Received: 27 February 2018 Revised: 18 May 2018 Accepted: 22 May 2018

DOI: 10.1002/glia.23469

RESEARCH ARTICLE

Comparison of the detrimental features of microglia

and infiltrated macrophages in traumatic brain injury:
A study using a hypnotic bromovalerylurea
Naoki Abe1,2 | Mohammed E. Choudhury1 | Minori Watanabe1 | Shun Kawasaki1,3 |
Tasuku Nishihara2 | Hajime Yano1 | Shirabe Matsumoto4 | Takehiro Kunieda4 |
Yoshiaki Kumon4 | Toshihiro Yorozuya2 | Junya Tanaka1

1
Department of Molecular and Cellular
Physiology, Graduate School of Medicine,
Ehime University, Toon, Ehime, Japan
2
Department of Anesthesia and Perioperative
Medicine, Graduate School of Medicine, Ehime
University, Toon, Ehime, Japan
3
Department of Surgery and Palliative
Medicine, Fujita Health University School of
Medicine, Toyoake, Aichi, Japan
4
Department of Neurosurgery, Graduate
School of Medicine, Ehime University, Toon,
Ehime, Japan
Correspondence
Junya Tanaka, Department of Molecular and
Cellular Physiology, Graduate School of
Medicine, Ehime University, Ehime 791-0295,
Japan.
Email: jtanaka@m.ehime-u.ac.jp
Funding information
Ministry of Education; Ehime University; Ehime
University
Abstract
Microglia and blood-borne macrophages in injured or diseased brains are difficult to distinguish
because they share many common characteristics. However, the identification of microglia-
specific markers and the use of flow cytometry have recently made it easy to discriminate these
types of cells. In this study, we analyzed the features of blood-borne macrophages, and activated
and resting microglia in a rat traumatic brain injury (TBI) model. Oxidative injury was indicated in
macrophages and neurons in TBI lesions by the presence of 8-hydroxy-2’-deoxyguanosine
(8-OHdG). Generation of mitochondrial reactive oxygen species (ROS) was markedly observed in
granulocytes and macrophages, but not in activated or resting microglia. Dihydroethidium staining
supported microglia not being the major source of ROS in TBI lesions. Furthermore, macrophages
expressed NADPH oxidase 2, interleukin-1β (IL-1β), and CD68 at higher levels than microglia. In
contrast, microglia expressed transforming growth factor β1 (TGFβ1), interleukin-6 (IL-6), and
tumor necrosis factor α at higher levels than macrophages. A hypnotic, bromovalerylurea (BU),
which has anti-inflammatory effects, reduced both glycolysis and mitochondrial oxygen consump-
tion. BU administration inhibited chemokine CCL2 expression, accumulation of monocytes/mac-
rophages, 8-OHdG generation, mitochondrial ROS generation, and proinflammatory cytokine
expression, and markedly ameliorated the outcome of the TBI model. Yet, BU did not inhibit
microglial activation or expression of TGFβ1 and insulin-like growth factor 1 (IGF-1). These results
indicate that macrophages are the major aggravating cell type in TBI lesions, in particular during
the acute phase. Activated microglia may even play favorable roles. Reduction of cellular energy
metabolism in macrophages and suppression of CCL2 expression in injured tissue may lead to
amelioration of TBI.

KEYWORDS

CCL2, FACS, glycolysis, mitochondria, oxidative stress, ROS

1 | I N T RO D UC T I O N chronic traumatic encephalopathy and Alzheimer’s disease. Although

Traumatic brain injury (TBI) is a leading cause of incurable neurological
disability that affects several million people in the world every year
(Corps, Roth, & McGavern, 2015; Gyoneva & Ransohoff, 2015; Nell &
Brown, 1991). Most TBI patients suffer from long-lasting neurological
problems, such as personality changes, deficits in learning and mem-
ory, and an increased risk of neurodegenerative changes, including
the primary injury in TBI is initiated by mechanical force, the second-
ary injury of subsequent neuroinflammatory events strongly aggra-
vates the outcomes of TBI, in which infiltrated leukocytes and
resident microglia play critical roles (Gyoneva & Ransohoff, 2015;
McKee & Lukens, 2016; Webster et al., 2017).
Shortly after TBI, brain parenchymal cells around the lesion start
to degenerate and then leukocytes infiltrate the lesion while exerting

Glia. 2018;1–16. wileyonlinelibrary.com/journal/glia © 2018 Wiley Periodicals, Inc. 1

