TBI including concussion bypasses the controversy of the presence or
not of authentic primary axonal injury in rodent species. Axonal injury
is certainly accentuated in gyrencephalic brains because of the abun-
dance of white matter tracts (Smith et al., 1997, 2003; Browne et al.,
2011). Although the rodent brain as a whole is not as vulnerable, the vi-
sual system and perhaps also some cerebellar tracts appear to be sus-
ceptible. Reasons for this phenomenon may be the relative abundance
of white matter in these systems or biomechanical peculiarities related
to distribution of shearing forces. Whatever the cause, the visual system
may be the ideal site to study primary TAI in rodent species.
The preferential injury to the visual system in our mouse model of
repetitive mild TBI, including degeneration of RGCs, may also have im-
plications for research on CTE. Compared with the corticospinal tract,
i.e. another system implicated in our model with a rather wide distribu-
tion of nerve cells of origin, the population of affected nerve cells in the
visual system is concentrated in the retina. Therefore, our model can be
used to explore the role of mild repetitive TBI in causing tauopathy in a
localized region of the nervous system that would be more straightfor-
ward to study. IHC for multiple phosphorylated tau epitopes failed to
demonstrate tau hyperphosphorylation in the brains of wild-type
mice injured as per protocols used here. In another mouse TBI model
based on mild blast exposure in a shock tube, a single blast event was
reported to have caused tauopathy (Goldstein et al., 2012). Reasons
causing this variability may include differences in injury mechanisms
between our model and theirs. The issue of phosphorylation of wild
tau in rodent species remains controversial and other investigators
have also reported negative results in blast injured rat (Gama Sosa
et al., 2014; Sosa et al., 2013) and wild type mouse (Xu et al., 2014).
Another advantage of our model is that it offers great opportunities to
explore experimental therapies for TAI. Neurons of interest in visual
TAI, i.e. RGCs, are exposed in the vitreous chamber and thus eminently
accessible for direct therapeutic manipulations with intravitreous injec-
tions of therapeutic compounds or grafts of cell preparations including
stem cells (Xin et al., 2013; Bharti et al., 2014).
In summary, the rat IA model can be adjusted for use in mild repet-
itive TBI protocols in mice. These protocols do not cause fractures or
focal contusions and also do not cause significant anoxic–ischemic
 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
Fig. 10. Neuroinflammation in the retinal ganglion cell (RGC) layer after single and repetitive mild injury. Repetitive mild injury (4 × 40 g) leads to a significantly higher number of CD68
(+) activated microglial cells (green) in the RGC layer compared to sham (B) 7 days after final injury. Confocal microscopy (C) verifies that CD68 (+) microglial cells (green) are localized
in the γ-synuclein (+) RGC layer (Sncg, red, arrows). Densities of CD68 (+) microglia in the retinas of sham, single and repetitive mild injured animals are shown in D and E at 7 days and
10 weeks post final injury, respectively. One-way ANOVA with Tukey's post hoc test is used for analysis. * p b 0.05. Scale bars: A and B, 20 μm; C, 50 μm.
injury. Instead, such mild repetitive protocols generate TAI in select CNS
sites coupled with local microglial activation and phagocytic transfor-
mation. The visual system is particularly vulnerable in these protocols.
Our model of mild repetitive injury and the involvement of the visual
system in particular support the notion that it is possible to generate
consistent primary TAI in the rodent brain. Moreover, this model can
serve for further investigations into mechanisms of mild repetitive TBI,
including tau-related neurodegenerative events in genetically vulnera-
ble subjects, and is eminently applicable to therapeutic interventions
targeting mild repetitive TBI and its chronic complications.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.expneurol.2014.11.004.
Conflict of interest
No competing financial interests exist.
Acknowledgment
This work was supported by Maryland Technology Development
Corporation funds to VEK.
References
Akiyama, H., Barger, S., Barnum, S., Bradt, B., Bauer, J., Cole, G.M., Cooper, N.R.,
Eikelenboom, P., Emmerling, M., Fiebich, B.L., Finch, C.E., Frautschy, S., Griffin, W.S.,
Hampel, H., Hull, M., Landreth, G., Lue, L., Mrak, R., Mackenzie, I.R., McGeer, P.L.,
O'Banion, M.K., Pachter, J., Pasinetti, G., Plata-Salaman, C., Rogers, J., Rydel, R., Shen,
Y., Streit, W., Strohmeyer, R., Tooyoma, I., Van Muiswinkel, F.L., Veerhuis, R.,
Walker, D., Webster, S., Wegrzyniak, B., Wenk, G., Wyss-Coray, T., 2000. Inflammation
and Alzheimer's disease. Neurobiol. Aging 21, 383–421.
