Intracisternally Localized Bacterial DNA
Containing CpG Motifs Induces Meningitis
Guo-Min Deng, Zai-Qing Liu and Andrej Tarkowski
This information is current as
of June 15, 2017.
Subscription
Permissions
Email Alerts
This article cites 61 articles, 26 of which you can access for free at:
http://www.jimmunol.org/content/167/8/4616.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2001 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
References
J Immunol 2001; 167:4616-4626; ;
doi: 10.4049/jimmunol.167.8.4616
http://www.jimmunol.org/content/167/8/4616
Intracisternally Localized Bacterial DNA Containing CpG
Motifs Induces Meningitis1
Guo-Min Deng,2 Zai-Qing Liu, and Andrej Tarkowski
B
acterial infections can be localized to the subarachnoid
space, causing bacterial meningitis, a relatively common
and devastating disease despite adequate use of antibiotics. Fever, headache, meningismus, and signs of cerebral dysfunction are found in ⬃85% of patients who present with acute bacterial meningitis (1). Despite early and adequate antibiotic
treatment, bacterial meningitis remains an infection with a high
mortality rate, particularly for very young and elderly patients. In
addition, survivors may have permanent neurological damage, e.g.,
learning deficits, hearing loss, seizures, motor handicaps, and other
sequelae (2). To study the pathogenesis of bacterial meningitis,
animal models of this disease have been developed. It has been
shown that certain bacterial virulence factors as well as host immune responses are important in the induction phase and progression of this disease (3). It is believed that exacerbation of meningitis is due to increased levels of proinflammatory cytokines, partly
a consequence of antibiotic therapy leading to disruption of bacterial cell walls, resulting in the local release of biologically active
cell wall products such as LPS. The release of bacterial cell wall
fragments giving rise to the brain production of IL-1, IL-6, and
TNF-␣ will not only exacerbate inflammation, but also further
damage the blood-brain barrier (4 –7). However, in recent studies
of experimental Escherichia coli meningitis, the amount of bacterial endotoxin ultimately released as a consequence of bacterioly-
Department of Rheumatology, Göteborg University, Göteborg, Sweden
Received for publication May 10, 2001. Accepted for publication July 17, 2001.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by grants from the Gothenburg Medical Society, the
Swedish Association against Rheumatism, King Gustaf V’s Foundation, the Nanna
Svartz Foundation, the Swedish Medical Research Council, the University of Gothenburg, the A.-G. Crafoord Foundation, Börje Dahlin Foundation, the Inflammation
Network, the Infection and Vaccination Network, and the AME Wolff Foundation.
2
Address correspondence and reprint requests to Dr. Guo-Min Deng, Department of
Rheumatology, Gothenburg University, Guldhedsgatan 10A, S-41346 Gothenburg,
Sweden. E-mail address: [email protected]
Copyright © 2001 by The American Association of Immunologists
sis was much lower than that released by bacteria not being exposed to antibiotics (8). This finding led us to hypothesize whether
there might be another virulence factor for exacerbating inflammation as a consequence of antibiotic treatment. Recently, bacterial DNA was reported to have an immunostimulatory effect on
leukocytes (9 –12). Indeed, intraarticularly deposited bacterial
DNA induces arthritis (13, 14). In the case of human bacterial
meningitis, the PCR has shown the occurrence of bacterial DNA in
cerebrospinal fluid (CSF)3 (15).
We propose that bacterial DNA triggers macrophage accumulation that in turn contributes to meningeal inflammation and
blood-brain barrier permeability. In addition, we suggest that adhesion molecules responsible for macrophage recruitment, and
macrophage-derived inflammatory cytokines/chemokines are
likely to contribute to the pathogenesis of the condition.
Materials and Methods
Animals and reagents
BALB/c, C57BL/6, and NMRI mice were purchased from ALAB (Stockholm, Sweden). C3H/HeJ, C3H/HeN mice, SCID mice, and their congeneic strain C.B.17, Sprague Dawley (SD) rat were purchased from M&B
(Bomholtvej, Denmark). IL-12 p40 knockout mice were kindly provided
by J. Magram (Nutley, NJ). All mice were housed in the animal facility of
the Department of Rheumatology, Göteborg University (Göteborg, Sweden). Male mice and rats of 6 – 8 wk of age were used in all the experiments. The hybridoma cells secreting RB6-8C5 were a gift from R. Coffman (DAX Research Institute, Palo Alto, CA). PK136 Ab was obtained
from Sjögren-Jansson (Göteborg, Sweden). Etoposide was supplied by
Bristol-Myers Squibb (Bromma, Sweden). ␣-Melanocyte-stimulating hormone (␣-MSH), fucoidin, and cobra venom factor (CVF) were bought
from Sigma. NG-Monomethyl-L-arginine monoacetate salt (L-NMMA) and
N␻-nitro-L-arginine methyl ester (L-NAME) were purchased from Kelab
(Göteborg, Sweden). CR1Abs (8C12) were kindly supplied by T. Kinoshita
3
Abbreviations used in this paper: CSF, cerebrospinal fluid; ␣-MSH, ␣-melanocytestimulating hormone; CR, complement receptor; CVF, cobra venom factor; NOS, NO
synthase; iNOS, inducible NOS; L-NAME, N␻-nitro-L-arginine methyl ester;
G
L-NMMA, N -monomethyl-L-arginine monoacetate salt; MCP, monocyte chemoattractant protein; ODN, oligodeoxynucleotide; PGN, peptidoglycan.
0022-1767/01/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Unmethylated CpG motifs are frequently found in bacterial DNA, and have recently been shown to exert immunostimulatory
effects on leukocytes. Since bacterial infections in the CNS will lead to local release of prokaryotic DNA, we wanted to investigate
whether such an event might trigger meningitis. To that end, we have intracisternally injected mice and rats with bacterial DNA
and oligonucleotides containing CpG motifs. Histopathological signs of meningitis were evident within 12 h and lasted for at least
14 days, and were characterized by an influx of monocytic, Mac-3ⴙ cells and by a lack of T lymphocytes. To study the mechanisms
whereby unmethylated CpG DNA gives rise to meningitis, we deleted the monocyte/macrophage population leading to abrogation
of brain inflammation. Also, interaction with NF-␬B using antisense technology led to down-regulation of proinflammatory
cytokine production and frequency of meningitis. Furthermore, specific interactions with vascular selectin expression and inhibition of NO synthase led to a significant amelioration of meningitis, altogether indicating that this condition is dependent on
macrophages and their products. In contrast, neutrophils, NK cells, T/B lymphocytes, IL-12, and complement system were not
instrumental in meningitis triggered by bacterial DNA containing CpG motifs. This study proves that bacterial DNA containing
unmethylated CpG motifs induces meningitis, and indicates that this condition is mediated in vivo by activated macrophages. The
Journal of Immunology, 2001, 167: 4616 – 4626.