2 ABE ET AL.

deleterious effects (Gyoneva & Ransohoff, 2015; Matsumoto et al., of Medicine. Male Wistar rats (7- to 8-weeks old, body weight
2007; Semple, Bye, Rancan, Ziebell, & Morganti-Kossmann, 2010). 250–270 g) were housed under standard laboratory conditions.
Furthermore, resident microglia that rapidly become activated in
response to brain injury are considered to play aggravating roles 2.2 | TBI and BU administration
(Bachstetter et al., 2013; Zhang et al., 2012). However, some studies
Stab wound injuries were made in the rat brain as described else-
demonstrate that infiltrated macrophages (Nishihara et al., 2011) and
where (Yokoyama, Sakamoto, Kameda, Imai, & Tanaka, 2006). In brief,
activated microglia (Act-MG) (Roth et al., 2014) have beneficial effects
rats were anesthetized with isoflurane, and a 15 mm longitudinal inci-
on damaged tissues and cells. Because of the conflicting data, the
sion was made in the skin to expose the skull. Two holes were made
term “double-edged sword” (Patel, Ritzel, McCullough, & Liu, 2013)
through the skull over the right hemisphere at approximately 2.5 and
has been used for the roles of these cells in injured brains.
4 mm right of the midline and 1 mm posterior to the bregma. Through
Act-MG are difficult to distinguish from infiltrated macrophages
each hole, a 26-gauge needle was inserted to a depth of approxi-
because of common functions and markers and similar morphological
mately 7 mm from the surface of the skull and the needle was moved
characteristics (Bennett et al., 2016). Therefore, the notion that micro-
in a fan like manner from the anterior then to the posterior in parallel
glia or macrophages have both ameliorative and destructive actions
with the midline. Thereafter, the needle was withdrawn, and the skin
might just reflect a confusion between the two types of cells.
incision was closed with quick-drying glue (Aron-Alpha; Toagosei,
Recently, microglia-specific proteins have been identified: transmem-
Tokyo, Japan). One hour after wounding, BU (1 mg/ml) dissolved in
brane protein 119 (TMEM119) (Bennett et al., 2016; Satoh et al.,
propylene glycol was subcutaneously injected at a dose of
2016) and Siglec-H (Konishi et al., 2017). These specific markers for
30 mg/kg. For control rats, propylene glycol only was injected.
microglia will bring about marked progress in understanding the roles
Wounded rats were then allowed to freely drink. Control rats were
of microglia in brain pathology.
given water and BU-treated rats were given water containing BU
Another method has recently been employed to distinguish macro-
(500 mg/l) for 7 days. Rats drank approximately 25 ml water per day;
phages and microglia from the brain; the combination of brain tissue
therefore, the daily BU dose was 50 mg/kg, which was comparable to
dissociation and flow cytometry (Hattori et al., 2017). Microglia express
the maximum clinically applicable dose. The Stroke Therapy Academic
CD45 at lower levels than macrophages, which is in contrast to
Industry Roundtable (STAIR) preclinical recommendations state that
CD11b, which is expressed at similar levels by both cell types. By the
multiple endpoints should be determined for anatomical and behav-
simultaneous use of antibodies to CD11b and CD45 after dissociation
ioral outcome at least 2–3 weeks or longer after onset of stroke to
of brain tissues into individual cells, flow cytometry can effectively dis-
determine the sustained benefit of experimental drugs (Fisher et al.,
criminate the two cell types from the brain. Furthermore, individually
2009). We, therefore, pursued the course of TBI for 15 weeks, using
sorted cells can be analyzed by quantitative real-time RT-PCR (qPCR).
magnetic resonance imaging (MRI) to measure cavity volumes and by
We have previously found marked anti-inflammatory effects of an
performing behavioral tests and anatomical evaluation at the endpoint
established hypnotic/sedative, bromovalerylurea (BU; C6H11BrN2O2,
as described below.
CAS: 496-67-3) (Higaki et al., 2016; Kikuchi et al., 2015). BU inhibits
CCL2 expression by lipopolysaccharide (LPS)-treated peritoneal macro-
phages (Kikuchi et al., 2015) and proinflammatory activation of microglia 2.3 | Measurement of lesion-induced cavity size
(Higaki et al., 2016). The antiinflammatory effects were at least partly
The stab wound injury caused extensive tissue loss. The resultant cav-
attributable to its suppressive effects on mitochondrial ATP production
ity volume was measured using the following two methods.
(Kawasaki et al., 2017). In this study using a rat stab-wound TBI model,
BU produced marked long-term amelioration of neurological deficits 2.3.1 | Magnetic resonance imaging
while reducing brain tissue loss. Furthermore, BU reduced the mitochon-
As described previously in Nishioka et al. (2016), MR images of
drial oxygen consumption rate (OCR) and glycolysis activity.
wounded rat heads were acquired with a 1.5 T MRminiSA (DS Pharma
By employing flow cytometry to discriminate microglia from mac-
Biomedical, Osaka, Japan) at 5 and 10 weeks after injury. A
rophages and by exploiting the marked ameliorative effects of BU, we
T1-weighted MR sequence was used with 1-mm slice thickness (total
differentially analyzed the contribution of blood-borne macrophages
13 slices). The lesion cavity area was calculated from the MR images
and microglia to the pathogenesis of TBI. Although microglia undergo
by marking low intensity areas, and the cavity volume was calculated
degeneration during the acute phase of TBI (Matsumoto et al., 2007),
by taking into account the slice thickness.
they may be ameliorating rather than aggravating compared with infil-
trated macrophages. 2.3.2 | Slices of fixed brains
Wounded rat brains were dissected out 15 weeks after TBI and
immersed in phosphate buffered saline (PBS) containing 4% parafor-
2 | MATERIALS AND METHODS
maldehyde and 2 mM MgCl2 for 7 days. Then, seven 2-mm thick
slices, centered on the injury site, were cut from each brain and
2.1 | Animals photographed. The images were processed to black (cavity) and white
All animal experiments were carried out in accordance with the Guide- (remaining brain tissue) images using Adobe Photoshop CS5 Extended
lines for Animal Experimentation of Ehime University Graduate School (Adobe Systems, San Jose, CA). The remaining brain tissue volume in
ABE ET AL.
the right hemisphere was calculated by multiplying the area by 2 mm
(thickness of each slice). The percentage of lost volume was obtained
by dividing the right remaining brain tissue volume by the volume of
the left hemisphere.
2.4 | Behavioral assessments
Learning and memory abilities after TBI were evaluated using Morris
water maze and passive avoidance tests.
2.4.1 | Morris water maze test
Spatial memory was assessed using a standard Morris water maze,
essentially as described elsewhere (Liu, Zhang, Wu, Wu, & Wang, 2013).
The animals were tested in a 150-cm diameter × 45-cm tall circular pool
filled with water at 24
C. A 12-cm diameter circular transparent plat-
form was placed 2 cm below the waterline in the center of one quadrant.
BU-treated and non-treated TBI rats were tested. Four different signs
were put on the walls of the quadrants (N, S, E, and W). On Day 0, the
platform was indicated with a small flag. Rats were forced to swim freely
twice for 60 s from points S and E of the pool to find the platform with
the flag. To memorize the location of the platform, rats were made to
stay on it for 10 s having successfully found the platform and for 15 s
when the rat could not find the platform. On Days 1 and 2, the flag was
removed and rats were given one trial for 60 s to find the platform. The
number of rats accomplishing the task was recorded and the mean
latency to the platform, total distance swum, and the swim speed was
measured for each trial using a video-tracking system (Ethovision XT
7, Noldus Info. Tech., Wageningen, The Netherlands).
2.4.2 | Passive avoidance test
Behavioral changes were also evaluated using a step-through passive
avoidance apparatus (Shuttle box; MPB-M001; Melquest, Toyama,
Japan) as described elsewhere (Lashgari, Motamedi, Zahedi Asl, Sha-
hidi, & Komaki, 2006). The apparatus consisted of an illuminated cham-
ber made of transparent plastic and a dark chamber made of dark
plastic. The size of both chambers was the same (45 × 27 × 26 cm).
The floors of both chambers were made of stainless steel rods and the
dark chamber floor was equipped with an electric shock generator (SG-
100; Melquest). A rectangular opening with an opaque guillotine door
was located between the two chambers. First, individual rats were put
in the light chamber and allowed to freely explore it for 30 s. Then, the
door was opened, and the latency for rats to completely enter the dark
chamber was measured (up to 300 s). When the rats were habituated
to enter the dark chamber within 20 s, the training was terminated.
Then, rats were again placed in the light chamber, and the door was
opened 30 sec later. When the rats completely entered the dark cham-
ber, a 1 mA electric shock was applied for 3 s. Twenty-four hours later,
trained rats were again placed into the light chamber, and the latency
to enter the dark chamber was measured (ceiling score 300 s).
2.5 | Quantitative real time RT-PCR
At 1.5 or 7 days postinjury (dpi) TBI rats were decapitated under deep
anesthesia, and their brains were dissected. The 2-mm thick coronal
sections centered on the wound holes were prepared. Rectangular
3
areas of injured tissue around the stab wounds were dissected
(Figure 2b). Total RNA was extracted from the tissue homogenates
using an RNeasy Lipid Tissue Mini Kit (Qiagen, Hilden, Germany). The
quality and quantity of purified RNA was examined using a NanoDrop
ND-2000 spectrophotometer (Thermo Fisher Scientific, Wilmington,
DE). cDNA samples were prepared from 0.5 μg RNA using ReverTra
Ace qPCR RT Master Mix with a gDNA remover kit (Toyobo, Osaka,
Japan) and were then diluted three-fold. qPCR analysis was then per-
formed in triplicate using an MJ mini instrument (BioRad, Hercules,
CA) and Fast Start Universal SYBR Green Master (Roche Diagnostic
Japan, Tokyo, Japan). All gene-specific mRNA measurements were
normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
mRNA levels. All PCR primer sequences are listed in Supporting Infor-
mation Table S1.
2.6 | Immunohistochemistry
Under deep anesthesia, TBI rats at 1.5 or 7 dpi were subjected to per-
fusion fixation using the fixative described in Section 2.3.2 and as
described elsewhere (Sugimoto et al., 2014). The brain region contain-
ing the injury was coronally cryosectioned at a thickness of 10 μm
and subjected to immunohistochemical staining using antibodies listed
in Supporting Information Table S2. Hoechst 33,342 (Sigma-Aldrich,
St. Louis, MO) was used for nuclear staining. To detect 8-hydroxy-20 -
deoxyguanosine (8-OHdG), brain sections were incubated in boiling
10 mM citrate buffer (pH 6.0) for 5 min followed by cooling in water
at room temperature, before incubation with an anti-8-OHdG anti-
body. Stained sections were observed with a Nikon A1 confocal laser
scanning microscope (Tokyo, Japan) or a scanning fluorescence micro-
scope (BZ-9000; Keyence, Osaka, Japan).
2.7 | Primary rat mixed glial and purified microglial
cell cultures
Rat primary mixed glial and purified microglial cell cultures were pre-
pared as described previously (Tanaka et al., 1998). Whole forebrains
from neonatal rats were mechanically dissociated into individual cells.
Mixed glial cultures were grown in Dulbecco’s modified Eagle’s medium
(DMEM) containing 10% fetal bovine serum in poly-L-lysine-coated
6-well culture plates for 11 days. Culture medium was then removed
and cells were incubated for 3 hr in serum-free E2 medium (DMEM
containing 10 mM HEPES, pH 7.3, Gibco, Grand Island, NY; 4.5 mg/ml
glucose, 5 μg/ml insulin, 5 nM sodium selenite, 5 μg/ml transferrin,
Gibco; and 0.2 mg/ml bovine serum albumin, Sigma-Aldrich) in the pres-
ence or absence of 1 μg/ml LPS (Sigma-Aldrich). Cells were then solubi-
lized for qPCR using an RNeasy Lipid Tissue Mini Kit (Qiagen). Purified
microglial cells were prepared from the mixed glial culture and were
used to determine cellular energy metabolism. BU (100 μg/ml; Higaki
et al., 2016) was added to cultures 24 hr before collecting total RNA
for qPCR or measurement of cellular metabolism.
2.8 | Flow cytometry
Under deep anesthesia, rats at 1.5 or 7 dpi were subjected to transcardial
perfusion with chilled PBS for 3 min to remove blood, followed by
4 ABE ET AL.

dissection of injured brains. Right brain hemispheres (~500 mg) of TBI 2.12 | Identification of cells generating superoxide
and normal rats were dissociated into single cells in a gentleMACS disso- anions using dihydroethidium
ciator using program 37C_ABDK_01 (Miltenyi Biotec, Tokyo, Japan) and
Dihydroethidium (DHE; Sigma-Aldrich) was used to monitor cells generat-
solutions provided with the adult brain dissociation kit (Miltenyi Biotec)
ing superoxide anions in TBI brain tissues (Choi et al., 2012). DHE
as previously described in Hattori et al. (2017). Nondissociated tissue
(10 mg/kg) was intraperitoneally administered to TBI model rats at 1.5 dpi
aggregates were removed using MACS SmartStrainers with 100-μm
and, 30 min later, the animals were euthanized and the brains dissected
pores (Miltenyi Biotec), and then debris and contaminating erythrocytes
out. Coronal slices were cut and fixed in 4% paraformaldehyde in PBS for
were removed from the cell suspensions using debris removal and red
30 min. After incubating in PBS containing 15% sucrose for 2 h, 10-μm
blood cell removal solutions provided with the adult brain dissociation kit.
thick cryosections were prepared. After immunohistochemical staining
The prepared cell suspensions were subjected to flow cytometry
using antibodies to Iba1 and CD45, the specimens were observed with a
analyses. The suspensions were diluted to 1 × 106 cells/100 μl with
Nikon A1 confocal laser scanning microscope (Tokyo, Japan).
PBS containing 2% fetal bovine serum and 2 mM ethylenediaminete-
traacetic acid. Cells were then incubated on ice with an anti-CD32
antibody for 20 min to block Fc receptors. Cells then were incubated
2.13 | Quantification of IGF-1 in brain tissues
with fluorescence-labeled antibodies (see Supporting Information IGF-1 protein in the right hemisphere of TBI rat brains administered
Table S3). To detect CD206+ cells, cells were incubated with an anti- vehicle or BU was quantified with an ELISA kit (R&D Systems). Before
CD206 antibody and then with an Alexa405-labeled anti-rabbit IgG the assay, brain tissues were homogenized in PBS and stored at
antibody (abcam) for 30 min on ice. To detect generated mitochon- −30
C overnight, followed by two freeze–thaw cycles to break cell
drial reactive oxygen species (ROS) (mitROS), cells isolated at 1.5 dpi membranes. Supernatants were then prepared and the ELISA assay
were suspended in E2 medium and incubated with MitoROS performed in triplicate.