Ballatore, C., Lee, V.M.Y., Trojanowski, J.Q., 2007. Tau-mediated neurodegeneration in
Alzheimer's disease and related disorders. Nat. Rev. Neurosci. 8, 663–672.
Banati, R.B., 2002. Visualising microglial activation in vivo. GLIA 40, 206–217.
Barth, J.T., 2011. Defense and Veterans Brain Injury Center. Sports Concussion and Combat
Blast Injury.
Beaumont, A., Marmarou, A., Czigner, A., Yamamoto, M., Demetriadou, K., Shirotani, T.,
Marmarou, C., Dunbar, J., 1999. The impact-acceleration model of head injury: injury
severity predicts motor and cognitive performance after trauma. Neurol. Res. 21,
742–754.
447
Beaumont, A., Marmarou, A., Hayasaki, K., Barzo, P., Fatouros, P., Corwin, F., Marmarou, C.,
Dunbar, J., 2000. The permissive nature of blood brain barrier (BBB) opening in
edema formation following traumatic brain injury. Acta Neurochir. Suppl. 76,
125–129.
Bharti, K., Rao, M., Hull, S.C., Stroncek, D., Brooks, B.P., Feigal, E., van Meurs, J.C., Huang,
C.A., Miller, S.S., 2014. Developing cellular therapies for retinal degenerative diseases.
Invest. Ophthalmol. Vis. Sci. 55, 1191–1202.
Bian, G.L., Wei, L.C., Shi, M., Wang, Y.Q., Cao, R., Chen, L.W., 2007. Fluoro-Jade C can
specifically stain the degenerative neurons in the substantia nigra of the 1-
methyl-4-phenyl-1,2,3,6-tetrahydro pyridine-treated C57BL/6 mice. Brain Res.
1150, 55–61.
Bowen, A.P., 2003. Second impact syndrome: a rare, catastrophic, preventable complica-
tion of concussion in young athletes. J. Emerg. Nurs. 29, 287–289.
Brahm, K.D., Wilgenburg, H.M., Kirby, J., Ingalla, S., Chang, C.Y., Goodrich, G.L., 2009. Visual
impairment and dysfunction in combat-injured servicemembers with traumatic
brain injury. Optom. Vis. Sci. 86, 817–825.
Bramlett, H.M., Kraydieh, S., Green, E.J., Dietrich, W.D., 1997. Temporal and regional pat-
terns of axonal damage following traumatic brain injury: a beta-amyloid precursor
protein immunocytochemical study in rats. J. Neuropathol. Exp. Neurol. 56,
1132–1141.
Brody, D.L., Mac, D.C., Kessens, C.C., Yuede, C., Parsadanian, M., Spinner, M., Kim, E.,
Schwetye, K.E., Holtzman, D.M., Bayly, P.V., 2007. Electromagnetic controlled cortical
impact device for precise, graded experimental traumatic brain injury. J. Neurotrauma
24, 657–673.
Browne, K.D., Chen, X.H., Meaney, D.F., Smith, D.H., 2011. Mild traumatic brain injury and
diffuse axonal injury in swine. J. Neurotrauma 28, 1747–1755.
Bu, G., 2009. Apolipoprotein E and its receptors in Alzheimer's disease: pathways,
pathogenesis and therapy. Nat. Rev. Neurosci. 10, 333–344.
Carter, S.F., Scholl, M., Almkvist, O., Wall, A., Engler, H., Langstrom, B., Nordberg, A., 2012.
Evidence for astrocytosis in prodromal Alzheimer disease provided by 11C-
deuterium-L-deprenyl: a multitracer PET paradigm combining 11C-Pittsburgh
compound B and 18F-FDG. J. Nucl. Med. 53, 37–46.
Cernak, I., Merkle, A.C., Koliatsos, V.E., Bilik, J.M., Luong, Q.T., Mahota, T.M., Xu, L.Y., Slack,
N., Windle, D., Ahmed, F.A., 2011. The pathobiology of blast injuries and blast-induced
neurotrauma as identified using a new experimental model of injury in mice.