The Journal of Immunology
4617
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 1. The histopathological appearance of bacterial DNA-induced meningitis. A, Histopathology of brain of a C57BL/6 mouse killed 3 days after
intracisternal inoculation of 6 ␮g ODN 1668, showing meningeal and submeningeal deposits of leukocytes. B–D, Histopathological appearance of a SD
rat killed 3 days after intracisternal inoculation of 6 ␮g ODN 1668 (B) or 60 ␮g ODN 1668 (C, D). Infiltration of mononuclear cells in meninges and around
blood vessels is apparent. E–G, Normal histopathological appearance of a mouse subarachnoid space after intracisternal inoculation of calf thymus DNA
(30 ␮g), and a SD rat subarachnoid space after intracisternal inoculation of 6 ␮g ODN 1668 m (F) and 6 ␮g ODN 1720 (G). Magnification ⫻20.
4618
BACTERIAL DNA INDUCES MENINGITIS
(Uppsala, Sweden). Peptidoglycan (PGN) from Staphylococcus aureus was
kindly provided by S. Foster (Sheffield, U.K.).
Genomic DNA and oligonucleotides
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
E. coli (strain B) DNA and calf thymus DNA were purchased from Sigma
(Stockholm, Sweden) and further purified by extraction with phenol-chloroform-isoamyl alcohol (25:24:1), followed by ethanol precipitation.
Genomic DNA from S. aureus strain LS-1 was prepared by lysing the
bacteria with Qiagen lysis buffer (Qiagen, Hilden, Germany) and a combination of lysostaphin (Sigma). The DNA was purified using Qiagen
genomic tips. DNA purity and concentration were determined using a spectrophotometer (Molecular Devices, Sunnyvale, CA) at 260 and 280 nm
wavelengths. DNA purity, that is, the relation between L260 and L280, was
1.83, which is well within expected 1.7 and 2. Phosphorothioate-modified
oligonucleotides 1668, 1668 m, 1720, 2006, 2006 m, 2041, and nonphosphorothioate-modified oligonucleotides 1668 were synthesized by Scandinavian Gene Synthesis (Köping, Sweden). Sequences of 1668 (containing
the CpG motif) and its methylated derivative 1668 m as well as 1720
(inverted CG motif) were as previously described (13). 1668, 5⬘-TCC ATG
A CG TTC CTG ATG CT-3⬘; 1668 m, 5⬘-TCC ATG A XG TTC CTG ATG
CT-3⬘ X ⫽ 5-methyl-deoxycytidine; 1720, 5⬘-TCC ATG A GC TTC CTG
ATG CT-3⬘; 2006, 5⬘-TCGTCGTTTTGTCGTTTTGTCGTT-3⬘; 2006 m,
5⬘-TXGTXGTTTGTXGTTTTGTXGTT-3⬘; 2041, 5⬘-CTGGTCTTTCTGGT
TTTTTTCTGG-3⬘.
Assessment of endotoxin levels
Endotoxin concentration in each stock of the E. coli, S. aureus, calf thymus
genomic DNA, and oligonucleotide was assayed using the endpoint chromogenic Limulus amebocyte lysate assay (BioWhittaker, Walkersville,
MD). Preparation of S. aureus DNA, calf thymus DNA, and oligodeoxynucleotides (ODN) had levels of LPS ⬍0.025 EU/ml. E. coli DNA
preparation had levels of LPS ⬍1 EU/ml.
Injection protocol
Mice and rat were anesthetized with mixture of hypnorm, dormicum, and
distilled water (1:1:2), and placed on clean table. After disinfection of
injecting area, 10 ␮l of either DNA, ODN, LPS, PGN, or PBS was injected
intracisternally into mice and rats.
Macrophage depletion was induced by s.c. injection of 12.5 mg/kg body
weight of etoposide in a volume of 100 ␮l, into the nuchal region on 3
consecutive days before and 2 consecutive days after injection of CpG
ODN; control mice received the same volume of the vehicle diluted in PBS
(16). Analysis by FACS showed that the etoposide depletes the monocyte/
macrophage population by ⬎90%. In addition, for direct depletion of macrophage in meninges and brain, the simultaneous intracisternal inoculation
of etoposide (1.25 mg/kg) and CpG ODN (6 ␮g) into CSF was performed.
Neutrophil depletion. BALB/c mice were injected i.p. with 1 mg mAb
RB6-8C5, or the IgG rat anti-OVA mAb as a control, 2 h before injection
of ODN 1668. Analysis of blood smears showed that the mAb RB6-8C5
depletes the granulocyte population by ⬎90% (17) .
NK cell depletion was performed in C57BL/6 mice 1 day before injection of CpG ODN and repeated 2 days after injection of CpG ODN.
C57BL/6 mice were administered with i.p. injection of either 200 ␮g PK
136 mAb or the IgG O1C5.B2 as a control. Flow cytometry analysis revealed that the population of NK 1.1⫹ spleen cells from mice injected with
a single dose of PK 136 mAb was reduced from 5.5% to 0.6% (18).
Selectin blockade. Fucoidin (10 mg/kg body weight) was administered
i.v. or s.c. 10 min before intracisternal injection of CpG ODN, and 12 h
later. Control mice received the same volume of PBS in the same way (19).
For evaluation of the role of TNF-␣ in CpG ODN-triggered meningitis,
␣-MSH (10 or 20 ␮g) plus CpG ODN (6 ␮g) were intracisternally injected
into CSF. CpG ODN (6 ␮g) alone was used as a control. In addition,
␣-MSH (50 ␮g) alone was also injected i.p.
Complement depletion. CVF was injected i.p. at the dosage of 0.8 ␮g/g
body weight every 48 h. Interaction with complement receptor 1 (CR1;
CD35) was mediated by i.p. injection with either 400 ␮g IgG rat antimouse CR1 Abs (8C12), or with control, 400 ␮g IgG rat anti-OVA, 24 h
before intracisternal injection of CpG ODN (20).
Inhibition of NO synthase (NOS). L-NMMA or L-NAME was injected
i.p. at the dosage of 0.4 mg/g, 0.1 mg/g body weight 2 h before intracisternal injection of CpG ODN or PBS (21).