520 (AAT Bioquest, Sunnyvale, CA) for 20 min at 37 C. The cells
labeled with antibodies for flow cytometry (see Supporting Informa- 2.14 | Determination of cell metabolic rates
tion Table S3) and/or MitoROS 520 were analyzed on a Gallios flow
Oxygen consumption rate (OCR) and extracellular acidification rate
cytometer (Beckman Coulter, Tokyo, Japan). Cell viability was ana-
(ECAR) were investigated to evaluate mitochondrial respiration and glycol-
lyzed using propidium iodide or Zombie NIR (BioLegend). Data were
ysis, respectively, in rat primary microglial cells and alveolar macrophages
analyzed using FlowJo Software (version.7.6.5, Treestar, Ashland, OR).
using XFp Extracellular Flux Analyzer (Agilent Seahorse Bioscience, Santa
Clara, CA) (Banh et al., 2016; Garaude et al., 2016). Rat alveolar macro-
2.9 | Cell sorting and qPCR phages, prepared as described elsewhere (Kikuchi et al., 2015), or micro-

mRNA levels in sorted cells were analyzed by qPCR. Before flow cyto- glial cells were seeded in E2 at 2.5 × 104 cells in eight-well plates provided

metry sorting, cells were resuspended with 5 volumes Cell Cover by the manufacturer. Two wells among the eight well were used as blank


(AL Anacyte Laboratories UG, Hamburg, Germany) at 4 C to stabilize wells. After overnight incubation in a CO2 incubator, the plates were

mRNA. Macrophages and microglial cells were collected on a BD FAC- placed in the XFp analyzer, which was operated according to the manufac-
turer’s instructions using oligomycin A (1 μM), carbonyl cyanide-p-
SAria flow cytometer (BD biosciences, Franklin Lakes, NJ). Total RNA
trifluoromethoxyphenylhydrazone (FCCP; 1 μM), rotenone (0.5 μM), and
was prepared from cells using an RNeasy micro kit (Qiagen) and then
antimycin A (0.5 μM), which were automatically and sequentially added to
reverse-transcribed to obtain cDNA. qPCR was performed as
cells to determine mitochondrial and glycolysis activities.
described earlier.

2.15 | Statistical analysis

2.10 | Quantification of 8-OHdG
Data expressed as means  SEM or SD were statistically analyzed using
To evaluate oxidative damage in injured brain tissues, 8-OHdG was
InStat3 software (GraphPad Software, La Jolla, CA). Data were sub-
quantified. Injured brain tissues were isolated at 1.5 dpi, and DNA
jected to two-tailed Student’s t test (unpaired) or ANOVA with Tukey’s
was extracted from the tissue (~150 mg) using a DNA Extractor TIS
post hoc test. Significance was set at p < .05 unless otherwise stated.
kit (Wako, Osaka, Japan) and hydrolyzed using a DNA hydrolysis kit
(Wako). The 8-OHdG levels were quantified using a highly sensitive
ELISA kit (Nikken SEIL, Shizuoka, Japan). 3 | RE SU LT S

2.11 | Quantification of interleukin-6 in 3.1 | Effects of BU on cell metabolism

cerebrospinal fluid
Cerebrospinal fluid (CSF) was taken at 1.5 dpi from the space over the
midbrain via the small holes in the skull. Interleukin-6 (IL-6) concentra-
tion was determined using an ELISA kit (R&D Systems; Minneapo-
lis, MN).
BU has marked antiinflammatory effects on microglia (Higaki et al.,
2016) and macrophages (Kikuchi et al., 2015), and the effects that are
at least partly attributable to suppressive effects on ATP synthesis
(Kawasaki et al., 2017). In this study, the effects of BU on cell metabo-
lism were investigated in primary microglial cells (Figure 1) and
ABE ET AL. 5

alveolar macrophages (Supporting Information Figure S1) using the
Seahorse Mito-Stress Test, which assesses Oxygen consumption rate
(OCR) and extracellular acidification rate (ECAR) (Figure 1a) (Banh
et al., 2016; Garaude et al., 2016). By measuring OCR, basal
(Figure 1b) and maximal respiration (Figure 1 da), ATP production
(Figure 1c), spare respiration capacity (Figure 1db), and non-
mitochondrial oxygen consumption (Figure 1e) were determined.
ECAR reflects glycolysis activity. BU decreased all of the indices for
glycolysis and mitochondrial metabolism that lead to reduced ATP
production. However, BU did not affect coupling efficiency
(Figure 1f ), which is calculated by dividing ATP production by basal
respiration, indicating that BU did not induce mitochondrial dysfunc-
tion but suppressed total cellular metabolism.
3.2 | Amelioration of TBI by BU
A penetrating stab injury was made in the right hemisphere of
8-week-old male Wistar rats as a TBI model. BU was subcutaneously
administered 1 h after TBI. BU dissolved in water was then given to
TBI rats for 7 days. Rats were then given normal water for 15 weeks.
Five and ten weeks after injury, the injured brains were monitored by
MRI (Figure 2a). At both time points, TBI-induced cavity or lost brain
tissue volumes were reduced in the BU administered rats. Rats were
euthanized 15 weeks after TBI and the cavity size in fixed brain slices
was quantified [Figure 2b; also see Supporting information Figure S2
displaying all rat brain slices examined (n = 12 for each group)]. No sig-
nificant differences in motor abilities were detected by open-field and
rota-rod tests (data not shown); however, BU administration amelio-
rated spatial cognitive impairments and learning/memory
dysfunctions as revealed by Morris water maze and passive avoidance
tests that were performed 15 weeks after TBI (Figure 2c–e). In the
water maze test, BU reduced the duration to reach the platform and
increased the number of rats that accomplished the task (Figure 2d)
but did not affect swimming velocity (Figure 2dd). BU administration
improved memory retention as revealed by the passive avoidance test
(Figure 2e). Thus, BU brought about long-term amelioration in terms
of TBI-induced brain tissue loss and neurological dysfunctions.
3.3 | CCL2 expression and proinflammatory
reactions
BU suppresses CCL2 expression in LPS-treated peritoneal macro-
phages (Kikuchi et al., 2015). CCL2 is also expressed by astrocytes
and microglial cells (Berman, Guida, Warren, Amat, & Brosnan, 1996;
Liu et al., 2013; Tei et al., 2013). We, therefore, investigated whether
BU inhibited CCL2 expression in a mixed glial culture containing astro-
cytes, microglial cells and NG2 glia (oligodendrocyte progenitor cells).
Unless stimulated with LPS, the mixed glial culture expressed CCL2
mRNA at a low level. Incubation with LPS markedly increased CCL2
mRNA levels, and BU significantly suppressed this upregulation
(Figure 3a).
The injured brain regions were dissected at 1.5 dpi as shown in
Figure 3b. The dissected brain tissues were used for qPCR analyses.
BU suppressed mRNA levels of CCL2 (Figure 3ca) and its receptor
CCR2 (Figure 3c[b]), the latter being mainly expressed by mono-
cytes and macrophages (Gyoneva & Ransohoff, 2015; Semple,
Kossmann, & Morganti-Kossmann, 2010). mRNAs for CD11b
(Figure 3da), a marker for myeloid leukocytes and microglia, Iba1

FIGURE 1 Effects of BU on energy metabolism in rat primary microglia as assessed by OCR and ECAR. (a) OCR (upper) and ECAR (lower) of cells treated
with vehicle or BU. OCR and ECAR values in the absence for measuring a basal respiration rate (b) or in the sequential presence of oligomycin for an ATP
production rate (c), FCCP for a maximal respiration rate (da) and spare respiratory capacity (db), and rotenone+antimycin a for nonmitochondrial OCR (e).
Coupling efficiency (f ) was calculated by dividing ATP production by basal respiration. Data (n = 3) are shown as means  SD and were statistically
analyzed with Student’s t test. *p < .05, ***p < .001 versus vehicle [Color figure can be viewed at wileyonlinelibrary.com]
6 ABE ET AL.

FIGURE 2 Amelioration of TBI in the rat forebrain by BU. (a) BU reduced the lesion cavity size 5 and 10 weeks after TBI as revealed by MRI.
Representative MR images of vehicle (aa) and BU-administered (ab) rats were taken 10 weeks after TBI. The measured cavity size or lost volume
(n = 12; mm3) based on MR images taken 5 (ac) or 10 (ad) weeks after TBI. (b) Brains were dissected, fixed and coronal slices were prepared
15 weeks after TBI. The volume of left (contralateral) and right (ipsilateral) hemispheres of brain slices were measured independently, and the
(left–right volume)/left volume values are shown. (c,d) BU administration ameliorated impaired learning ability as revealed by the Morris water
maze test. (c) Representative traces recorded on days 1 and 2 of rats administered with vehicle (water; ca) or BU (cb). (d) Duration to reach the
platform (da), distance moved to the platform (db), and the percentage of rats that reached the platform (dc) are shown. Swimming velocity was
not different between the two groups (dd). (e) Transfer latency (sec) in the passive avoidance test. Data are shown as means  SD and were
statistically analyzed with Student’s t test (n = 12). *p < .05, **p < .01, ***p < .001 versus vehicle [Color figure can be viewed at
wileyonlinelibrary.com]

(AIF1) (Figure 3db), a specific marker for microglia and monocytes/ IL-6 (Figure 3ed). BU also decreased the concentration of IL-6 in
macrophages (Matsumoto et al., 2007), and neutrophil elastase CSF (Figure 3f ). BU decreased mRNA levels of p40phox
(ELANE) (Figure 3dc) were all suppressed by BU administration. BU (Figure 3ga) and p67phox (Figure 3gb), subunits of NADPH oxidase
reduced mRNA levels of the proinflammatory cytokines, (NOX), which generates ROS (Ansari, Roberts, & Scheff, 2014).
interleukin-1β (IL-1β) (Figure 3ea), tumor necrosis factor α (TNFα; However, BU did not reduce mRNA levels for TGFβ1
Figure 3eb), inducible nitric oxide synthase (iNOS; Figure 3ec) and (Figure 3ha) or IGF-1 (Figure 3hb).
ABE ET AL. 7