Neurobiol. Dis. 41, 538–551.
Chen, S., Pickard, J.D., Harris, N.G., 2003. Time course of cellular pathology after controlled
cortical impact injury. Exp. Neurol. 182, 87–102.
Chidlow, G., Wood, J.P., Sarvestani, G., Manavis, J., Casson, R.J., 2009. Evaluation of Fluoro-
Jade C as a marker of degenerating neurons in the rat retina and optic nerve. Exp. Eye
Res. 88, 426–437.
Conte, V., Uryu, K., Fujimoto, S., Yao, Y., Rokach, J., Longhi, L., Trojanowski, J.Q., Lee, V.M.Y.,
McIntosh, T.K., Pratico, D., 2004. Vitamin E reduces amyloidosis and improves cogni-
tive function in Tg2576 mice following repetitive concussive brain injury. J.
Neurochem. 90, 758–764.
Corsellis, J.A.N., Bruton, C.J., Freeman-Browne, D., 1973. The aftermath of boxing. Psychol.
Med. 3, 270–303.
448 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
Creeley, C.E., Wozniak, D.F., Bayly, P.V., Olney, J.W., Lewis, L.M., 2004. Multiple episodes of
mild traumatic brain injury result in impaired cognitive performance in mice. Acad.
Emerg. Med. 11, 809–819.
Deford, S.M., Wilson, M.S., Rice, A.C., Clausen, T., Rice, L.K., Barabnova, A., Bullock, R.,
Hamm, R.J., 2002. Repeated mild brain injuries result in cognitive impairment in
B6C3F1 mice. J. Neurotrauma 19, 427–438.
Donnelly, D.J., Popovich, P.G., 2008. Inflammation and its role in neuroprotection, axonal
regeneration and functional recovery after spinal cord injury. Exp. Neurol. 209,
378–388.
Dratviman-Storobinsky, O., Hasanreisoglu, M., Offen, D., Barhum, Y., Weinberger, D.,
Goldenberg-Cohen, N., 2008. Progressive damage along the optic nerve following in-
duction of crush injury or rodent anterior ischemic optic neuropathy in transgenic
mice. Mol. Vis. 14, 2171–2179.
Erlanger, D.M., Kutner, K.C., Barth, J.T., Barnes, R., 1999. Neuropsychology of sports-related
head injury: dementia pugilistica to post concussion syndrome. Clin. Neuropsychol.
13, 193–209.
Ferrari, R., Hardy, J., Momeni, P., 2011. Frontotemporal dementia: from Mendelian genet-
ics towards genome wide association studies. J. Mol. Neurosci. 45, 500–515.
Flierl, M.A., Stahel, P.F., Beauchamp, K.M., Morgan, S.J., Smith, W.R., Shohami, E., 2009.
Mouse closed head injury model induced by a weight-drop device. Nat. Protoc. 4,
1328–1337.
Foda, M.A.A., Marmarou, A., 1994. A new model of diffuse brain injury in rats. 2. Morpho-
logical characterization. J. Neurosurg. 80, 301–313.
Fujita, M., Wei, E.P., Povlishock, J.T., 2012. Intensity- and interval-specific repetitive trau-
matic brain injury can evoke both axonal and microvascular damage. J. Neurotrauma
29, 2172–2180.
Gama Sosa, M.A., De Gasperi, R., Janssen, P.L., Yuk, F.J., Anazodo, P.C., Pricop, P.E., Paulino,
A.J., Wicinski, B., Shaughness, M.C., Maudlin-Jeronimo, E., Hall, A.A., Dickstein, D.L.,
McCarron, R.M., Chavko, M., Hof, P.R., Ahlers, S.T., Elder, G.A., 2014. Selective vulner-
ability of the cerebral vasculature to blast injury in a rat model of mild traumatic
brain injury. Acta Neuropathol. Commun. 2, 67.
Gentleman, S.M., Leclercq, P.D., Moyes, L., Graham, D.I., Smith, C., Griffin, W.S., Nicoll, J.A.,
2004. Long-term intracerebral inflammatory response after traumatic brain injury.
Forensic Sci. Int. 146, 97–104.
