Minocycline reduces glioma expansion and invasion by attenuating microglial
D.S. Markovica,b,1, K. Vinnakotaa,1, N. van Rooijenc, J. Kiwitb, M. Synowitza,c, R. Glassa,1, H. Kettenmanna,⇑,1
aMax Delbrück Center for Molecular Medicine, Cellular Neuroscience, Robert Rössle Str. 10, 13125 Berlin, Germany
bHelios Clinic, Dept. of Neurosurgery, Schwanebecker Chausse 50, 13125 Berlin, Germany
cVrije Universiteit, VUMC, Department of Molecular Cell Biology, Faculty of Medicine, Van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands
a r t i c l ei n f o
Received 24 November 2010
Received in revised form 18 January 2011
Accepted 25 January 2011
Available online 13 February 2011
a b s t r a c t
Glioma cells release soluble factors, which induce the expression of membrane type 1 matrix metallopro-
tease (MT1-MMP) in tumor associated microglia and then exploit MT1-MMP mediated matrix degrada-
tion for invasion. Here, we show that minocycline blocked the increase in MT1-MMP expression and
activity in cultivated microglia stimulated with glioma conditioned medium. Glioma growth within an
organotypic brain slice preparation was reduced by minocycline and this reduction depended on the
presence of microglia. Glioma growth in an experimental mouse model was strongly reduced by the addi-
tion of minocycline to drinking water, compared to untreated controls. Coherently, we observed in our
orthotopic glioma implantation model, that MT1-MMP was abundantly expressed in glioma associated
microglia in controls, but was strongly attenuated in tumors of minocycline treated animals. Overall,
our study indicates that the clinically approved antibiotic minocycline is a promising new candidate
for adjuvant therapy against malignant gliomas.
? 2011 Elsevier Inc. All rights reserved.
Malignant gliomas are the most frequent primary tumors of the
brain. To invade the brain parenchyma, glioma cells secrete extra-
cellular matrix (ECM) degrading enzymes. An important group of
proteolytic enzymes mediating this process are the matrix metal-
loproteases (Osenkowski et al., 2004). The glioma infiltrative prop-
erties are mainly promoted by matrix metalloprotease 2 (MMP-2)
and its extracellular activator, the membrane type 1 matrix metal-
loprotease (MT1-MMP or MMP-14) (Rao, 2003; Mayes et al., 2006;
Guo et al., 2005; Markovic et al., 2005, 2009). Both human and
experimental gliomas are characterized by a high density of
microglial cells residing within and around the invasive edge of
the tumor (Roggendorf et al., 1996; Badie and Schartner, 2001).
Glioma tissue consists of as much as 30% of microglial cells while
other immune cells like T-lymphocytes are much less abundant
(Strik et al., 2004). We and others have recently shown that
microglial cells promote glioma growth (Watters et al., 2005).
We demonstrated that soluble factor(s) released from glioma cells
increased the expression of microglial MT1-MMP, the enzyme that
cleaves the inactive pro-MMP-2 into the active MMP-2, via activa-
tion of toll-like receptors and the p38 MAP kinase pathway in
microglia (Markovic et al., 2009).
Minocycline hydrochloride, hereafter referred to as minocy-
cline, has been demonstrated to block the p38 MAP kinase path-
activation into a pro-inflammatory phenotype (Suk, 2004). Mino-
cycline is a semi-synthetic broad spectrum tetracycline antibiotic
with bacteriostatic functions and has been approved for over
30 years by the FDA to treat chronic conditions such as acne and
rosacea (Seukeran et al., 1997; Yong et al., 2004). It is a small,
highly lipophilic molecule (495 kDa), readily absorbed from the
gut after oral intake and capable of crossing the intact blood–brain
barrier (Seukeran et al., 1997). We have therefore tested whether
minocycline would interfere with the pro-tumorigenic action of
microglial cells in vitro and in vivo.
2. Materials and methods
2.1. Minocycline hydrochloride
Minocycline (#M9511, Sigma–Aldrich, Germany) was freshly
prepared in sterile water and used in vitro (200 nM) and in vivo
(10 ng/ml) as indicated.
2.2. Cell culture
Murine GL261 glioma cells (National Cancer Institute, NCI-
Frederick, MD, USA) were cultivated in 10% FCS/DMEM while pri-
mary microglia were grown as described previously (Markovic
0889-1591/$ - see front matter ? 2011 Elsevier Inc. All rights reserved.