FIGURE 3 Suppressive effects of BU on proinflammatory responses in vitro and in vivo. (a) BU suppressed LPS-induced CCL2 mRNA expression
in a mixed glial culture. Total RNA was collected 3 h after LPS addition. Data are shown as means  SEM (n = 4). **p < .01, *** p < .001 versus
control/LPS(−). ## p < .01, versus control/LPS(+). ANOVA and Tukey’s post hoc test. (b) the injured brain tissues denoted with dotted lines were
isolated at 1.5 dpi from brain slices for qPCR. (c) BU administration suppressed levels of CCL2 and CCR2 mRNA. (d) BU suppressed mRNA levels
of CD11b, Iba1, and neutrophil elastase. (e) mRNA levels of IL-1β, TNFα, iNOS and IL-6 were also suppressed by BU. (f) BU administration
reduced IL-6 protein levels in CSF at 1.5 dpi. (g) BU reduced mRNA levels of NOX subunits, p40 and p67phox. (h) BU did not affect mRNA levels
of TGFβ1 or IGF-1. * p < .05, **p < .01 versus vehicle. Data analyzed with Student’s t test are shown as means  SEM (n = 7 for qPCR and n =
4 for ELISA) [Color figure can be viewed at wileyonlinelibrary.com]

3.4 | Flow cytometry analyses of leukocytes and
microglia at 1.5 dpi
Flow cytometry analyses using mixtures of antibodies to CD45 (clone
OX1) and granulocytes (Figure 4a, see Supporting information
Figure S3 for gating strategy) or CD11b/c (clone OX42; Figure 4b)
revealed only a few macrophages and granulocytes in the normal brain
(Figure 4aa, ba, and c). CD45hi macrophages and CD45+/granulocyte+
cells (granulocytes or presumptive neutrophils) were abundantly pre-
sent in vehicle-treated TBI brains, whereas the number of CD45lo
microglia was decreased (Figure 4ab, bb, and c). BU did not affect the
numbers of accumulated granulocytes or microglia (Figure 4c). How-
ever, the number of macrophages in BU-treated brains was less than
that in vehicle-treated brains. The discrimination of microglia from mac-
rophages by flow cytometry based on the level of CD45 expression
was validated by determining the mRNA level for microglia-specific cell
surface protein, TMEM119 (Bennett et al., 2016; Satoh et al., 2016)
(Figure 4d). TMEM119 expression in both resting and Act-MG fractions
was higher than that in the macrophage fraction.
Resting and Act-MG were separated by forward (FS) and side scat-
ter (SS) plots (Figure 4e). Microglia in normal brains were a very homog-
enous population (Figure 4ea). We presupposed that 96% of microglial
cells (96.0  1.40%; mean  SD; n = 5) are classified as resting microglia
(Rest-MG) in normal rat brains, and the gating (dotted line in
Figure 4ea) was set based on this estimation. Microglia excluded from
the gating were defined as Act-MG, and are surrounded with a red line.
The FS and SS values of the Act-MG were comparable with those of
macrophages, which are denoted with blue dots surrounded with blue
lines (Figure 4eb, ec). Although the total number of microglia decreased,
the percentages of Act-MG were much increased in TBI brains regard-
less of treatment (Figure 4f ). Act-MG characteristically had higher FS
and SS values than Rest-MG (Supporting Information Figure S4). TBI
8 ABE ET AL.

increased FS and SS values of granulocytes, macrophages and Act-MG.
Act-MG characteristically increased their CD11b/c expression com-
pared with resting microglia, but their CD45 expression was not chan-
ged in response to TBI.
Immunohistochemical analyses showed the presence of a few
Act-MG with weak CD45-immunofluorescence and ramified morphol-
− +
ogy (Figure 4g,h; pink arrows). Also present were Iba1 /CD45 cells
that were granulocytes because of their segmented nuclei (yellow
arrows and the inset in Hc) and Iba1+/CD45+ macrophages (blue
arrows), which displayed round morphology, a feature that was dis-
tinctive from the ramified microglial morphology (Sugimoto et al.,
2014). BU administration did not affect CD45 expression and mor-
phology of microglia in the normal brain (Supporting Information
Figure S5).

FIGURE 4
ABE ET AL. 9

3.5 | ROS-generating cells and oxidative stress
at 1.5 dpi
ROS oxidizes DHE to produce red fluorescence from ethidium
(Choi et al., 2012). Cells emitting ethidium fluorescence were
Iba1−/CD45+ granulocytes and Iba1+/CD45hi macrophages, but not
+ lo/−
Iba1 /CD45 microglia with ramified processes (Figure 5a). Note
the absence of ramified microglia in the core lesion where abundant
granulocytes and macrophages are located as described elsewhere
(Matsumoto et al., 2007). BU did not affect CD45 expression and
also it did not cause ethidium fluorescence in the normal brains
(Supporting Information Figure S6A). BU reduced the ethidium fluo-
rescence in and around the lesion core (Supporting Information
Figure S6B).
Approximately 90% of ROS are generated by mitochondria (Gao,
Laude, & Cai, 2008); therefore, mitROS generation by the individual
cell types was evaluated using flow cytometry with a superoxide-
sensitive fluorescent molecule, MitoROS 520, which quickly enters
mitochondria and generates green fluorescence when oxidized by
superoxide (Serrano-Nascimento et al., 2014) (Figure 5b). When com-
pared with granulocytes and macrophages, both activated and resting
microglia generated much lower levels of mitROS. TBI treatment
increased mitROS-generation in granulocytes, macrophages and rest-
ing microglia but not in Act-MG. BU significantly reduced mitROS
generation by macrophages and microglia.
To evaluate oxidative damage, 8-OHdG in DNA, a marker for
oxidative stress (Valavanidis, Vlachogianni, & Fiotakis, 2009), was
immunohistochemically identified on cryosections of TBI brains at
1.5 dpi (Figure 5c). The majority of strongly 8-OHdG-positive cells
were macrophages (green arrowheads). A small number of NeuN+
neurons were weakly 8-OHdG-positive (pink arrowheads). In
vehicle- and BU-treated normal rat brains, no significant 8-OHdG-
immunoreactivity was found (Supporting Information Figure S7a).
Figure 5d shows that BU reduced 8-OHdG-immunoreactivity in and
around the lesion cores. 8-OHdG levels were quantified by ELISA.
TBI increased 8-OHdG levels, and BU administration abolished this
increase to the control level (Figure 5e, also see Supporting Informa-
tion Figure S7b for enlarged micrographs detecting 8-OHdG, Iba1,
and NeuN).
3.6 | Characterization of sorted macrophages and
activated and resting microglia by qPCR
Macrophages and microglia accumulated in injured hemispheres were
collected by cell sorting at 1.5 dpi, and cDNA prepared from the iso-
lated cells was subjected to qPCR (Figure 6). NADPH-oxidase
2 (NOX2; gp91phox) mRNA was found in macrophages at much higher
levels than in microglia (Figure 6a). Other NOX components, p40phox
and p67phox, were expressed by microglia and macrophages at similar
levels, and BU did not affect their expression (Supporting Information
Figure S8). Glutathione peroxidase (GPx) mRNA was also expressed at
higher levels by macrophages than by microglia (Figure 6b). BU did
not affect the expression of these mRNAs. Higher mRNA levels of
iNOS (Figure 6c) and IL-1β (Figure 6d) were expressed by macro-
phages compared with microglia, and Act-MG expressed these
mRNAs at higher levels than Rest-MG. BU reduced iNOS and IL-1β
mRNA levels in macrophages. Conversely, microglia expressed TNFα,
IL-6, and TGFβ1 mRNAs at higher levels than macrophages. BU sup-
pressed TNFα and IL-6, but not TGFβ1 mRNA expression. Microglia
and macrophages expressed IGF-1 mRNA at similar levels, and BU did
not affect its expression. Macrophages expressed mRNA encoding
CD68, a marker for mononuclear phagocytes (Song, Lee, & Schindler,
2011), at higher levels than microglia (Figure 6i). The high expression
of CD68 in macrophages was confirmed by immunohistochemical
staining (Supporting Information Figure S9). Microglia in the close
vicinity of the lesion core displayed faint CD68-immunoreactvity.
3.7 | Effects of BU administration on TBI brains
at 7 dpi
CD45+ cells in TBI brains at 7 dpi were analyzed by flow cytometry
(Figure 7a–d; Supporting Information Figure S10). At this time point,
the number of macrophages was less than that of total microglia in
the injured hemisphere regardless of treatment. The number of Act-
MG was much greater than that of Rest-MG. BU did not affect the
percentage of Act-MG or Rest-MG. BU decreased the number of mac-
rophages expressing CD86, a presumable M1 marker, and it weakly
increased the number of CD206+ macrophages, which may be M2
cells (Figure 7c). NG2 chondroitin sulfate proteoglycan (NG2), a

FIGURE 4 Granulocytes, macrophages and microglia in normal, vehicle- and BU-administered TBI brains. (a) Flow cytometry analyses using
antibodies against granulocytes and CD45 distinguished granulocytes, macrophages (MΦ) and microglia (MG). Only a few granulocytes and
macrophages were present in normal brain (aa). The number of total microglia was reduced and the numbers of macrophages and granulocytes
were increased in TBI brains (ab, vehicle and ac, BU-administered). Representative results are shown. (b) Analyses using antibodies against
CD11b/c and CD45. Reduced numbers of microglia and increased numbers of macrophages are apparent in TBI brains (bb, vehicle and Bc, BU-
administered). (c) The frequency of each cell type with respect to total live cells is shown. Granulocyte and macrophage numbers were increased
and microglia numbers were reduced in TBI brains. BU administration significantly reduced the accumulation of macrophages (### indicates
significant reduction; p < .001). (d) Significantly higher levels of TMEM119 mRNA were found in sorted Act-MG and Rest-MG compared with
macrophages. (e) Act-MG and Rest-MG were distinguished by FS and SS plots. (ea) a 96% of microglial cells with small FS and SS values in normal
brains were defined as Rest-MG, denoted with the dotted line. A few Act-MG and macrophages were present in normal brain. The numbers of
Act-MG and macrophages in a TBI brain denoted with red or blue dots and lines, respectively, were increased after TBI (eb). BU decreased the
number of macrophages but not Act-MG (ec). (f ) TBI increased the percentage of Act-MG and reduced the percentage of Rest-MG, but BU did
not affect these percentages. Flow cytometry data analyzed with ANOVA and Tukey's post-hoc test are shown as means  SEM (n = 5). **p <
.01, *** p < .001 versus control of each group. $$$ p < .001, versus act-MG in each group. (g,h) Immunohistochemical staining of TBI brains at 1.5
dpi using antibodies against Iba1 and CD45. Nuclear staining was performed using Hoechst 33,342. (h) 3D-constructed images are shown in (g).
Blue arrows denote Iba1+/CD45+ macrophages, pink arrows Iba1+/CD45weak Act-MG, and yellow arrows Iba1−/CD45+ granulocytes. The inset in
(hc) shows a segmented nucleus of a granulocyte [Color figure can be viewed at wileyonlinelibrary.com]
10 ABE ET AL.