Goldstein, L.E., Fisher, A.M., Tagge, C.A., Zhang, X.L., Velisek, L., Sullivan, J.A., Upreti, C.,
Kracht, J.M., Ericsson, M., Wojnarowicz, M.W., Goletiani, C.J., Maglakelidze, G.M.,
Casey, N., Moncaster, J.A., Minaeva, O., Moir, R.D., Nowinski, C.J., Stern, R.A., Cantu,
R.C., Geiling, J., Blusztajn, J.K., Wolozin, B.L., Ikezu, T., Stein, T.D., Budson, A.E.,
Kowall, N.W., Chargin, D., Sharon, A., Saman, S., Hall, G.F., Moss, W.C., Cleveland,
R.O., Tanzi, R.E., Stanton, P.K., McKee, A.C., 2012. Chronic traumatic encephalopathy
in blast-exposed military veterans and a blast neurotrauma mouse model. Sci. Transl.
Med. 4, 134ra60.
Greer, J.E., McGinn, M.J., Povlishock, J.T., 2011. Diffuse traumatic axonal injury in the
mouse induces atrophy, c-Jun activation, and axonal outgrowth in the axotomized
neuronal population. J. Neurosci. 31, 5089–5105.
Hartman, R.E., Izumi, Y., Bales, K.R., Paul, S.M., Wozniak, D.F., Holtzman, D.M., 2005. Treat-
ment with an amyloid-beta antibody ameliorates plaque load, learning deficits, and
hippocampal long-term potentiation in a mouse model of Alzheimer's disease. J.
Neurosci. 25, 6213–6220.
Hellerstein, L.F., Freed, S., Maples, W.C., 1995. Vision profile of patients with mild brain in-
jury. J. Am. Optom. Assoc. 66, 634–639.
Homsi, S., Piaggio, T., Croci, N., Noble, F., Plotkine, M., Marchand-Leroux, C., Jafarian-
Tehrani, M., 2010. Blockade of acute microglial activation by minocycline promotes
neuroprotection and reduces locomotor hyperactivity after closed head injury in
mice: a twelve-week follow-up study. J. Neurotrauma 27, 911–921.
Horn, K.P., Busch, S.A., Hawthorne, A.L., van, R.N., Silver, J., 2008. Another barrier to regen-
eration in the CNS: activated macrophages induce extensive retraction of dystrophic
axons through direct physical interactions. J. Neurosci. 28, 9330–9341.
Hu, X., Li, P., Guo, Y., Wang, H., Leak, R.K., Chen, S., Gao, Y., Chen, J., 2012. Microglia/
macrophage polarization dynamics reveal novel mechanism of injury expansion
after focal cerebral ischemia. Stroke 43, 3063–3070.
Jefferson, S.C., Tester, N.J., Howland, D.R., 2011. Chondroitinase ABC promotes recovery
of adaptive limb movements and enhances axonal growth caudal to a spinal
hemisection. J. Neurosci. 31, 5710–5720.
Johnson, V.E., Stewart, J.E., Begbie, F.D., Trojanowski, J.Q., Smith, D.H., Stewart, W., 2013.
Inflammation and white matter degeneration persist for years after a single traumatic
brain injury. Brain 136, 28–42.
Kelley, B.J., Lifshitz, J., Povlishock, J.T., 2007. Neuroinflammatory responses after experi-
mental diffuse traumatic brain injury. J. Neuropathol. Exp. Neurol. 66, 989–1001.
Koliatsos, V.E., Cernak, I., Xu, L., Song, Y., Savonenko, A., Crain, B.J., Eberhart, C.G.,
Frangakis, C.E., Melnikova, T., Kim, H., Lee, D., 2011. A mouse model of blast injury
to brain: initial pathological, neuropathological, and behavioral characterization. J.
Neuropathol. Exp. Neurol. 70, 399–416.
Langlois, J.A., Rutland-Brown, W., Wald, M.M., 2006. The epidemiology and impact of
traumatic brain injury: a brief overview. J. Head Trauma Rehabil. 21, 375–378.
Laurer, H.L., Bareyre, F.M., Lee, V.M.Y.C., Trojanowski, J.Q., Longhi, L., Hoover, R., Saatman,
K.E., Raghupathi, R., Hoshino, S., Grady, M.S., McIntosh, T.K., 2001. Mild head injury
increasing the brain's vulnerability to a second concussive impact. J. Neurosurg. 95,
859–870.