Administration of antisense of NF-␬B. Mice received an intracisternal
injection of CpG ODN (6 ␮g) ⫹ antisense of NF-␬B (6 ␮g) or i.p. administration of antisense (120 ␮g). Control mice received only an intracisternal injection with ODN. The sequences of phosphorothioate oligonucleotides were as previously described (22): murine p65 antisense 1, 5⬘-
FIGURE 2. Effect of bacterial DNA and ODN 1668 on development of
meningitis. A, Incidence of meningitis in C3H/HeJ mice (left, HeJ; n ⫽ 5)
or C3H/HeN mice (right, HeN; n ⫽ 5), killed 3 days after inoculation with
either E. coli DNA (30 ␮g), ODN 1668 (6 ␮g) or LPS (10 ␮g). B, Incidence of meningitis in C57BL/6 mice after inoculation of genomic E. coli
DNA (30 ␮g), S. aureus DNA (30 ␮g), calf thymus DNA (30 ␮g), ODN
1668, ODN 1668 m, ODN 1720 (all 6 ␮g), S. aureus DNA ⫹ DNase I, E.
coli DNA ⫹ DNase I (n ⫽ 5 per group), or PBS. All mice were killed 3
days after the inoculation. C, Kinetics of ODN 1668-induced meningitis in
C57BL/6 mice (n ⫽ 5 per group). Data represent the mean ⫾ SD score of
inflammatory cell infiltrate (severity).
The Journal of Immunology
4619
GAAACAGATCGTCCATGGT-3⬘; murine p65 antisense 2, 5⬘-GAGGGG
AAACAGATCGTCCATGGT-3⬘; murine p65 sense, 5⬘-ACCATGGACGAT
CTGTTTC-3⬘; murine p65 nonsense, 5⬘-GTACTACTCTGAGCAAGG A-3⬘.
All of these phosphorothioate oligonucleotides were synthesized by Scandinavian Gene Synthesis.
Assessment of blood-brain barrier integrity and leukocyte
numbers in CSF
To assess the integrity of the blood-brain barrier in animals exposed to
CpG ODN, Evans blue was used. Meningitis was induced in SD rats as
described above, but 1 h before termination of the experiment, 1 ml 1%
Evans blue (Sigma) was injected i.v. (23). CSF was collected from SD rats
injected with CpG ODN (12 ␮g) or PBS. Evans blue concentration in the
CSF was determined by measuring the absorbance at 650 nm with a spectrophotometer (Molecular Devices); serially diluted Evans blue in PBS
served as a standard. In other rats, the CSF was collected following intracisternal injection with CpG ODN (12 ␮g) or PBS, and centrifuged
(1500 ⫻ g, 10 min). Total and differential leukocyte counts of CSF were
done immediately after the CSF collection.
Histopathological and immunohistochemical examination of
brain
In situ hybridization for cytokine mRNA detection
In situ hybridization was conducted to detect mRNA expression of TNF-␣,
IL-1␤, IL-12, as well as monocyte chemoattractant protein-1 (MCP-1), as
previously detailed (14). Briefly, synthetic oligonucleotide probes for
TNF-␣, IL-1␤, IL-12 (gift from T. Olsson, Karolinska Institute, Stockholm, Sweden), and MCP-1 were labeled at the 3⬘ end using terminal
deoxynucleotidyl transferase (Advanced Biotechnologies, Leatherhead,
U.K.) and [␣-35S]ATP (DuPont Scandinavia, Stockholm, Sweden). Sections of 4-␮m-thick freshly frozen brain were thaw mounted onto slides
and were hybridized with 1 ⫻ 106 cpm of labeled probe/100 ␮l hybridization mixture. After emulsion autoradiography, development, and fixation, the coded slides were examined by dark field microscopy for positive
cells containing ⬎15 silver grains/cell in a starlike distribution.
Assessment of TNF-␣ levels in culture supernatants, CSF
Spleens from C57BL/6 mice were obtained aseptically and passed through
a nylon mesh. Erythrocytes were depleted by hypotonic lysis. The resulting
single cell suspension was resuspended in Iscove’s complete medium (10%
FCS, 5 ⫻ 10⫺5 M 2-ME, 2 mM L-glutamine, and 50 ␮g/ml gentamicin).
Subsequently, 1 ⫻ 106 cells/ml were incubated with 1 ␮M ODN 1668,
Results
Bacterial DNA and synthetic ODN containing unmethylated
CpG dinucleotides (CpG ODN) induce meningitis
To study the impact of bacterial DNA in mediating brain inflammation, we injected intracisternally DNA and CpG ODN (ODN
1668) into the cisterna magna of mice and rats, and found that
meningitis was induced by bacterial DNA and CpG ODN (Fig. 1,
A–D), but not by vertebrate DNA or PBS (Fig. 1E). In addition, we
used another ODN, ODN 2006, containing unmethylated CpG sequences. In this case, 9 of 10 mice injected intracisternally with
ODN 2006 developed meningitis. To exclude possibility of LPS
contamination in DNA preparations, we injected intracisternally
both types of bacterial DNA and CpG ODN to LPS nonresponder
C3H/HeJ mice. There was no difference in the incidence or score
of inflammatory cell infiltrate induced after the injection of bacterial DNA and CpG ODN in LPS-resistant strain C3H/HeJ mice
compared with congeneic LPS-responder strain C3H/HeN mice. In
contrast, intracisternal injection of LPS induced meningitis in
C3H/HeN, but not in C3H/HeJ mice (Fig. 2A). To further rule out
other bacterial contaminants in our DNA, the purified DNA was
digested with DNase I (Sigma). Intact bacterial DNA induced
meningitis, while DNase-exposed bacterial DNA did not (Fig. 2B).
These data indicate that induction of meningitis was due to bacterial DNA and CpG ODN rather than to LPS or other contaminants. We used CpG ODN (ODN 1668) as representative of bacterial DNA in all the following experiments unless notified.
To further evaluate meningitis triggered by bacterial DNA, we
assessed potential changes of blood-brain permeability by measuring influx of systemically injected Evans blue to CSF and total and
differential leukocyte counts in CSF in SD rats exposed to CpG
ODN. Rats rather than mice were used in this experiment since
CSF from the former species can be more easily collected. Table
I shows that blood-brain permeability was markedly increased in
CSF from rats injected with CpG ODN compared with control rats.
In addition, CpG ODN-triggered meningitis was characterized by
an increase of mononuclear cells in CSF (Table I).