FIGURE 5 ROS generation by leukocytes and microglia in TBI brains and oxidative stress. (a) Immunohistochemical detection of ROS-generating
cells in TBI lesions at 1.5 dpi using DHE staining. Cells with strong DHE fluorescence were macrophages (blue arrows) and neutrophils (yellow)
with segmented nuclei. Microglial cells (pink) did not emit DHE fluorescence. (b) Granulocytes (ba), macrophages (bb), act-MG (bc) and rest-MG
(bd) were individually examined for mitROS generation by flow cytometry. Representative results (n = 5) are shown. Mean fluorescence intensity
for mitROS generation by granulocytes and macrophages in TBI brain was much stronger than that generated by microglia (be). BU significantly
suppressed macrophage and microglia-derived mitROS. Data were analyzed with ANOVA and Tukey’s post hoc test and are shown as means 
SEM (n = 5). ***p < .001 versus control of each group. ### p < .001, cells in vehicle-treated (Vehi) versus BU-treated TBI brains.
(c) Immunohistochemical staining of 8-OHdG in a vehicle-treated TBI brain. Iba1+ macrophages were strongly immunostained with anti8-OHdG
antibody (green arrowheads). NeuN+ degenerating neurons were weakly 8-OHdG-immunoreactive (pink arrowheads). (d) Immunohistochemical
staining of 8-OHdG in the right hemisphere of TBI brains. (dc) morphometric results show that BU treatment reduced 8-OHdG immunoreactivity.
Data were analyzed with Student's t-test and are shown as means  SEM *p < .05 versus vehicle (n = 3). (e) Quantification of 8-OHdG levels in
sham, vehicle- and BU-administered TBI brains by ELISA showed a TBI-induced increase of 8-OHdG levels and prevention of oxidative stress
effects by BU. data were analyzed with ANOVA and Tukey’s post hoc test and are shown as means  SEM (n = 4). *,# p < .05 versus sham and BU,
respectively [Color figure can be viewed at wileyonlinelibrary.com]
ABE ET AL. 11

FIGURE 6 Expression of mRNA by macrophages and microglia isolated from TBI brains by flow cytometry was analyzed by qPCR.
(a) Macrophages (MΦ) expressed NOX2 (a) and GPx (b) mRNA at higher levels than microglia. BU did not affect the expression. Macrophages
expressed iNOS (c) and IL-1β (d) mRNA at higher levels than microglia and BU suppressed this expression. Act-MG expressed iNOS mRNA at a
higher level than Rest-MG. Microglia expressed TNFα (e), IL-6 (f) and TGFβ1 (g) mRNAs at higher levels than macrophages. BU suppressed
expression of TNFα and IL-6 mRNAs but not that of TGFβ1. IGF-1 mRNA (h) was expressed at similar levels by all cells and BU did not affect its
expression. Macrophages expressed CD68 mRNA (i) at higher levels than microglia. BU did not affect CD68 expression. Data were analyzed with
ANOVA and Tukey's post-hoc test and are shown as means  SEM (n = 4). * $ p < .05, ** p < .01, ### p < .001. * versus vehicle-treated the same
cell type. $ versus vehicle-treated Act-MG. # versus vehicle-treated macrophages [Color figure can be viewed at wileyonlinelibrary.com]

marker for oligodendrocyte progenitor cells or NG2 glia, is expressed
by macrophages and microglia in injured brains that may be beneficial
cells for injured brains (Matsumoto et al., 2008; Smirkin et al., 2010).
BU treatment did not affect the number of NG2+ macrophages.
Figure 7e shows mRNA expression in TBI brain tissues at 7 dpi. BU
administration decreased levels of mRNAs encoding CCL2, CCR2, and
iNOS. Note that the level of CCL2 mRNA was much decreased com-
pared with that at 1.5 dpi. In contrast, BU increased levels of Iba1,
GFAP, and IGF-1 mRNA. Proliferating cell nuclear antigen (PCNA) and
TGFβ1 mRNA levels were not affected by BU treatment. BU treat-
ment did not change the IGF-1 protein content in injured right hemi-
spheres (Figure 7k). Immunohistochemical observations showed that
BU did not cause marked differences in Iba1, CD11b, NG2, and Ki67
expression in the injured hemisphere at 7 dpi (Supporting Information
Figure S11). However, dense accumulation of NG2+ macrophages
were often observed in the close vicinity of the lesions in BU-treated
brains (Supporting Information Figure S12).
4 | DI SCU SSION
4.1 | Strategy to determine the aggravating cell type
in the TBI pathology
It has been a focus of debate for decades whether Act-MG are benefi-
cial or detrimental to injured brain tissues (Corps et al., 2015; Dud-
varski Stankovic, Teodorczyk, Ploen, Zipp, & Schmidt, 2016; Karve,
Taylor, & Crack, 2016; Patel et al., 2013). Because Act-MG display
similar phenotypes to those of infiltrated macrophages, they are diffi-
cult to distinguish from each other. In this study, we took advantage
of flow cytometry to distinguish microglia from macrophages based
on different levels of CD45 expression. We confirmed much higher
expression of TMEM119 (Bennett et al., 2016; Satoh et al., 2016) in
isolated microglia compared with macrophages. Activated and resting
microglia were discriminated on the basis of FS and SS values. CD11b
expression was also higher in Act-MG compared with Rest-MG.
12 ABE ET AL.

FIGURE 7 CD45+ cells from TBI brains were analyzed by flow cytometry (a–d) and effects of BU on mRNA and protein expression at 7 dpi were
analyzed (e–k). (a) Analyses with anti-CD11b and anti-CD45 antibodies. BU administration for 7 days did not affect the numbers of macrophages
or total microglia accumulated in injured hemispheres at 7 dpi. Representative flow cytometry plot data are shown in Supporting Information
Figure S10. (b) BU administration did not affect the ratio of Act-/Rest-microglia. Data (n = 4) are shown as means  SD. Statistical analyses using
ANOVA and Tukey’s post hoc test (a,b) showed significant differences (**, p < .01) compared with macrophage or Act-MG numbers in vehicle-
treated TBI brains, respectively. (c) CD86+ and CD206+ macrophages in the injured sites. BU administration weakly decreased the number of
CD86+ macrophages and increased the number of CD206+ macrophages. (d) BU administration did not affect the number of NG2+ macrophages.
BU-induced changes in mRNA levels and IGF-1 content in TBI brains at 7 dpi. (e) Changes in mRNA levels in injured tissues at 7 dpi as revealed
by qPCR. BU administration for 7 days reduced levels of mRNAs encoding CCL2, CCR2, and iNOS, and increased those encoding Iba1, NG2,
GFAP, and IGF-1. BU administration did not affect TGFβ1 or PCNA mRNA levels. (k) BU did not affect IGF-1 protein contents in injured brain
tissues. Data analyzed with Student’s t test (c–k) are shown as means  SEM (n = 4 for [c,d]; n = 5 for [e,k]. *, p < .05, ** p < .01, *** p < .001
versus vehicle [Color figure can be viewed at wileyonlinelibrary.com]