⇑Corresponding author. Fax: +49 30 94 06 38 19.
E-mail address: firstname.lastname@example.org (H. Kettenmann).
1These authors contributed equally.
Brain, Behavior, and Immunity 25 (2011) 624–628
Contents lists available at ScienceDirect
Brain, Behavior, and Immunity
journal homepage: www.elsevier.com/locate/ybrbi
et al., 2005). Briefly, the cerebral cortices of newborn C57Bl/6 mice
were carefully cleaned of blood vessels and meninges, then the tis-
sue was enzymatically digested with Trypsin/DNase and mechan-
ically dissociated with a fire-polished glass pasteur pipette.
Cultures were incubated at 37 ?C, 5% CO2atmosphere and the next
day adherent cells were washed 5 times with PBS. After culturing
for one week followed by culturing with L929 (fibroblast cell line)
conditioned medium for two days the microglia are now seen as
floating cells or as semi-adherent cells on top of an astrocytic
monolayer. These were then harvested by shaking the culture flask.
2.3. Organotypic brain slice culture (OBSC) model
Brain tissue was derived from 16 day-old male C57BL/6 mice.
Coronal brain slices (250 lm) were sectioned by a Vibratome
(Leica, Germany) and cultured in the upper chamber of a transwell
tissue insert (Falcon cell culture inserts, 0.4 lm polycarbonate
membrane, 6-well format, BD Labware, Germany), which was in-
serted into a 6-well plate (BD Labware, Germany). Then, 5000 EGFP
transfected GL261 cells were injected into the OBSCs using a 1 ll
syringe with a blunt tip (Mikroliterspritze 7001N, Hamilton,
Switzerland) mounted onto a self constructed micromanipulator.
Microglial cells were depleted from the slices by the application
of clodronate-filled liposomes for one day (Markovic et al., 2005).
Finally, the tumor sizes were analyzed in fluorescent images using
the Image Pro plus software (Media Cybernetics, USA).
2.4. Western blot
Whole-cell protein extracts were prepared from microglia cells
as described earlier (Lohi et al., 2000). The antibodies recognizing
Rabbit anti- MT1-MMP (M5808) and Mouse monoclonal anti-b-Ac-
tin-Peroxidase (A3854) were purchased from Sigma–Aldrich,
2.5. MT1-MMP activity assay
determine the amount of active MT1-MMP (Matrix Metalloprotein-
assay is based on a detection enzyme (pro-urokinase) that is acti-
vated only by active MT1-MMP through a single proteolytic event.
The activated urokinase activity was detected by a specific chromo-
genic peptide substrate (S-2444™ peptide) and the color reaction
was quantified at 405 nm in a microplate reader (Perkin Wallac,
2.6. Semi-quantitative and quantitative RT-PCR
RNA was isolated using the RNeasy-Microkit (Qiagen, Hilden,
Germany) and first-strand cDNA synthesis was done using Super-
Script II reverse transcriptase (Invitrogen, Karlsruhe, Germany)
taking 1 lg RNA and oligo dT primers. PCR was performed with
the High Fidelity SuperMix PCR kit (Invitrogen, Karlsruhe,
Germany). Real-time quantitative PCR (Q-PCR) was carried out
using SYBR Green (Roche Diagnostics GmbH, Mannheim, Germany)
in the Realplex thermal cycler (Eppendorf, Hamburg, Germany).
Total RNAs isolated from four independent experiments were used
for these analyses. The cDNAs were amplified in duplicates. The
mouse primers used are as follows- MT1-MMP sense 50-GTGCCCT
ATGCCTACATCCG-30, anti-sense 50-CAGCCACCAAGAAGATGTCA-30
(Q-PCR): MT1-MMP sense 50-GGATACCCAATGCCCATTGGCCA-30,
anti-sense 50-CCATTGGGCATCCAGAAGAGA-30(RT-PCR) and for Ac-
tin sense 50-CCCTGAAGTACCCCATTGAA-30, anti-sense 50-GTGGACA
GTGAGGCCAAGAT-30(both RT & Q-PCRs). The primers used for
The stainings were performed using the primary anti-mouse
antibodies to Rabbit anti-Iba1 (Wako Pure Chemicals, Japan) and
Mouse anti-MT1-MMP (Calbiochem, Darmstadt, Germany) on
40 lm thick, free-floating brain sections as previously described
(Markovic et al., 2009).