A requirement for the induction of meningitis is that the bacterial DNA should be deposited locally in the cisterna rather than
Table I. Mean CSF leukocyte counts and CSF Evans blue concentration and score of inflammatory cell infiltrate in SD rats
injected intracisternally with ODN 1668 (12 ␮g) or PBSa
CSF index
2
Leukocytes (total) (10 /ml)
Polymorphonuclear
Mononuclear cells
Score of inflammatory cell
infiltrate in brain tissue
Evans blue (␮g/ml)
ODN 1668
PBS
ODN 2006
ODN 2041
603 ⫾ 292*
4⫾3
599 ⫾ 293*
2.25 ⫾ 0.5*
3⫾5
0
3⫾5
0⫾0
NA
NA
NA
1.4 ⫾ 0.7
NA
NA
NA
0⫾0
7.3 ⫾ 2.5*
0.08 ⫾ 0.04
NA
NA
CSF was collected from SD rats 3 days after injection with ODN 1668. n ⫽ 4 per group for CSF leukocyte counts. Evans blue applied i.v. was
quantified in CSF by spectrophotometry at 650 nm and used as a measure of blood-brain integrity. Evans blue experimental group (n ⫽ 3) and control
group (n ⫽ 2). C57BL/6 mice injected intracisternally with ODN 2006 (6 ␮g/mice, n ⫽ 10) and with ODN 2041 (6 ␮g/mice, n ⫽ 5). ⴱ, p ⬍ 0.05 vs PBS
controls. The score of inflammation in the subarachnoid space was scored by a predetermined scheme: 0 ⫽ no inflammation; 1 ⫽ occasional occurrence
of inflammatory cells; 2 ⫽ inflammatory cells forming an infiltrate not involving the entire depth of the subarachnoid space; 3 ⫽ inflammatory infiltrate
involving the entire subarachnoid space. NA, not analyzed.
a
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Histopathological examination of brain was performed after routine fixation and paraffin embedding. Tissue sections from brain were cut to 5 ␮m
thick and stained with H&E. All the slides were coded and evaluated
blindly.
For immunohistochemical examination, the brain was removed,
mounted on cryostat chucks, frozen in isopentane prechilled in liquid nitrogen, and kept at ⫺70° until cryosectioned. Serial cryosections, 6 ␮m
thick, were stained with rat mAbs directed against mouse CD11b (Mac-1),
CD4⫹ (GK1.5), CD8⫹ (53.6.7), and Mac-3, followed by incubation with
biotinylated secondary Abs and avidin-biotin-peroxidase complexes and
3-amino-9-ethyl-carbazole containing H2O2. All sections were counterstained with Mayer’s hematoxylin.
1668 (1 ␮M) ⫹ ␣-MSH (0.2 mg/ml), 1668 (1 ␮M) ⫹ antisense p65 (1
␮M), 1668 (1 ␮M) ⫹ sense p65 (1 ␮M), 1668 (1 ␮M) ⫹ nonsense (1 ␮M),
respectively. The cultures were kept in 24-well plates (Nunc A.S., Roskilde, Denmark) at 37°C in 5% CO2 and 95% humidity. The supernatants
were collected after 18 h for the detection of TNF-␣.
TNF-␣ levels of supernatants and CSF were determined using a TNF-␣
ELISA kit from R&D Systems (Minneapolis, MN). The assay was performed as recommended by the manufacturer.
4620
administered systemically. Indeed, i.p. injection of mice with either 10 nmol (i.e., 60 ␮g) or 20 nmol (i.e., 120 ␮g) CpG ODN did
not produce histological evidence of meningitis. In contrast, the
CpG ODN given intracisternally at a dose of 1 or 10 nmol (6 or 60
␮g) gave rise to meningitis in close to 100% of mice (data not
shown).
We found that 1 nmol (6 ␮g) CpG ODN was the lowest dose for
triggering meningitis (data not shown). The score of inflammatory
cell infiltrate was dose dependent (Fig. 1, B–D). Bacterial DNAand CpG ODN-triggered meningitis was inducible in six mouse
BACTERIAL DNA INDUCES MENINGITIS
strains, including NMRI, BALB/c, C57BL/6, C3H/HeN, C3H/
HeJ, and CB17, as well as SD rats (data not shown). Phosphorothioate-modified CpG ODN induced meningitis 3 days after inoculation, and did so at a magnitude higher than that of
phosphodiester CpG ODN (data not shown).
Meningitis as analyzed by histopathology appeared within 12 h
and lasted for at least 14 days. The maximal incidence and score
of inflammatory cell infiltrate were noted on day 3 after injection
of CpG ODN (Fig. 2C). There was hypertrophy of the leptomeninges, and infiltrating leukocytes were found in the meningeal
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 3. Cellular requirement for induction of CpG ODN-triggered meningitis. A–D, Immunohistochemical analysis of brain, showing meningeal
infiltration of cells expressing Mac-1 (brown)-positive cells (A), absence of CD4 (B), CD8 (C) staining, and high frequency of Mac-3⫹ cells (D) in meninges
and around blood vessels of a mouse killed 3 days after intracisternal inoculation with ODN 1668 (6 ␮g). E and F, Incidence and score of inflammatory
cell infiltrate (severity) in BALB/c mice depleted of neutrophils using mAb RB6-8C5 (n ⫽ 10 per group), C57BL/6 mice depleted of NK cells using mAb
PK136 (n ⫽ 10 per group), SCID mice deficient in T and B cells and congeneic CB17 control mice (n ⫽ 11 per group), and C57BL/6 mice depleted of
macrophages using etoposide treatment by s.c. injection or intracisternal injection (i.c.) (n ⫽ 10 per group). All values are expressed as mean ⫾ SD. All
the mice were killed 3 days after intracisternal inoculation of 6 ␮g ODN 1668.
The Journal of Immunology
4621
The role of different immunocompetent cells in meningitis
induced by CpG ODN
The role of cytokines, chemokines, and complement in
meningitis induced by CpG ODN
To study the role of different immune cells in meningitis induced
by CpG ODN, we first checked which immune cells participated in
meningitis using immunohistochemistry. Immunohistochemical
analysis of mouse brain injected with CpG ODN demonstrated that
there was abundance of Mac-1⫹ and Mac-3⫹ mononuclear cells at
all stages of meningitis, but lack of T lymphocytes (Fig. 3, A–D).
These data suggest that cells of monocyte/macrophage lineage
may exert a role in development of this condition.
We treated mice i.p. with a cytotoxic drug (etoposide) known to
selectively deplete monocyte/macrophage population to assess the
role of this cell lineage in CpG ODN-triggered meningitis (25).
Depletion of monocytes almost completely abrogated CpG ODNtriggered meningitis (Fig. 3, E and F). Also, simultaneous inoculation of CpG ODN and etoposide into cerebral ventricles inhibited
development of meningitis (Fig. 3, E and F). These results strongly
indicate that the monocyte/macrophage lineage is responsible for
the induction of CpG ODN-triggered meningitis.