Macrophages, and Act-/Rest-MG were thus individually characterized
to evaluate their contribution to TBI pathogenesis. Total RNA was
purified from sorted cells and mRNA level was standardized with
GAPDH mRNA to evaluate the changes in the nature of cells but not
accumulated cell numbers.
Because BU had marked ameliorative actions on TBI brains, it
may be possible to infer the pathogenic mechanisms in TBI from the
BU-induced changes. For example, BU reduced the generation of
8-OHdG in TBI brains; therefore, oxidative injury may play a deleteri-
ous role in TBI, as has been proposed (Choi et al., 2012; Corps et al.,
2015; Hall, Wang, & Miller, 2012; Zhang et al., 2012). Proinflamma-
tory mediators, whose expression and/or release were suppressed by
BU are, therefore, likely to play aggravating roles in TBI. Based on
these proposals, we compared the pathophysiological contributions of
microglia and macrophages to TBI.
4.2 | Detrimental factors derived from macrophages
and microglia
BU-mediated reduction in 8-OHdG generation in TBI lesions may be
attributable to reduced macrophage accumulation as well as inhibi-
tion of their mitROS generation. Rest- and Act-MG generated much
less amount of mitROS than macrophages. mitROS, which are 90%
of total cellular ROS (Gao et al., 2008), may be the main cause of
ABE ET AL.
oxidative stress-induced brain injury (Chouchani et al., 2014). NOX2
is another source for ROS (Zhang et al., 2012), and it was more
highly expressed by macrophages than microglia. Therefore, macro-
phages may be the major cell type responsible for the oxidative
stress in TBI. However, BU did not affect NOX2 or p40phox and
p67phox mRNA levels in macrophages, indicating that NOX2 may be
less critical for the oxidative stress. Macrophages expressed iNOS
mRNA at higher levels than microglia. iNOS-derived NO may play a
neuroprotective role (Sinz et al., 1999), but it can still be deleterious
because it forms highly toxic peroxynitrite (ONOO−) by reacting with
superoxide (Hall et al., 2012).
IL-1β expression is strongly increased after TBI and plays a central
role in enhancement of proinflammatory responses by increasing
blood brain barrier permeability, immune cell recruitment, and resident
glial cell activation (McKee & Lukens, 2016). Macrophages expressed
IL-1β mRNA at a higher level than microglia. BU inhibited IL-1β mRNA
expression in TBI lesions and macrophages. IL-1β is produced as an
inactive form that requires cleavage before its secretion by caspase-1,
whose activation is regulated by inflammasomes (Latz, 2010). Endoge-
nous Toll-like receptor (TLR) ligands, referred to as damage-associated
molecular patterns (DAMPs) (Shichita et al., 2012; Webster et al.,
2017) induce mitROS generation (West et al., 2011), which increases
IL-1β release by activation of the major inflammasome, the
nucleotide-binding oligomerization domain-like receptor family, pyrin
domain-containing 3 inflammasome (Garaude et al., 2016; Heid et al.,
2013). BU reduced mitROS generation by macrophages. Therefore,
BU presumably reduced IL-1β actions by two ways; inhibition of
DAMPs-mediated mRNA expression and reduced mitROS-induced
secretion. These observations suggest that macrophages but not Act-
MG play a central aggravating role in TBI.
4.3 | Favorable factors derived from macrophages
and microglia
Act- and Rest-MG expressed TNFα, IL-6, and TGFβ1 mRNA at higher
levels than macrophages. Microglia and macrophages expressed IGF-1
mRNA at similar levels. Although both TNFα and IL-6 are classified as
proinflammatory cytokines, they have also been reported to exert
neuroprotective actions (Lenzlinger, Morganti-Kossmann, Laurer, &
McIntosh, 2001; Toku et al., 1998; Woodcock & Morganti-Kossmann,
2013). However, BU suppressed the expression of TNFα and IL-6,
indicating that they have some detrimental roles. BU did not affect
the expression of TGFβ1, which has an ameliorative action in ischemic
and hemorrhagic brains (Taylor et al., 2017; Wattananit et al., 2016).
TGFβ1 inhibits TLR ligands-induced phosphorylation of IκB kinase in a
sustained manner, leading to the sustained inhibition of NFκB translo-
cation into the nucleus (Islam et al., 2018). This action may be corre-
lated with the induction of anti-inflammatory macrophage and
microglia phenotypes in the subacute phase of severely damaged
brain tissues (Islam et al., 2018; Taylor et al., 2017). The secretion of
favorable cytokines together with much less ROS and IL-1β produc-
tion should make microglia more neuroprotective than macrophages,
even when they are activated in TBI lesions.
Although both microglia and macrophages have been implicated
in removal of cell and tissue debris in injured brains through their
13
phagocytic activities (Matsumoto et al., 2008; Ritzel et al., 2015), mac-
rophages may have stronger phagocytic activity than microglia
because of their higher expression of CD68. Phagocytic elimination of
degenerated materials in injured brains may enhance regeneration
(Neumann, Kotter, & Franklin, 2009). On the other hand, it has also
been insisted that the phagocytes accelerate neuronal death by elimi-
nating viable neurons (Neher et al., 2013; Vilalta & Brown, 2017). BU
did not affect the expression of CD68 by macrophages and microglia,
suggesting that phagocytosis may have both beneficial and detrimen-
tal effects on injured brains.
4.4 | Neutrophils in the acute and macrophages
in the subacute phases
BU did not significantly affect accumulation of granulocytes, almost all
of which were presumably neutrophils. Furthermore, BU did not
inhibit neutrophil mitROS generation. Neutrophils massively accumu-
late in the core lesion within hours after the onset of TBI (Gyoneva &
Ransohoff, 2015; McKee & Lukens, 2016), causing blood brain barrier
breakdown and brain edema. However, blockade of neutrophil entry
into TBI lesions or genetic deletion/pharmacological inhibition of neu-
trophil elastase does not lead to marked amelioration of TBI (Semple,
Bye, Ziebell, & Morganti-Kossmann, 2010; Semple, Trivedi, Gimlin, &
Noble-Haeusslein, 2015; Weaver, Branch, Hernandez, Miller, & Quat-
trocchi, 2000). The weak effects of BU on neutrophils also suggest
that neutrophils may not be critical cells for secondary neuroinflam-
matory aggravation of TBI.
The high-level of mitROS may cause 8-OHdG generation in mac-
rophages and their subsequent oxidative stress-induced degeneration,
likely leading to the decreased number of macrophages at 7 dpi. Fur-
thermore, iNOS mRNA levels were also higher in macrophages, and
the generated NO may produce peroxynitrite (Hall et al., 2012), which
would affect their viability. Although macrophages expressed GPx
mRNA at very high levels, the scavenging effect of GPx may not be
sufficient to protect macrophages. Different from those present at 1.5
dpi, macrophages at 7 dpi may be ameliorative, given their polarization
toward the M2 phenotype and previous reports describing ameliora-
tive macrophages in injured brains that express IGF-1 and TGFβ1
(Nishihara et al., 2011; Smirkin et al., 2010; Sugimoto et al., 2014;
Wattananit et al., 2016).
The majority of studies that inhibit CCR2-CCL2 interactions in
TBI models to prevent the infiltration of macrophages have shown the
detrimental effects of monocyte-derived macrophages (Gyoneva
et al., 2015; Hsieh et al., 2014; Morganti et al., 2015; Semple, Koss-
mann, & Morganti-Kossmann, 2010; Song et al., 2017). However,
accumulated macrophages may be ameliorative at the subacute phase
as shown here and also elsewhere (Chu et al., 2015; Miro-Mur et al.,
2016; Smirkin et al., 2010; Wattananit et al., 2016). Regarding this,
there should be two possibilities; phenotypic change into ameliorative
cells of monocyte-derived macrophages belonging to a single lineage
and the presence of two different lineages. According to the litera-
tures, the phenotypic change may be more probable (Miro-Mur et al.,
2016; Wattananit et al., 2016).
14
4.5 | The BU actions on cell metabolism
BU suppressed mitochondrial activities (basal, maximal and spare res-
piration, and ATP production) but not coupling efficacy, indicating that
BU suppressed mitochondrial activities without causing dysfunction.
Furthermore, BU also reduced ECAR. Taken together, BU appeared to
reduce overall cellular energy consumption, while suppressing mitROS
generation. Elevation of brain extracellular lactate concentration in
TBI patients is correlated with poor outcomes (Carpenter, Jalloh, &
Hutchinson, 2015). The BU-induced decrease in ECAR implies a
decrease in lactate production or a reduced glycolysis rate. This BU
effect is likely correlated with improved outcome of the TBI model
rats. Non-mitochondrial OCR is an indicator of the activities of
hypoxia-inducible factor (HIF) prolyl hydroxylases, which consume
α-ketoglutarate and O2 as co-substrates causing HIF ubiquitination
(Banh et al., 2016). Because BU reduced the nonmitochondrial OCR, it
might increase the levels of HIFs, which is possibly correlated with the
BU-induced metabolic changes. Although the ameliorative effects of
BU on TBI may mainly be attributable to its antiinflammatory effects,
BU might also have direct neuroprotective effects in TBI brains by
decreasing neuronal cell metabolism, because hypothermia and