2.8. In vivo tumor model
Wild type C57BL/6 mice (Charles River Breeding Laboratories,
Sulzfeld, Germany) were bred and maintained according to
German animal care regulations (Glass et al., 2005). Mice were
anesthetized, immobilized and mounted onto a stereotactic head
holder (David Kopf Instruments, USA) in the flat-skull position.
After skin incision 1 mm anterior and 1.5 mm lateral to the breg-
ma, the skull was carefully drilled with 20 G needle tip. Then a
1 ll syringe with a blunt tip (Mikroliterspritze 7001 N, Hamilton,
Switzerland) was inserted to a depth of 4 mm and retracted to a
depth of 3 mm from the dural surface into the right caudate puta-
men. Over 2 min, 1 ll (2 ? 104cells/ll) of glioma cell suspension
was slowly injected. The needle was then slowly retracted from
the injection canal and the skin was sutured with a surgical sewing
cone (Johnson & Johnson International). After the surgery the mice
were kept warm until awake and their postoperative condition was
monitored daily. To analyze tumor growth 14 or 21 days after sur-
gery, mice were euthanized with Ketamine, brains perfused and
fixed (4% paraformaldehyde), cryopreserved and immunolabeled
for specific markers as stated above.
2.9. Unbiased stereology
The glioma tumor volume was quantified according to the
Cavalieri principle by determining tumor area in every 12th
40 lm brain slice and then multiplying this area by 12 ? 40 lm
(Glass et al., 2005) to calculate the tumor volume in both the
experimental and control groups of mice.
3.1. Minocycline blunts the pro-tumorigenic effect of microglial MT1-
MMP expression on gliomas
To test if minocycline modulates the increase in MT1-MMP
expression by glioma conditioned medium (GCM), we stimulated
microglia for 3 h and 6 h with GCM and performed RT-PCR,
Q-PCR, western blot and an activity assay. Indeed minocycline
attenuated the increase in MT1-MMP mRNA 3 h and 6 h after stim-
ulation with GCM as quantified through semi-quantitative (Fig. 1A)
and quantitative RT-PCRs (Fig. 1B). We detected an average two-
fold amplification of MT1-MMP transcription 6 h after GCM stimu-
lation while minocycline abrogated this increase (Fig. 1B). Protein
extracts from primary microglia stimulated with GCM alone or
GCM and minocycline were analyzed by western blot for MT1-
MMP protein expression. We obtained a similar result as seen with
the RT-PCR analyses, namely a stimulation of MT1-MMP expres-
sion after 3 h and 6 h and a blockade by minocycline, demonstrated
by decreased amounts of degraded product at 57 kDa (Fig. 1C).
Hence the MT1-MMP mRNA was dramatically reduced after
minocycline treatment (Fig. 1A and B) while in the western blot
the degraded products at 57 kDa were also reduced, indicating a
decreased turnover of MT1-MMP (Fig. 1C). An MT1-MMP activity
D.S. Markovic et al./Brain, Behavior, and Immunity 25 (2011) 624–628
assay further confirmed the mRNA and protein data. The
MT1-MMP activity in microglia was significantly elevated after
incubation with GCM (Fig. 1D).The enzymatic activity increased
by 220% and 217% after 3 h and 6 h, respectively. This difference
between 3 h and 6 h stimulation with GCM was not significant
and we assume that we reached the saturation level of the assay.
In the presence of minocycline, GCM no longer stimulated MT1-
MMP activity, but rather reduced the activity to 51% and 73% after
3 h and 6 h, respectively, as compared to the untreated control.
While GL261 cells expressed the matrix metalloprotease 2
(MMP-2) as we previously reported, we did not observe an expres-
sion of MMP-2 in unstimulated or GCM stimulated microglial cells
3.2. Minocycline interferes with glioma growth in a cultured slice
model depending on the presence of microglia
We tested the effect of minocycline on glioma expansion in
organotypic brain slices by injecting EGFP labeled GL261 cells in
the slices and analyzing the area occupied by the glioma cells
5 days after injection. Minocycline reduced the average size occu-
pied by glioma cells to 57% as compared to control slices. By apply-
ing clodronate filled liposomes, we could effectively eliminate
microglial cells from the slices. As compared to control slices,
microglia depletion resulted in a reduction of tumor size to 58%.