To decide the role of other immune cells such as neutrophils,
NK cell, and T and B cells, we depleted neutrophil and NK cells
in the mice and also used SCID mice. As shown in Fig. 3, E and
F, the incidence and score of inflammatory cell infiltrate were not
significantly different between neutrophil or NK cell-depleted mice
or SCID mice and their control mice. They reflect that neutrophil,
NK cells, and T and B cells do not affect development of meningitis induced by CpG ODN. Taken together, these studies provided
strong evidence for the role of macrophages in initiation and development of meningitis triggered by CpG ODN.
To analyze how macrophages trigger meningitis, we studied effect
of products released from macrophages that might be important in
induction of meningitis. For this purpose, we first measured
mRNA expression for cytokines in mouse brain using in situ hybridization. Fig. 4, A and B, shows that mRNA expression for the
cytokines TNF-␣, IL-1␤, IL-12, and MCP-1 was increased in
brains exposed to CpG ODN compared with brains exposed to calf
thymus DNA or PBS.
As TNF-␣, a cytokine mainly produced by activated monocytes/
macrophages, is one of major mediators of bacterial meningitis (7,
28, 29), and bacterial DNA and CpG ODN is known to stimulate
macrophages to release TNF-␣ (12), we assessed its impact on the
course of meningitis. We collected CSF samples from SD rats
injected with CpG ODN and analyzed TNF-␣ in CSF by ELISA.
TNF-␣ levels in CSF were clearly elevated in rats injected with
CpG ODN compared with uninjected controls (Fig. 4C). In contrast, only background levels of TNF-␣ were found in CSF from
control rats inoculated with calf thymus DNA or with PBS (data
not shown). In addition, intracisternal administration of TNF-␣
Which property of bacterial DNA is responsible for induction of
meningitis induced by bacterial DNA?
The role of NF-␬B in meningitis induced by CpG ODN
What factor controls and regulates the activity of macrophages in
CpG ODN-triggered meningitis? To answer this question, we investigated the role of NF-␬B, a major intracellular factor controlling and regulating gene expression of proinflammatory cytokines
at the transcriptional level (26, 27). Importantly, previous studies
have shown that bacterial DNA and CpG ODN induce NF-␬B
activation in macrophages (12). For this purpose, we locally ad-
Table II. Impact of local p65 antisense treatment on meningitis
induced by CpG ODNa
Oligonucleotide
ODN 1668
1668⫹ antisense 1
1668⫹ antisense 2
1668⫹ sense
1668⫹ nonsense
Antisense 1 alone
Antisense 2 alone
p65 sense alone
Nonsense alone
No. of
Mice
Incidence of
Meningitis (%)
Score of Inflammatory
Cell Infiltrate
4
9
8
3
3
4
4
3
3
100
22
50
100
100
0
0
33
0
2.8 ⫾ 0.5
0.2 ⫾ 0.4
0.6 ⫾ 0.7
2.3 ⫾ 0.6
2.3 ⫾ 0.6
0.0 ⫾ 0.0
0.0 ⫾ 0.0
0.7 ⫾ 1.2
0.0 ⫾ 0.0
a
Incidence and score of inflammatory cell infiltrate in C57BL/6 mice treated with
intracisternal injection of p65 antisense of NK-␬B, sense, nonsense (all 6 ␮g/mouse).
After 3 days, the mice were killed, and brains were examined by histopathology.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
To answer this question, one needs to consider the structure of
bacterial DNA and eukaryotic DNA. Unmethylated CpG motifs
are present at the expected frequency of one per sixteen dinucleotides in the bacterial DNA, but are underrepresented (CpG suppression) and predominantly methylated in eukaryotic DNA (24).
This structural difference seems to explain the finding that bacterial
DNA and certain synthetic oligonucleotides containing unmethylated CpG dinucleotides activate immune cells (9, 10, 12). To determine whether the methylation of CpG dinucleotides protects
against the induction of meningitis, we used oligonucleotides containing either an embedded unmethylated CpG dinucleotide
(1668), a methylated CpG dinucleotide of the same sequence
(1668 m), or a CpG inverted to nonmethylated GpC dinucleotide
(1720) in mice and SD rat. Only ODN 1668 was able to cause
meningitis (Figs. 1, B, F, and G, and 2B). In agreement, only in
minority of cases (one of four mice) methylated ODN 2006 triggered meningitis. Thus, oligonucleotides containing unmethylated
CpG dinucleotides are responsible for the induction of meningitis.
ministered antisense phosphorothioate oligonucleotides targeted
against the translation initiation site of the p65 subunit of NF-␬B,
which has previously been shown to abrogate experimental colitis
in mice (22). Phosphorothioate oligonucleotides were administered
to mice either as a single i.p. injection or as a single intracisternal
injection. We found that a single i.p. injection of p65 antisense
oligonucleotide (60 or 120 ␮g) did not decrease the incidence and
score of the inflammatory cell infiltrate (data not shown). In contrast, a single intracisternal inoculation of the p65 antisense oligonucleotide (6 ␮g) significantly decreased the incidence and score
of the inflammatory cell infiltrate of CpG ODN-triggered meningitis (Table II). As expected, significant changes of incidence and
score of inflammatory cell infiltrate were not observed in mice
treated with control oligonucleotides (6 ␮g p65 sense and nonsense). However, there was an obvious toxicity resulting in an
increased mortality when the dose of p65 antisense was higher
than 6 ␮g (data not shown). Notably, mRNA expression of TNF-␣
and IL-1␤ was significantly reduced in brains of mice treated with
p65 antisense (Fig. 4G). Also, in vitro antisense-mediated downregulation of p65 expression in CpG ODN-stimulated mononuclear cells was accompanied by significantly reduced secretion of
TNF-␣. This effect was absent using various control oligonucleotides (data not shown). Taken together, the above evidence implies
that NF-␬B p65 is a major transcriptional regulator of expression
of CpG ODN-triggered meningitis.
lining cell layer as well as around blood vessels. This meningeal
infiltration was apparent around the basal cisterns, on the cortical
surface, and in the longitudinal fissure of the cerebral hemispheres,
with higher density around blood vessels (Fig. 1A–D).
4622
BACTERIAL DNA INDUCES MENINGITIS
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 4. The role of TNF-␣ in development of meningitis. A, Cytokine and chemokine mRNA expression in C57BL/6 mice 3 days after intracisternal
inoculation with 6 ␮g CpG ODN per mouse. Values are the mean ⫾ SEM of number of positive cells/cm2 (n ⫽ 5 mice per group). CT-DNA, calf thymus
DNA. B, In situ hybridization of inflamed meningeal tissue showing TNF-␣ mRNA expression 3 days after inoculation of ODN 1668. Magnification ⫻40.