hypothyroidism-induced hypometabolic states exert neuroprotective
effects (Rastogi et al., 2006; Robertson, Scafidi, McKenna, & Fiskum,
2009). Therefore, it is necessary to determine whether BU has a sig-
nificant suppressive effect on neuronal cell metabolism.
5 | CO NC LUSIO N
By employing a combination of flow cytometry analyses, qPCR of iso-
lated cells, and immunohistochemistry, macrophages but not Act-MG
were found to be a major source of ROS and IL-1β. However, such
detrimental macrophages may undergo degeneration due to their own
ROS generation during the acute phase. Later, they might turn into
ameliorating cells for brain tissues by producing neuroprotective fac-
tors. Granulocytes may play less significant roles in aggravating TBI
pathology. Resident microglia may be neuroprotective cells because
they produce less IL-1β, NOX2, and mitROS while expressing more
TGFβ1 than macrophages and IGF-1 at a comparable level. These
results support the notion that prevention of CCL2-CCR2 interaction
leads to amelioration of TBI by reducing macrophage accumulation in
TBI brains in the acute phase. Reduction of energy metabolism in
macrophages and microglia likely leads to amelioration of TBI. BU is a
strong candidate to achieve this. The BU dosage for in vivo study was
approximately 50 mg/kg/day, which is comparable to the maximum
clinical dose. BU administration in the acute phase would ameliorate
the outcome of TBI by suppressing CCL2 expression and cellular
energy metabolism, resulting in reduced accumulation of macrophages
and reduced oxidative and inflammatory injuries.
ACKNOWLEDGMENTS
We are grateful to the staff of the Animal Center, Ehime University,
for their careful and gentle handling of the animals. This study was
partly supported by grants from Ehime University (#0503003000) and
ABE ET AL.
the Ministry of Education, Culture, Sports, Science, and Technology,
Japan (16 K20014 to SM; 17 K10836 to YK). We thank Jeremy Allen,
PhD, from Edanz Group (www.edanzediting.com/ac) for editing a
draft of this manuscript.
CONFLIC T OF INT E RE ST
Nothing to declare.
RE FE RE NC ES
Ansari, M. A., Roberts, K. N., & Scheff, S. W. (2014). A time course of
NADPH-oxidase up-regulation and endothelial nitric oxide synthase
activation in the hippocampus following neurotrauma. Free Radical
Biology & Medicine, 77, 21–29.
Bachstetter, A. D., Rowe, R. K., Kaneko, M., Goulding, D., Lifshitz, J., & Van
Eldik, L. J. (2013). The p38alpha MAPK regulates microglial responsive-
ness to diffuse traumatic brain injury. The Journal of Neuroscience, 33,
6143–6153.
Banh, R. S., Iorio, C., Marcotte, R., Xu, Y., Cojocari, D., Rahman, A. A., …
Neel, B. G. (2016). PTP1B controls non-mitochondrial oxygen con-
sumption by regulating RNF213 to promote tumour survival during
hypoxia. Nature Cell Biology, 18, 803–813.
Bennett, M. L., Bennett, F. C., Liddelow, S. A., Ajami, B., Zamanian, J. L.,
Fernhoff, N. B., … Barres, B. A. (2016). New tools for studying microglia
in the mouse and human CNS. Proceedings of the National Academy of
Sciences of the United States of America, 113, E1738–E1746.
Berman, J. W., Guida, M. P., Warren, J., Amat, J., & Brosnan, C. F. (1996).
Localization of monocyte chemoattractant peptide-1 expression in the
central nervous system in experimental autoimmune encephalomyelitis
and trauma in the rat. Journal of Immunology, 156, 3017–3023.
Carpenter, K. L., Jalloh, I., & Hutchinson, P. J. (2015). Glycolysis and the
significance of lactate in traumatic brain injury. Frontiers in Neurosci-
ence, 9, 112.
Choi, B. Y., Jang, B. G., Kim, J. H., Lee, B. E., Sohn, M., Song, H. K., &
Suh, S. W. (2012). Prevention of traumatic brain injury-induced neuro-
nal death by inhibition of NADPH oxidase activation. Brain Research,
1481, 49–58.
Chouchani, E. T., Pell, V. R., Gaude, E., Aksentijevic, D., Sundier, S. Y.,
Robb, E. L., … Murphy, M. P. (2014). Ischaemic accumulation of succi-
nate controls reperfusion injury through mitochondrial ROS. Nature,
515, 431–435.
Chu, H. X., Broughton, B. R., Kim, H. A., Lee, S., Drummond, G. R., &
Sobey, C. G. (2015). Evidence that Ly6C(hi) monocytes are protective
in acute ischemic stroke by promoting M2 macrophage polarization.
Stroke, 46, 1929–1937.
Corps, K. N., Roth, T. L., & McGavern, D. B. (2015). Inflammation and neu-
roprotection in traumatic brain injury. JAMA Neurology, 72, 355–362.
Dudvarski Stankovic, N., Teodorczyk, M., Ploen, R., Zipp, F., &
Schmidt, M. H. (2016). Microglia-blood vessel interactions: A
double-edged sword in brain pathologies. Acta Neuropathologica, 131,
347–363.
Fisher, M., Feuerstein, G., Howells, D. W., Hurn, P. D., Kent, T. A.,
Savitz, S. I., & Lo, E. H. (2009). Update of the stroke therapy academic
industry roundtable preclinical recommendations. Stroke, 40,
2244–2250.
Gao, L., Laude, K., & Cai, H. (2008). Mitochondrial pathophysiology, reac-
tive oxygen species, and cardiovascular diseases. The Veterinary Clinics
of North America. Small Animal Practice, 38(137–155), vi.
Garaude, J., Acin-Perez, R., Martinez-Cano, S., Enamorado, M., Ugolini, M.,
Nistal-Villan, E., … Sancho, D. (2016). Mitochondrial respiratory-chain
adaptations in macrophages contribute to antibacterial host defense.
Nature Immunology, 17, 1037–1045.
Gyoneva, S., Kim, D., Katsumoto, A., Kokiko-Cochran, O. N., Lamb, B. T., &
Ransohoff, R. M. (2015). Ccr2 deletion dissociates cavity size and tau
pathology after mild traumatic brain injury. Journal of Neuroinflamma-
tion, 12, 228.
ABE ET AL.
Gyoneva, S., & Ransohoff, R. M. (2015). Inflammatory reaction after trau-
matic brain injury: Therapeutic potential of targeting cell-cell communi-
cation by chemokines. Trends in Pharmacological Sciences, 36, 471–480.
Hall, E. D., Wang, J. A., & Miller, D. M. (2012). Relationship of nitric oxide
synthase induction to peroxynitrite-mediated oxidative damage during
the first week after experimental traumatic brain injury. Experimental
Neurology, 238, 176–182.
Hattori, T., Kaji, M., Ishii, H., Jureepon, R., Takarada-Iemata, M., Minh
Ta, H., … Hori, O. (2017). CD38 positively regulates postnatal develop-
ment of astrocytes cell-autonomously and oligodendrocytes
non-cell-autonomously. Glia, 65, 974–989.
Heid, M. E., Keyel, P. A., Kamga, C., Shiva, S., Watkins, S. C., & Salter, R. D.
(2013). Mitochondrial reactive oxygen species induces
NLRP3-dependent lysosomal damage and inflammasome activation.
Journal of Immunology, 191, 5230–5238.
Higaki, H., Choudhury, M. E., Kawamoto, C., Miyamoto, K., Islam, A.,
Ishii, Y., … Tanaka, J. (2016). The hypnotic bromovalerylurea amelio-
rates 6-hydroxydopamine-induced dopaminergic neuron loss while
suppressing expression of interferon regulatory factors by microglia.
Neurochemistry International, 99, 158–168.
Hsieh, C. L., Niemi, E. C., Wang, S. H., Lee, C. C., Bingham, D., Zhang, J., …
Nakamura, M. C. (2014). CCR2 deficiency impairs macrophage infiltra-
tion and improves cognitive function after traumatic brain injury. Jour-
nal of Neurotrauma, 31, 1677–1688.
Islam, A., Choudhury, M. E., Kigami, Y., Utsunomiya, R., Matsumoto, S.,
Watanabe, H., … Tanaka, J. (2018). Sustained anti-inflammatory effects
of TGF-beta1 on microglia/macrophages. Biochimica et Biophysica Acta,
1864, 721–734.
Karve, I. P., Taylor, J. M., & Crack, P. J. (2016). The contribution of astro-
cytes and microglia to traumatic brain injury. British Journal of Pharma-
cology, 173, 692–702.
Kawasaki, S., Abe, N., Ohtake, F., Islam, A., Choudhury, M. E.,
Utsunomiya, R., … Tanaka, J. (2017). Effects of hypnotic bromovalery-
lurea on microglial BV2 cells. Journal of Pharmacological Sciences, 134,
116–123.
Kikuchi, S., Nishihara, T., Kawasaki, S., Abe, N., Kuwabara, J.,
Choudhury, M. E., … Tanaka, J. (2015). The ameliorative effects of a
hypnotic bromvalerylurea in sepsis. Biochemical and Biophysical
Research Communications, 459, 319–326.
Konishi, H., Kobayashi, M., Kunisawa, T., Imai, K., Sayo, A., Malissen, B., …
Kiyama, H. (2017). Siglec-H is a microglia-specific marker that discrimi-
nates microglia from CNS-associated macrophages and
CNS-infiltrating monocytes. Glia, 65, 1927–1943.
Lashgari, R., Motamedi, F., Zahedi Asl, S., Shahidi, S., & Komaki, A. (2006).
Behavioral and electrophysiological studies of chronic oral administra-
tion of L-type calcium channel blocker verapamil on learning and mem-
ory in rats. Behavioural Brain Research, 171, 324–328.
Latz, E. (2010). NOX-free inflammasome activation. Blood, 116,
1393–1394.
Lenzlinger, P. M., Morganti-Kossmann, M. C., Laurer, H. L., &
McIntosh, T. K. (2001). The duality of the inflammatory response to
traumatic brain injury. Molecular Neurobiology, 24, 169–181.
Liu, S., Zhang, L., Wu, Q., Wu, Q., & Wang, T. (2013). Chemokine CCL2
induces apoptosis in cortex following traumatic brain injury. Journal of
Molecular Neuroscience, 51, 1021–1029.
Matsumoto, H., Kumon, Y., Watanabe, H., Ohnishi, T., Shudou, M.,
Chuai, M., … Tanaka, J. (2008). Accumulation of macrophage-like cells
expressing NG2 proteoglycan and Iba1 in ischemic core of rat brain
after transient middle cerebral artery occlusion. Journal of Cerebral
Blood Flow and Metabolism, 28, 149–163.
Matsumoto, H., Kumon, Y., Watanabe, H., Ohnishi, T., Shudou, M., Ii, C., …
Tanaka, J. (2007). Antibodies to CD11b, CD68, and lectin label neutro-