Leung, C.K., Lindsey, J.D., Crowston, J.G., Lijia, C., Chiang, S., Weinreb, R.N., 2008.
Longitudinal profile of retinal ganglion cell damage after optic nerve crush
with blue-light confocal scanning laser ophthalmoscopy. Invest. Ophthalmol.
Vis. Sci. 49, 4898–4902.
Liao, B., Zhao, W., Beers, D.R., Henkel, J.S., Appel, S.H., 2012. Transformation from a neuro-
protective to a neurotoxic microglial phenotype in a mouse model of ALS. Exp.
Neurol. 237, 147–152.
Loane, D.J., Byrnes, K.R., 2010. Role of microglia in neurotrauma. Neurotherapeutics 7,
366–377.
Longhi, L., Saatman, K.E., Fujimoto, S., Raghupathi, R., Meaney, D.F., Davis, J., McMillan, A.,
Conte, V., Laurer, H.L., Stein, S., Stocchetti, N., McIntosh, T.K., 2005. Temporal window
of vulnerability to repetitive experimental concussive brain injury. Neurosurgery 56,
364–373.
Marmarou, A., Foda, M.A.A., Vandenbrink, W., Campbell, J., Kita, H., Demetriadou, K., 1994.
A new model of diffuse brain injury in rats. 1. Pathophysiology and biomechanics. J.
Neurosurg. 80, 291–300.
Martinez, F.O., Sica, A., Mantovani, A., Locati, M., 2008. Macrophage activation and polar-
ization. Front. Biosci. 13, 453–461.
Martland, H.S., 1928. Punch drunk. JAMA 91, 1103–1107.
McKee, A.C., Cantu, R.C., Nowinski, C.J., Hedley-Whyte, E.T., Gavett, B.E., Budson, A.E.,
Santini, V.E., Lee, H.S., Kubilus, C.A., Stern, R.A., 2009. Chronic traumatic encephalop-
athy in athletes: progressive tauopathy after repetitive head injury. J. Neuropathol.
Exp. Neurol. 68, 709–735.
McPhail, L.T., Stirling, D.P., Tetzlaff, W., Kwiecien, J.M., Ramer, M.S., 2004. The contribution
of activated phagocytes and myelin degeneration to axonal retraction/dieback fol-
lowing spinal cord injury. Eur. J. Neurosci. 20, 1984–1994.
Millspaugh, 1937. Dementia pugilistica. U.S. Naval Med. Bull. 35, 297–303.
Morganti-Kossmann, M.C., Rancan, M., Stahel, P.F., Kossmann, T., 2002. Inflammatory re-
sponse in acute traumatic brain injury: a double-edged sword. Curr. Opin. Crit. Care 8,
101–105.
Mouzon, B., Chaytow, H., Crynen, G., Bachmeier, C., Stewart, J., Mullan, M., Stewart, W.,
Crawford, F., 2012. Repetitive mild traumatic brain injury in a mouse model produces
learning and memory deficits accompanied by histological changes. J. Neurotrauma
29, 2761–2773.
Neumann, H., Kotter, M.R., Franklin, R.J., 2009. Debris clearance by microglia: an essential
link between degeneration and regeneration. Brain 132, 288–295.
Ojo, J.O., Mouzon, B., Greenberg, M.B., Bachmeier, C., Mullan, M., Crawford, F., 2013. Repet-
itive mild traumatic brain injury augments tau pathology and glial activation in aged
hTau mice. J. Neuropathol. Exp. Neurol. 72, 137–151.
Omalu, B.I., DeKosky, S.T., Minster, R.L., Kamboh, M.I., Hamilton, R.L., Wecht, C.H., 2005.
Chronic traumatic encephalopathy in a National Football League player. Neurosur-
gery 57, 128–133.
Omalu, B.I., DeKosky, S.T., Hamilton, R.L., Minster, R.L., Kamboh, M.I., Shakir, A.M., Wecht,
C.H., 2006. Chronic traumatic encephalopathy in a national football league player:
part II. Neurosurgery 59, 1086–1092.
Oppenheimer, D.R., 1968. Microscopic lesions in the brain following head injury. J. Neurol.
Neurosurg. Psychiatry 31, 299–306.