C, Since CSF from rats can be more reproducibly harvested than from mice, rats were used for collection of CSF and measured for TNF-␣ levels. CSF
TNF-␣ levels in SD rat on days 0, 3, and 5 after intracisternal inoculation with ODN 1668 (12 ␮g) (n ⫽ 4 per group). D and E, Incidence and score of
inflammatory cell infiltrate (severity) in C57BL/6 mice, treated with ␣-MSH given intracisternally (10 or 20 ␮g) or i.p. (50 ␮g), 3 days after inoculation
with ODN 1668 (6 ␮g) (n ⫽ 8 per group). Values are the mean ⫾ SD score of inflammatory cell infiltrate score. F, ␣-MSH inhibits TNF-␣ release from
mononuclear cells stimulated by ODN 1668 (n ⫽ 4 per group). G, Effects of p65 antisense 1 treatment on brain cytokine mRNA expression of TNF-␣ and
The Journal of Immunology
4623
and B, incidence of meningitis and the score of inflammatory cell
infiltrate were only marginally affected in IL-12 knockout mice
compared with control congeneic mice.
Next, we analyzed the effect of complement because the complement anaphylatoxin C5a and C3a are powerful chemoattractants that recruit macrophages and polymorphonuclear cells into
the inflammatory site. In addition, monocytes/macrophages can
produce almost all of the complement components in a functionally active form, and bear CR1 and CR3 (31–33). To assess the
role of complement cascade in CpG ODN-triggered meningitis, the
complement system was depleted using CVF. In another set of
experiments, CR1 was blocked using rat anti-mouse CR1 Ab
(8C12). Results from these experiments revealed that the complement system does not participate in development of meningitis
because the incidence and score of inflammatory cell infiltrate
were not significantly different between depleted and control mice
(Fig. 5, A and B).
The role of NO in meningitis induced by CpG ODN
The role of selectins in meningitis induced by CpG ODN
FIGURE 5. The role of macrophage products in development of meningitis. A and B, Incidence and score of inflammatory cell infiltrate (severity) in IL-12 knockout mice (n ⫽ 10 per group), depleted of complement
using CVF treatment, neutralized CR1 using CR1-specific mAbs (n ⫽ 9
per group), inhibited iNOS using L-NMMA or L-NAME (n ⫽ 10 per
group). All mice were killed 3 days after inoculation with ODN 1668 (6
␮g). Values are the mean ⫾ SD score of inflammatory cell infiltrate.
(0.2 ng; Genzyme, Cambridge, MA) in mice triggered meningitis
(data not shown), further confirming role of this cytokine for the
induction of brain inflammation. Furthermore, we used ␣-MSH,
which can abrogate the effects of TNF-␣-mediated brain inflammation (30). The incidence and score of inflammatory cell infiltrate
of CpG ODN-triggered meningitis were significantly lower in
mice treated with ␣-MSH than in control mice (Fig. 4, D and E).
In addition, in vitro experiments displayed that the level of TNF-␣
was significantly decreased in supernatants from CpG ODN-stimulated mononuclear cells containing ␣-MSH compared with control supernatants (Fig. 4F). Collectively, these results indicate that
TNF-␣ is operational in CpG ODN-triggered meningitis.
To study the role of IL-12 in induction of meningitis, we used
IL-12 knockout mice to determine whether this cytokine is required for induction of meningitis. However, as shown in Fig. 5, A
How do monocytes migrate to the subarachnoid space from blood
vessels? To understand this process, we analyzed the role of selectins since these molecules mediate leukocyte rolling, and thus
control early steps of extravasation of leukocytes during inflammation (36). To this end, we used fucoidin, which has the ability
to block P- and L-selectins (19, 37, 38). Mice were treated with
fucoidin either i.v. or s.c. Histopathological results 24 h after inoculation of CpG ODN demonstrated that the score of inflammatory cell infiltrate and incidence of meningitis were markedly reduced in mice treated with fucoidin compared with control mice
(Fig. 6, A and B), indicating that selectins are instrumental in development of meningitis.
Synergy of bacterial DNA with LPS and PGN
Since LPS from Gram-negative bacteria and PGN from Grampositive bacteria and bacterial DNA stimulate macrophages by different signal pathways (9, 39, 40), and any of these compounds
may be present during bacterial meningitis, we wished to analyze
whether LPS or PGN together with CpG ODN act synergistically
in inducing meningitis. There were an intermediate incidence and
score of inflammatory cell infiltrate when suboptimal amounts of
either LPS or PGN were injected intracisternally. In contrast, a
combination of a suboptimal dose of CpG ODN together with
above doses of either LPS or PGN led to a high incidence and
score of inflammatory cell infiltrate (Table III).
IL-1␤ in C57BL/6 mice 3 days after intracisternal inoculation of 6 ␮g CpG ODN. Mice with CpG ODN-induced meningitis received intracisternal
inoculation with 6 ␮g p65 antisense 1. Values are the mean ⫾ SEM of number of positive cells/cm2 (n ⫽ 4 mice per group). ⴱ, p ⬍ 0.05 as compared
with mice inoculated with CpG ODN only.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Since NO plays a detrimental role in bacterial meningitis (34), we
decided to assess its role in CpG ODN-triggered meningitis. To
ascertain whether NO elicits neurotoxicity in this condition, we
administered inhibitors of NOS. TNF-␣ and IL-1␤ are potent inducers of inducible NOS (iNOS) (35). In our study, iNOS was
inhibited using L-NMMA or L-NAME, injected i.p. Fig. 5, A and
B, shows that the incidence and score of inflammatory cell infiltrate were markedly down-regulated in mice treated with
L-NMMA or L-NAME compared with control mice. These observations suggest a critical role of NO in development of CpG ODNtriggered meningitis.
4624
BACTERIAL DNA INDUCES MENINGITIS
Discussion
Our results indicate that intracisternally deposited bacterial DNA
in general and unmethylated CpG oligonucleotides in particular
trigger meningitis. This condition was not only assessed as inflammatory infiltrates in the meningeal tissue, but also proved to trigger
pleocytosis and increased permeability of blood-brain barrier.