In microglia depleted slices, minocycline did not affect the tumor
size and its effect was similar to the microglia-depleted slices with-
out minocycline, namely 56% (Fig. 2A). This indicates that minocy-
cline reduces the tumor-promoting activity of microglia.
3.3. Oral administration of minocycline reduced glioma growth in vivo
To quantify the effect of minocycline on glioma expansion
in vivo, we induced experimental gliomas by stereotactic injection
of EGFP labeled GL261 cells into brains of 30-day old C57Bl/6 mice.
We used two experimental paradigms to apply minocycline. In the
first, mice were provided with minocycline immediately after gli-
oma implantation for the entire two weeks of tumor growth. In
the second approach, we let the experimental gliomas develop
for one week and then administered minocycline to the tumor
bearing mice for an additional two weeks. In both paradigms the
drug was added to the drinking water at a dosage of 10 ng/ml,
while the control group received pure water. Mice were sacrificed
after two weeks of minocycline treatment and tumor volumes
were calculated according to the Cavalieri method. In mice receiv-
ing minocycline during the entire time of tumor growth, the differ-
ence in average tumor volume was reduced strongly to 1.05 mm3
as compared to 4.71 mm3in the control group (Fig. 2B). Also in
the second experimental paradigm, minocycline significantly
reduced glioma growth to 4.07 mm3as compared to the control
mice where the average tumor volume was 7.52 mm3(Fig. 2C).
Hence, minocycline treatment reduced glioma growth in the
experimental mouse model, even when minocycline was applied
at a later time point during the tumor growth.
Fig. 1. Minocycline blunts the pro-tumorigenic effect of MT1-MMP expression in glioma-associated microglia. (A) Semi-quantitative RT-PCR analysis of MT1-MMP
expression revealed changes in primary cultured microglia stimulated with GCM or GCM and minocycline after 3 h and 6 h (with b-Actin as the internal control). A reduction
in MT1-MMP gene expression was observed when the microglia were stimulated with GCM together with 200 nM minocycline. (B) This was further supported through real-
time quantitative PCR analysis to assess MT1-MMP induction in microglia by GCM alone or GCM and minocycline for 3 h and 6 h. GCM treatment for 6 h alone lead up to two
fold enhancement of the MT1-MMP expression. After GCM stimulation together with minocycline, the overexpression of MT1-MMP is abrogated. (C) A similar result on MT1-
MMP expression was observed by western blot analysis of whole cell protein extracts from microglia after they were stimulated with GCM alone or GCM and minocycline;
note that GCM induced an increase in MT1-MMP expression after 6 h whereas co-treatment with 200nM minocycline prevented an increase in MT1-MMP expression (arrow).
Fragments of degraded MT1-MMP are also increased (arrowhead), indicating a higher turnover activity of the active MT1-MMP after stimulation with GCM without
minocycline. (D) The increase in MT1-MMP activity after GCM treatment is inhibited by minocycline treatment. For this, primary microglia cells were stimulated with GCM
alone or with GCM and minocycline for the indicated time points (3 h and 6 h). The activity was normalized to controls (i.e., the baseline activity was measured in non-
stimulated control microglia).⁄Significance was assessed at P < 0.05 and⁄⁄significance at P < 0.01.
D.S. Markovic et al./Brain, Behavior, and Immunity 25 (2011) 624–628
3.4. The MT1-MMP expression in glioma-associated microglia is
reduced after minocycline administration in vivo
We analyzed the expression level of MT1-MMP by immunola-
beling with a MT1-MMP specific antibody in tissue sections from
mice injected with the EGFP-GL261 cells, two weeks after intrace-
rebral injection. The microglial cells were identified by labeling
with the microglia/macrophage-specific antibody against Iba1.