Pohl, H.B., Porcheri, C., Mueggler, T., Bachmann, L.C., Martino, G., Riethmacher, D.,
Franklin, R.J., Rudin, M., Suter, U., 2011. Genetically induced adult oligodendrocyte
cell death is associated with poor myelin clearance, reduced remyelination, and axo-
nal damage. J. Neurosci. 31, 1069–1080.
Prewitt, C.M., Niesman, I.R., Kane, C.J., Houle, J.D., 1997. Activated macrophage/microglial
cells can promote the regeneration of sensory axons into the injured spinal cord. Exp.
Neurol. 148, 433–443.
Rees, P.M., 2003. Contemporary issues in mild traumatic brain injury. Arch. Phys. Med.
Rehabil. 84, 1885–1894.
Schmued, L.C., Albertson, C., Slikker Jr., W., 1997. Fluoro-Jade: a novel fluorochrome for
the sensitive and reliable histochemical localization of neuronal degeneration. Brain
Res. 751, 37–46.
Schmued, L.C., Stowers, C.C., Scallet, A.C., Xu, L., 2005. Fluoro-Jade C results in ultra
high resolution and contrast labeling of degenerating neurons. Brain Res. 1035,
24–31.
Shitaka, Y., Tran, H.T., Bennett, R.E., Sanchez, L., Levy, M.A., Dikranian, K., Brody, D.L.,
2011. Repetitive closed-skull traumatic brain injury in mice causes persistent
multifocal axonal injury and microglial reactivity. J. Neuropathol. Exp. Neurol.
70, 551–567.
Siopi, E., Cho, A.H., Homsi, S., Croci, N., Plotkine, M., Marchand-Leroux, C., Jafarian-Tehrani,
M., 2011. Minocycline restores sAPPalpha levels and reduces the late histopatholog-
ical consequences of traumatic brain injury in mice. J. Neurotrauma 28, 2135–2143.
Siopi, E., Calabria, S., Plotkine, M., Marchand-Leroux, C., Jafarian-Tehrani, M., 2012a.
Minocycline restores olfactory bulb volume and olfactory behavior after traumatic
brain injury in mice. J. Neurotrauma 29, 354–361.
Siopi, E., Llufriu-Daben, G., Fanucchi, F., Plotkine, M., Marchand-Leroux, C., Jafarian-
Tehrani, M., 2012b. Evaluation of late cognitive impairment and anxiety states fol-
lowing traumatic brain injury in mice: the effect of minocycline. Neurosci. Lett. 511,
110–115.
Smith, D.H., Chen, X.H., Xu, B.N., McIntosh, T.K., Gennarelli, T.A., Meaney, D.F., 1997. Char-
acterization 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., Meaney, D.F., Shull, W.H., 2003. Diffuse axonal injury in head trauma. J. Head
Trauma Rehabil. 18, 307–316.
Smith, C., Gentleman, S.M., Leclercq, P.D., Murray, L.S., Griffin, W.S., Graham, D.I., Nicoll,
J.A., 2013. The neuroinflammatory response in humans after traumatic brain injury.
Neuropathol. Appl. Neurobiol. 39, 654–666.
Sosa, M.A., de Gasperi, R., Paulino, A.J., Pricop, P.E., Shaughness, M.C., Maudlin-Jeronimo, E.,
Hall, A.A., Janssen, W.G., Yuk, F.J., Dorr, N.P., Dickstein, D.L., McCarron, R.M., Chavko, M.,
Hof, P.R., Ahlers, S.T., Elder, G.A., 2013. Blast overpressure induces shear-related inju-
ries in the brain of rats exposed to a mild traumatic brain injury. Acta Neuropathol.
Commun. 1, 51.
Soto, I., Oglesby, E., Buckingham, B.P., Son, J.L., Roberson, E.D., Steele, M.R., Inman, D.M.,
Vetter, M.L., Horner, P.J., Marsh-Armstrong, N., 2008. Retinal ganglion cells downreg-
ulate gene expression and lose their axons within the optic nerve head in a mouse
glaucoma model. J. Neurosci. 28, 548–561.
 L. Xu et al. / Experimental Neurology 275 (2016) 436–449
Spain, A., Daumas, S., Lifshitz, J., Rhodes, J., Andrews, P.J., Horsburgh, K., Fowler, J.H., 2010.
Mild fluid percussion injury in mice produces evolving selective axonal pathology
and cognitive deficits relevant to human brain injury. J. Neurotrauma 27, 1429–1438.