Meningitis was induced within 12 h after intracisternal injection
and lasted for at least 14 days. The maximal incidence and score
of the inflammatory cell infiltrate were noted on day 3 after the
injection of bacterial DNA or CpG ODN. Unmethylated CpG
dinucleotides were responsible for induction of meningitis trig-
Table III. CpG ODN acts synergistically with LPS and PGN in the
induction of meningitisa
Treatment
No. of
Mice
Incidence of
Meningitis (%)
Score of
Inflammatory
Cell Infiltrate
None
LPS
PGN
1668
LPS ⫹ 1668
PGN ⫹ 1668
5
5
4
5
6
6
0
60
25
80
100
100
0.0 ⫾ 0.0
0.8 ⫾ 0.8
0.5 ⫾ 1.0
0.8 ⫾ 0.5
2.7 ⫾ 0.5
1.5 ⫾ 0.8
a
C57BL/6 mice were injected intracisternally with LPS (0.1 ␮g), PGN (1 ␮g),
ODN 1668 (0.6 ␮g), a combination of LPS (0.1 ␮g) and ODN 1668 (0.6 ␮g), or a
combination of PGN (1 ␮g) and ODN 1668 (0.6 ␮g). After 3 days, the mice were
killed, and brains were examined by histopathology.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 6. Impact of selectin expression on the development of meningitis. A and B, Incidence and score of inflammatory cell infiltrate (severity) in BALB/c mice following blockade of P- and L-selectin using i.v. or
s.c. fucoidin treatment (10 mg/kg). Fucoidin was given 5 min before injection of ODN 1668 (6 ␮g) and 12 h thereafter; the mice were killed 24 h
after inoculation of ODN 1668 (n ⫽ 8 per group). Values are the mean ⫾
SD score of inflammatory cell infiltrate.
gered by bacterial DNA. The magnitude of meningeal inflammation triggered by CpG ODN may be synergistically enhanced by
the inflammatory response to LPS and PGN. Meningitis following
injection of bacterial DNA was not due to surgical trauma or nonDNA bacterial contamination, since intracisternal inoculation of
PBS, calf thymus DNA, or ODN lacking unmethylated CpG motifs did not trigger meningitis. In addition, bacterial DNA and CpG
ODN triggered meningitis equally well both in LPS nonresponder
and in LPS-sensitive mice. Finally, destruction of bacterial DNA
using DNase I treatment abolished the induction of meningitis.
These studies suggest that unmethylated CpG containing bacterial
DNA may be a potent trigger of inflammatory cell infiltration in
meningitis.
Macrophages play an important role in the pathogenesis of experiment bacterial (41, 42) and human meningitis (43), such as
tuberculous meningitis. In children, tuberculous meningitis is associated with ⬃50% mortality; most of the survivors have permanent neurologic sequelae and experience considerable disability.
There are several lines of evidence that support the importance of
macrophages also in CpG ODN-mediated meningitis. First, histochemical analysis of brain tissue demonstrated that meningitis was
characterized by influx of monocytic, Mac-3⫹ cells. More importantly, our experiments demonstrated that absence of neutrophils,
NK cells, and T/B cells did not affect development of meningitis.
In contrast, depletion of monocytes totally abrogated development
of inflammation. Third, leukocyte differential count demonstrated
that mononuclear cells predominated in CSF. Fourth, the pattern of
mRNA expression of cytokines in the inflamed brain was indicative of macrophage products, such as the expression of TNF-␣. In
addition, previous study has demonstrated that phosphodiester and
phosphorothioate ODN predominantly bind to the surface of leukocytes expressing Mac-1⫹, and this binding inhibits migration of
polymorphonuclear cells (44).
What are the mechanisms of bacterial DNA-mediated meningitis? Previous studies showed that bacterial DNA and CpG ODN
directly activate macrophages (12). The first step of activation encompasses uptake of bacterial DNA or synthetic oligonucleotides
by macrophages in a saturable, sequence-independent, temperature- and energy-dependent manner (45, 46) into an acidified intracellular compartment, in which DNA becomes degraded to
ODN (47). Once there, unmethylated CpG dinucleotides activate
within minutes the stress-kinase/jun pathway, yielding NF-␬B
(48). This transcription factor controls mRNA expression of a variety of cytokines and secretion of proinflammatory cytokines,
such as TNF-␣, IL-1␤, IL-6, as well as MCP-1 (26, 49). TNF-␣
and IL-1␤ initiate meningeal inflammation (6, 7), elicit selectin
expression on the endothelium (50), and promote synthesis of NO
(35), as well as activate NF-␬B (51, 52).
TNF-␣ is mainly released from monocyte and macrophage, and
acts as a mediator of inflammation (53). It causes endothelial cell
activation, up-regulates expression of adhesion molecules; and
stimulates macrophage production of IL-1 and IL-6 (53). IL-1 and
TNF-␣ display synergistic actions and stimulate the release of each
other, thereby amplifying the cascade of other inflammatory mediators (54). A high level of TNF-␣ and IL-1 is found in the CSF
of patients suffering from bacterial meningitis and experimental
bacterial meningitis (55). Also, in case of CpG ODN-triggered
meningitis, the levels of intrathecal TNF-␣ were clearly elevated.
It has been shown that murine rTNF-␣ is able to induce meningitis
(7, 56). Indeed, neutralization of TNF-␣ can attenuate bacterial
meningitis (55). Inhibition of TNF-␣ release by ␣-MSH decreased
the incidence and score of inflammatory cell infiltrate triggered by
The Journal of Immunology
lated by p65 subunit of NF-␬B. These findings may have implications for the treatment of bacterial meningitis. Thus, eradication
of bacteria using antibiotics might not be an efficient enough procedure to clear the disease since released bacterial DNA from resident bacteria is proinflammatory, and in concert with LPS or PGN
might promote a continuous inflammatory response. Indeed, antiinflammatory treatment in conjunction with antibiotics has been
shown to reduce inflammation and its clinical sequelae in both
experimental and clinical studies (4, 61, 62). Future treatment
strategies should include attempts to minimize bacterial growth
together with blockade of the proinflammatory effects of
bacterial DNA.
References
1. Tunkel, A. R., and W. M. Scheld. 1995. Acute bacterial meningitis. Lancet 346:
1675.
2. Quagliarello, V., and W. M. Scheld. 1992. Bacterial meningitis: pathogenesis,
pathophysiology, and progress. N. Engl. J. Med. 327:864.
3. Tauber, M. G., U. Borschberg, and M. A. Sande. 1988. Influence of granulocytes
on brain edema, intracranial pressure, and cerebrospinal fluid concentrations of
lactate and protein in experimental meningitis. J. Infect. Dis. 157:456.
4. Mustafa, M. M., J. Mertsola, O. Ramilo, X. Saez-Llorens, R. C. Risser, and
G. H. McCracken, Jr. 1989. Increased endotoxin and interleukin-1␤ concentrations in cerebrospinal fluid of infants with coliform meningitis and ventriculitis
associated with intraventricular gentamicin therapy. J. Infect. Dis. 160:891.