phils rather than microglia in traumatic and ischemic brain lesions. Jour-
nal of Neuroscience Research, 85, 994–1009.
McKee, C. A., & Lukens, J. R. (2016). Emerging roles for the immune sys-
tem in traumatic brain injury. Frontiers in Immunology, 7, 556.
Miro-Mur, F., Perez-de-Puig, I., Ferrer-Ferrer, M., Urra, X., Justicia, C.,
Chamorro, A., & Planas, A. M. (2016). Immature monocytes recruited
to the ischemic mouse brain differentiate into macrophages with fea-
tures of alternative activation. Brain, Behavior, and Immunity, 53,
18–33.
15
Morganti, J. M., Jopson, T. D., Liu, S., Riparip, L. K., Guandique, C. K.,
Gupta, N., … Rosi, S. (2015). CCR2 antagonism alters brain macrophage
polarization and ameliorates cognitive dysfunction induced by trau-
matic brain injury. The Journal of Neuroscience, 35, 748–760.
Neher, J. J., Emmrich, J. V., Fricker, M., Mander, P. K., Thery, C., &
Brown, G. C. (2013). Phagocytosis executes delayed neuronal death
after focal brain ischemia. Proceedings of the National Academy of Sci-
ences of the United States of America, 110, E4098–E4107.
Nell, V., & Brown, D. S. (1991). Epidemiology of traumatic brain injury in
Johannesburg--II. Morbidity, mortality and etiology. Social Science &
Medicine, 33, 289–296.
Neumann, H., Kotter, M. R., & Franklin, R. J. (2009). Debris clearance by
microglia: An essential link between degeneration and regeneration.
Brain, 132, 288–295.
Nishihara, T., Ochi, M., Sugimoto, K., Takahashi, H., Yano, H., Kumon, Y., …
Tanaka, J. (2011). Subcutaneous injection containing IL-3 and GM-CSF
ameliorates stab wound-induced brain injury in rats. Experimental Neu-
rology, 229, 507–516.
Nishioka, R., Sugimoto, K., Aono, H., Mise, A., Choudhury, M. E.,
Miyanishi, K., … Tanaka, J. (2016). Treadmill exercise ameliorates
ischemia-induced brain edema while suppressing Na(+)/H(+) exchanger
1 expression. Experimental Neurology, 277, 150–161.
Patel, A. R., Ritzel, R., McCullough, L. D., & Liu, F. (2013). Microglia and
ischemic stroke: A double-edged sword. International Journal of Physiol-
ogy, Pathophysiology, and Pharmacology, 5, 73–90.
Rastogi, L., Godbole, M. M., Ray, M., Rathore, P., Rathore, P., Pradhan, S.,
… Pandey, C. M. (2006). Reduction in oxidative stress and cell death
explains hypothyroidism induced neuroprotection subsequent to
ischemia/reperfusion insult. Experimental Neurology, 200, 290–300.
Ritzel, R. M., Patel, A. R., Grenier, J. M., Crapser, J., Verma, R.,
Jellison, E. R., & McCullough, L. D. (2015). Functional differences
between microglia and monocytes after ischemic stroke. Journal of
Neuroinflammation, 12, 106.
Robertson, C. L., Scafidi, S., McKenna, M. C., & Fiskum, G. (2009). Mito-
chondrial mechanisms of cell death and neuroprotection in pediatric
ischemic and traumatic brain injury. Experimental Neurology, 218,
371–380.
Roth, T. L., Nayak, D., Atanasijevic, T., Koretsky, A. P., Latour, L. L., &
McGavern, D. B. (2014). Transcranial amelioration of inflammation and
cell death after brain injury. Nature, 505, 223–228.
Satoh, J., Kino, Y., Asahina, N., Takitani, M., Miyoshi, J., Ishida, T., &
Saito, Y. (2016). TMEM119 marks a subset of microglia in the human
brain. Neuropathology, 36, 39–49.
Semple, B. D., Bye, N., Rancan, M., Ziebell, J. M., &
Morganti-Kossmann, M. C. (2010). Role of CCL2 (MCP-1) in traumatic
brain injury (TBI): Evidence from severe TBI patients and CCL2−/−
mice. Journal of Cerebral Blood Flow and Metabolism, 30, 769–782.
Semple, B. D., Bye, N., Ziebell, J. M., & Morganti-Kossmann, M. C. (2010).
Deficiency of the chemokine receptor CXCR2 attenuates neutrophil
infiltration and cortical damage following closed head injury. Neurobiol-
ogy of Disease, 40, 394–403.
Semple, B. D., Kossmann, T., & Morganti-Kossmann, M. C. (2010). Role of
chemokines in CNS health and pathology: A focus on the CCL2/CCR2
and CXCL8/CXCR2 networks. Journal of Cerebral Blood Flow and
Metabolism, 30, 459–473.
Semple, B. D., Trivedi, A., Gimlin, K., & Noble-Haeusslein, L. J. (2015). Neu-
trophil elastase mediates acute pathogenesis and is a determinant of
long-term behavioral recovery after traumatic injury to the immature
brain. Neurobiology of Disease, 74, 263–280.
Serrano-Nascimento, C., da Silva, T. S., Nicola, J. P., Nachbar, R. T.,
Masini-Repiso, A. M., & Nunes, M. T. (2014). The acute inhibitory
effect of iodide excess on sodium/iodide symporter expression and
activity involves the PI3K/Akt signaling pathway. Endocrinology, 155,
1145–1156.
Shichita, T., Hasegawa, E., Kimura, A., Morita, R., Sakaguchi, R., Takada, I.,
… Yoshimura, A. (2012). Peroxiredoxin family proteins are key initiators
of post-ischemic inflammation in the brain. Nature Medicine, 18,
911–917.
Sinz, E. H., Kochanek, P. M., Dixon, C. E., Clark, R. S., Carcillo, J. A.,
Schiding, J. K., … Billiar, T. R. (1999). Inducible nitric oxide synthase is
16
an endogenous neuroprotectant after traumatic brain injury in rats and
mice. The Journal of Clinical Investigation, 104, 647–656.
Smirkin, A., Matsumoto, H., Takahashi, H., Inoue, A., Tagawa, M., Ohue, S.,
… Tanaka, J. (2010). Iba1(+)/NG2(+) macrophage-like cells expressing a
variety of neuroprotective factors ameliorate ischemic damage of the
brain. Journal of Cerebral Blood Flow and Metabolism, 30, 603–615.
Song, L., Lee, C., & Schindler, C. (2011). Deletion of the murine scavenger
receptor CD68. Journal of Lipid Research, 52, 1542–1550.
Song, S., Kong, X., Acosta, S., Sava, V., Borlongan, C. V., & Sanchez-Ramos, J.
(2017). Effects of an inhibitor of monocyte recruitment on recovery from
traumatic brain injury in mice treated with granulocyte Colony-stimulating
factor. International Journal of Molecular Science, 18, E1418.
Sugimoto, K., Nishioka, R., Ikeda, A., Mise, A., Takahashi, H., Yano, H., …
Tanaka, J. (2014). Activated microglia in a rat stroke model express
NG2 proteoglycan in peri-infarct tissue through the involvement of
TGF-beta1. Glia, 62, 185–198.
Tanaka, J., Toku, K., Matsuda, S., Sudo, S., Fujita, H., Sakanaka, M., …
Maeda, N. (1998). Induction of resting microglia in culture medium
devoid of glycine and serine. Glia, 24, 198–215.
Taylor, R. A., Chang, C. F., Goods, B. A., Hammond, M. D., Mac Grory, B.,
Ai, Y., … Sansing, L. H. (2017). TGF-beta1 modulates microglial pheno-
type and promotes recovery after intracerebral hemorrhage. The Jour-
nal of Clinical Investigation, 127, 280–292.
Tei, N., Tanaka, J., Sugimoto, K., Nishihara, T., Nishioka, R., Takahashi, H.,
… Ohnishi, T. (2013). Expression of MCP-1 and fractalkine on endothe-
lial cells and astrocytes may contribute to the invasion and migration
of brain macrophages in ischemic rat brain lesions. Journal of Neurosci-
ence Research, 91, 681–693.
Toku, K., Tanaka, J., Yano, H., Desaki, J., Zhang, B., Yang, L., … Maeda, N.
(1998). Microglial cells prevent nitric oxide-induced neuronal apoptosis
in vitro. Journal of Neuroscience Research, 53, 415–425.
Valavanidis, A., Vlachogianni, T., & Fiotakis, C. (2009). 8-hydroxy-2’
-deoxyguanosine (8-OHdG): A critical biomarker of oxidative stress
and carcinogenesis. Journal of Environmental Science and Health. Part C,
Environmental Carcinogenesis & Ecotoxicology Reviews, 27, 120–139.
Vilalta, A., & Brown, G. C. (2017). Neurophagy, the phagocytosis of live
neurons and synapses by glia, contributes to brain development and
disease. The FEBS Journal. https://doi.org/10.1111/febs.14323
ABE ET AL.
Wattananit, S., Tornero, D., Graubardt, N., Memanishvili, T., Monni, E.,
Tatarishvili, J., … Kokaia, Z. (2016). Monocyte-derived macrophages
contribute to spontaneous long-term functional recovery after stroke
in mice. The Journal of Neuroscience, 36, 4182–4195.
Weaver, K. D., Branch, C. A., Hernandez, L., Miller, C. H., &
Quattrocchi, K. B. (2000). Effect of leukocyte-endothelial adhesion
antagonism on neutrophil migration and neurologic outcome after cor-
tical trauma. The Journal of Trauma, 48, 1081–1090.
Webster, K. M., Sun, M., Crack, P., O’Brien, T. J., Shultz, S. R., &
Semple, B. D. (2017). Inflammation in epileptogenesis after traumatic
brain injury. Journal of Neuroinflammation, 14, 10.
West, A. P., Brodsky, I. E., Rahner, C., Woo, D. K., Erdjument-Bromage, H.,
Tempst, P., … Ghosh, S. (2011). TLR signalling augments macrophage
bactericidal activity through mitochondrial ROS. Nature, 472,
476–480.
Woodcock, T., & Morganti-Kossmann, M. C. (2013). The role of markers of
inflammation in traumatic brain injury. Frontiers in Neurology, 4, 18.
Yokoyama, A., Sakamoto, A., Kameda, K., Imai, Y., & Tanaka, J. (2006). NG2
proteoglycan-expressing microglia as multipotent neural progenitors in
normal and pathologic brains. Glia, 53, 754–768.
Zhang, Q. G., Laird, M. D., Han, D., Nguyen, K., Scott, E., Dong, Y., …
Brann, D. W. (2012). Critical role of NADPH oxidase in neuronal oxida-
tive damage and microglia activation following traumatic brain injury.
PLoS One, 7, e34504.
SUPPOR TI NG I NFORMATION
Additional supporting information may be found online in the Sup-
porting Information section at the end of the article.
How to cite this article: Abe N, Choudhury ME,
Watanabe M, et al. Comparison of the detrimental features of
microglia and infiltrated macrophages in traumatic brain injury:
A study using a hypnotic bromovalerylurea. Glia. 2018;1–16.
https://doi.org/10.1002/glia.23469