The level of MT1-MMP labeling was increased in microglia in close
proximity to the tumor border and MT1-MMP immunoreactivity
was particularly pronounced when the microglia were in close
MT1-MMP was significantly reduced in mice which had received
minocycline in the drinking water during the two week period
Fig. 2. Minocycline interferes with the tumor growth and expansion in our in situ and in vivo glioma mouse models. (A) The effect of minocycline on the growth of the glioma
was assessed in an organotypic brain slice model. The brain slices were inoculated with glioma cells and the relative tumor size was measured in control and microglia
depleted brain slices, with or without minocycline treatment. Defining the tumor size in the control slices as 100%, we observed that minocycline treatment resulted in a
significant reduction in tumor size to nearly 60%. Furthermore, in (B) and (C) in vivo minocycline administration resulted in a decrease in the tumor volume relative to control
group which did not receive any minocycline. Two groups of mice (control and experimental) were implanted with EGFP-GL261 glioma cells. While one group received
minocycline through drinking water from day 0, the other group received minocycline after 1 week. In both groups, minocycline was applied for two weeks and the relative
tumor volume was significantly reduced upon minocycline treatment as compared to the controls. In mice which received minocycline from day 0, the relative tumor volume
was 1 mm3compared to nearly 5 mm3in the untreated control group. In mice which received minocycline from day 7 after tumor inoculation, the tumor volume was 4 mm3
after three weeks compared to nearly 8 mm3in the untreated control group.⁄⁄Significance was calculated at P < 0.01. Note that the tumor sizes in C are larger than in B due to
the prolonged tumor development time (three weeks). (D) Expression of MT1-MMP was analyzed by immunolabeling slices obtained from mice treated with minocycline for
two weeks after EGFP-GL261 glioma cell inoculation. Microglia cells were labeled by Iba1 (in red) and it was evident that they accumulated less at the tumor boundary after
minocycline treatment. The MT1-MMP labeling shown in brown is strongly reduced in microglial cells associated with glioma tumors in minocycline treated mice. Note that
the in vivo administration of minocycline not only reduced the MT1-MMP expression in mice which received minocycline compared to the control mice but also the tumor
size. The scale bar is 300 lm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
D.S. Markovic et al./Brain, Behavior, and Immunity 25 (2011) 624–628
compared to the control animals which received plain drinking Download full-text
water (Fig. 2D). Importantly, the demonstrated downregulation
of MT1-MMP in in vivo experiments furthermore indicated the
strong therapeutical effect of continuous minocycline treatment.
Hence, in the glioma bearing mice treated with minocycline, the
MT1-MMP expression in microglia was lower as compared to the
enhanced expression in control mice.
We have previously shown that microglial cells promote glioma
growth by upregulation of MT1-MMP (Markovic et al., 2009). Here
we have shown that minocycline, a medically approved antibiotic,
blunts the pro-tumorigenic effect of glioma associated microglia.
Minocycline has been reported to block p38 MAP kinase activation
in microglia (Nikodemova et al., 2006) and we have previously
shown that the upregulation of MT1-MMP is p38 MAP kinase
dependent (Markovic et al., 2009). In an animal model of multiple
sclerosis, minocycline inhibited MMP activity, reduced production
of MMP-9 and decreased the transmigration of T lymphocytes
(Brundula et al., 2002). In vivo studies on intracerebral hemorrhage
in rodents demonstrated that MMP-2 and MMP-9 (Machado et al.,
2006) and MMP-12 (Power et al., 2003) are downregulated and
inhibited in their activity and expression after treatment with min-
ocycline. Moreover, minocycline can also affect other microglial
functions since it has been previously reported that it is also capa-
ble of suppressing chemokine secretion (Kremlev et al., 2004). The
microglial MT1-MMP overexpression may not only promote gli-
oma invasion, but may also contribute to revascularization of the
tumor (Beliën et al., 1999).
Recent findings suggest that minocycline can be neuroprotec-
tive and anti-inflammatory in various neurodegenerative disorders
(Yong et al., 2004). The neuroprotective action of minocycline may
include its inhibitory effect on 5-lipoxygenase, an inflammatory
enzyme associated with brain aging, and is being studied for its
therapeutic use in Alzheimer disease patients (Seabrook et al.,
2006). Minocycline also confers neuroprotection in mouse models
of amyotrophic lateral sclerosis (Kriz et al., 2002) and Huntington’s
disease (Chen et al., 2000;Wang et al., 2003) and has recently been
shown to stabilize the course of Huntington’s disease in humans
over a 2-year period (Bonelli and Kapfhammer, 2003). Our own
study implies that minocycline targets the pro-tumorigenic effects
of glioma associated microglia and may thus be a candidate for
concomitant adjuvant therapies to current treatment modalities
in glioma patients.
Badie, B., Schartner, J., 2001. Microsc. Res. Tech. 54 (2), 106–113.
Beliën, A.T., Paganetti, P.A., Schwab, M.E., 1999. J. Cell Biol. 144 (2), 373–384.