Strittmatter, W.J., Saunders, A.M., Schmechel, D., Pericak-Vance, M., Enghild, J., Salvesen,
G.S., Roses, A.D., 1993. Apolipoprotein E: high-avidity binding to b-amyloid and in-
creased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc.
Natl. Acad. Sci. U. S. A. 90, 1977–1981.
Summers, M.J., 2006. Increased inattentional blindness in severe traumatic brain injury:
evidence for reduced distractibility? Brain Inj. 20, 51–60.
Turgut, M., Kaplan, S., Unal, B.Z., Bozkurt, M., Yuruker, S., Yenisey, C., Sahin, B., Uyanikgil,
Y., Baka, M., 2010. Stereological analysis of sciatic nerve in chickens following neona-
tal pinealectomy: an experimental study. J. Brachial. Plex. Peripher. Nerve Inj. 5, 10.
Tzekov, R., Quezada, A., Gautier, M., Biggins, D., Frances, C., Mouzon, B., Jamison, J., Mullan,
M., Crawford, F., 2014. Repetitive mild traumatic brain injury causes optic nerve and
retinal damage in a mouse model. J. Neuropathol. Exp. Neurol. 73, 345–361.
Uryu, K., Laurer, H., McIntosh, T., Pratico, D., Martinez, D., Leight, S., Lee, V.M., Trojanowski,
J.Q., 2002. Repetitive mild brain trauma accelerates Abeta deposition, lipid peroxida-
tion, and cognitive impairment in a transgenic mouse model of Alzheimer amyloid-
osis. J. Neurosci. 22, 446–454.
Vandrovcova, J., Anaya, F., Kay, V., Lees, A., Hardy, J., de Silva, R., 2010. Disentangling the
role of the tau gene locus in sporadic tauopathies. Curr. Alzheimer Res. 7, 726–734.
Wang, Y.P., Biernat, J., Pickhardt, M., Mandelkow, E., Mandelkow, E.M., 2007. Stepwise
proteolysis liberates tau fragments that nucleate the Alzheimer-like aggregation of
full-length tau in a neuronal cell model. Proc. Natl. Acad. Sci. U. S. A. 104,
10252–10257.
449
Wang, J., Hamm, R.J., Povlishock, J.T., 2011. Traumatic axonal injury in the optic nerve:
evidence for axonal swelling, disconnection, dieback, and reorganization. J.
Neurotrauma 28, 1185–1198.
Wang, J., Fox, M.A., Povlishock, J.T., 2013. Diffuse traumatic axonal injury in the optic
nerve does not elicit retinal ganglion cell loss. J. Neuropathol. Exp. Neurol. 72,
768–781.
Warden, D., 2006. Military TBI during the Iraq and Afghanistan wars. J. Head Trauma
Rehabil. 21, 398–402.
Xin, X., Rodrigues, M., Umapathi, M., Kashiwabuchi, F., Ma, T., Babapoor-Farrokhran, S.,
Wang, S., Hu, J., Bhutto, I., Welsbie, D.S., Duh, E.J., Handa, J.T., Eberhart, C.G., Lutty,
G., Semenza, G.L., Montaner, S., Sodhi, A., 2013. Hypoxic retinal Muller cells promote
vascular permeability by HIF-1-dependent up-regulation of angiopoietin-like 4. Proc.
Natl. Acad. Sci. U. S. A. 110, E3425–E3434.
Xu, L., Ryugo, D.K., Pongstaporn, T., Johe, K., Koliatsos, V.E., 2009. Human neural stem cell
grafts in the spinal cord of SOD1 transgenic rats: differentiation and structural inte-
gration into the segmental motor circuitry. J. Comp. Neurol. 514, 297–309.
Xu, H., Rosler, T.W., Carlsson, T., Bruch, J., de Andrade, A., Hollerhage, M., Oertel, W.H.,
Hoglinger, G.U., 2014. Memory deficits correlate with tau and spine pathology in
P301S MAPT transgenic mice. Neuropathol. Appl. Neurobiol. 40 (7), 833–843.
Yamada, J., Jinno, S., 2013. Novel objective classification of reactive microglia following
hypoglossal axotomy using hierarchical cluster analysis. J. Comp. Neurol. 521,
1184–1201.