Bonelli, R.M., Kapfhammer, H.P., 2003. J. Psychopharmacol. 17 (4), 461.
Brundula, V., Rewcastle, N.B., Metz, L.M., Bernard, C.C., Yong, V.W., 2002. Brain 125
(Pt. 6), 1297–1308.
Chen, M., Ona, V.O., Li, M., Ferrante, R.J., Fink, K.B., Zhu, S., Bian, J., Guo, L., Farrell,
L.A., Hersch, S.M., Hobbs, W., Vonsattel, J.P., Cha, J.H., Friedlander, R.M., 2000.
Nat. Med. 6 (7), 797–801.
Glass, R., Synowitz, M., Kronenberg, G., Walzlein, J.H., Markovic, D.S., Wang, L.P.,
Gast, D., Kiwit, J., Kempermann, G., Kettenmann, H., 2005. J. Neurosci. 25, 2637–
Guo, P., Imanishi, Y., Cackowski, F.C., Jarzynka, M.J., Tao, H.Q., Nishikawa, R., Hirose,
T., Hu, B., Cheng, S.Y., 2005. Am. J. Pathol. 166, 877–890.
Kremlev, S.G., Roberts, R.L., Palmer, C., 2004. J. Neuroimmunol. 149 (1–2), 1–9.
Kriz, J., Nguyen, M.D., Julien, J.P., 2002. Neurobiol. Dis. 10 (3), 268–278.
Lohi, J., Lehti, K., Valtanen, H., Parks, W.C., Keski-Oja, J., 2000. Gene 242, 75–86.
Machado, L.S., Kozak, A., Ergul, A., Hess, D.C., Borlongan, C.V., Fagan, S.C., 2006. BMC
Neurosci. 7, 56.
Markovic, D.S., Glass, R., Synowitz, M., Rooijen, N., Kettenmann, H., 2005. J.
Neuropathol. Exp. Neurol. 64, 754–762.
Markovic, D.S., Vinnakota, K., Chirasani, S., Synowitz, M., Raguet, H., Stock, K., Sliwa,
M., Lehmann, S., Kälin, R., van Rooijen, N., Holmbeck, K., Heppner, F.L., Kiwit, J.,
Matyash, V., Lehnardt, S., Kaminska, B., Glass, R., Kettenmann, H., 2009. Proc.
Natl. Acad. Sci. USA 106, 12530–12535.
Mayes, D.A., Hu, Y., Teng, Y., Siegel, E., Wu, X., Panda, K., Tan, F., Yung, W.K., Zhou,
Y.H., 2006. Cancer Res. 66, 9809–9817.
Nikodemova, M., Duncan, I.D., Watters, J.J., 2006. J. Neurochem. 96 (2), 314–323.
Osenkowski, P., Toth, M., Fridman, R., 2004. J. Cell Physiol. 200, 2–10.
Power, C., Henry, S., Del Bigio, M.R., Larsen, P.H., Corbett, D., Imai, Y., Yong, V.W.,
Peeling, J., 2003. Ann. Neurol. 53 (6), 731–742.
Rao, J.S., 2003. Nat. Rev. Cancer 3, 489–501.
Roggendorf, W., Strupp, S., Paulus, W., 1996. Acta Neuropathol. (Berlin) 92, 288–
Seabrook, T.J., Jiang, L., Maier, M., Lemere, C.A., 2006. Glia 53 (7), 776–782.
Seukeran, D.C., Eady, E.A., Cunliffe, W.J., 1997. Lancet 349 (9060), 1251–1252.
Strik, H.M., Stoll, M., Meyermann, R., 2004. Anticancer Res. 24, 37–42.
Suk, K., 2004. Neurosci. Lett. 366 (2), 167–171.
Wang, X., Zhu, S., Drozda, M., Zhang, W., Stavrovskaya, I.G., Cattaneo, E., Ferrante,
R.J., Kristal, B.S., Friedlander, R.M., 2003. Proc. Natl. Acad. Sci. USA 100 (18),
Watters, J.J., Schartner, J.M., Badie, B., 2005. J. Neurosci. Res. 81, 447–455.
Yong, V.W., Wells, J., Giuliani, F., Casha, S., Power, C., Metz, L.M., 2004. Lancet
Neurol. 3 (12), 744–751.
D.S. Markovic et al./Brain, Behavior, and Immunity 25 (2011) 624–628