5. Mustafa, M. M., O. Ramilo, J. Mertsola, R. C. Risser, B. Beutler, E. J. Hansen,
and G. H. McCracken, Jr. 1989. Modulation of inflammation and cachectin activity in relation to treatment of experimental Haemophilus influenzae type b
meningitis. J. Infect. Dis. 160:818.
6. Quagliarello, V. J., B. Wispelwey, W. J. Long, Jr., and W. M. Scheld. 1991.
Recombinant human interleukin-1 induces meningitis and blood-brain barrier injury in the rat: characterization and comparison with tumor necrosis factor.
J. Clin. Invest. 87:1360.
7. Ramilo, O., X. Saez-Llorens, J. Mertsola, H. Jafari, K. D. Olsen, E. J. Hansen,
M. Yoshinaga, S. Ohkawara, H. Nariuchi, and G. H. McCracken, Jr. 1990. Tumor
necrosis factor ␣/cachectin and interleukin 1␤ initiate meningeal inflammation.
J. Exp. Med. 172:497.
8. Friedland, I. R., H. Jafari, S. Ehrett, S. Rinderknecht, M. Paris, M. Coulthard,
H. Saxen, K. Olsen, and G. H. McCracken, Jr. 1993. Comparison of endotoxin
release by different antimicrobial agents and the effect on inflammation in experimental Escherichia coli meningitis. [Published erratum appears in 1993 J. Infect. Dis. 168:1342.] J. Infect. Dis. 168:657.
9. Krieg, A. M., A. K. Yi, S. Matson, T. J. Waldschmidt, G. A. Bishop, R. Teasdale,
G. A. Koretzky, and D. M. Klinman. 1995. CpG motifs in bacterial DNA trigger
direct B-cell activation. Nature 374:546.
10. Klinman, D. M., A. K. Yi, S. L. Beaucage, J. Conover, and A. M. Krieg. 1996.
CpG motifs present in bacteria DNA rapidly induce lymphocytes to secrete interleukin 6, interleukin 12, and interferon ␥. Proc. Natl. Acad. Sci. USA 93:2879.
11. Sparwasser, T., T. Miethke, G. Lipford, K. Borschert, H. Hacker, K. Heeg, and
H. Wagner. 1997. Bacterial DNA causes septic shock. Nature 386:336.
12. Sparwasser, T., T. Miethke, G. Lipford, A. Erdmann, H. Hacker, K. Heeg, and
H. Wagner. 1997. Macrophages sense pathogens via DNA motifs: induction of
tumor necrosis factor-␣-mediated shock. Eur. J. Immunol. 27:1671.
13. Deng, G. M., I. M. Nilsson, M. Verdrengh, L. V. Collins, and A. Tarkowski.
1999. Intra-articularly localized bacterial DNA containing CpG motifs induces
arthritis. Nat. Med. 5:702.
14. Deng, G. M., and A. Tarkowski. 2000. The features of arthritis induced by CpG
motifs in bacterial DNA. Arthritis Rheum. 43:356.
15. Ni, H., A. I. Knight, K. Cartwright, W. H. Palmer, and J. McFadden. 1992.
Polymerase chain reaction for diagnosis of meningococcal meningitis. Lancet
340:1432.
16. Verdrengh, M., and A. Tarkowski. 2000. Role of macrophages in Staphylococcus
aureus-induced arthritis and sepsis. Arthritis Rheum. 43:2276.
17. Verdrengh, M., and A. Tarkowski. 1997. Role of neutrophils in experimental
septicemia and septic arthritis induced by Staphylococcus aureus. Infect. Immun.
65:2517.
18. Nilsson, N., T. Bremell, A. Tarkowski, and H. Carlsten. 1999. Protective role of
NK1.1⫹ cells in experimental Staphylococcus aureus arthritis. Clin. Exp. Immunol. 117:63.
19. Granert, C., J. Raud, A. Waage, and L. Lindquist. 1999. Effects of polysaccharide
fucoidin on cerebrospinal fluid interleukin-1 and tumor necrosis factor ␣ in pneumococcal meningitis in the rabbit. Infect. Immun. 67:2071.
20. Sakiniene, E., B. Heyman, and A. Tarkowski. 1999. Interaction with complement
receptor 1 (CD35) leads to amelioration of sepsis-triggered mortality but aggravation of arthritis during Staphylococcus aureus infection. Scand. J. Immunol.
50:250.
21. Sakiniene, E., T. Bremell, and A. Tarkowski. 1997. Inhibition of nitric oxide
synthase (NOS) aggravates Staphylococcus aureus septicaemia and septic arthritis. Clin. Exp. Immunol. 110:370.
22. Neurath, M. F., S. Pettersson, K. H. Meyer zum Buschenfelde, and W. Strober.
1996. Local administration of antisense phosphorothioate oligonucleotides to the
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
CpG ODN. It has been demonstrated that TNF-␣ can initiate meningeal inflammation (7), resulting in breach of the blood-brain barrier leading to brain edema and increased intracranial pressure (2);
and it plays a critical role in neuronal apoptosis in the hippocampus (57).
Nitrite levels are usually significantly elevated in CSF of patients with bacterial meningitis and in experimental meningitis (34,
58). Inhibition of iNOS has been shown to attenuate alteration of
the blood-brain barrier permeability and meningeal inflammation
(58). As TNF-␣ is a crucial mediator in meningitis triggered by
CpG ODN, it was an expected finding that iNOS inhibition
blocked the development of CpG ODN-triggered meningitis.
Selectins are cell surface glycoproteins that mediate the initial
adhesion and rolling, a transient contact leading to a deceleration
of leukocytes along the vascular endothelium. This is the first and
essential step in the process of leukocyte extravasation into inflamed sites (31, 36). In bacterial meningitis, treatment with L- and
P-selectin-binding fucoidin (37, 38) has been shown to attenuate
meningeal inflammation, abolish pleocytosis, and inhibit the release of TNF-␣ and IL-1 in a rabbit model of meningitis induced
by pneumococcal cell wall fragments (19, 59). The fact that fucoidin did not completely abrogate meningitis induced by CpG
ODN indicates that E-selectin might also be required for development of the meningitis. Indeed, in TNF-␣-induced meningitis, Eselectin has been shown to participate in the development of meningitis since CSF leukocyte influx and permeability were partially
inhibited in P-selectin-deficient mice, and near complete inhibition
of these parameters was displayed in mice being double deficient
with respect to P- and E-selectin (56).
One of the major control mechanisms of gene expression occurs
at the transcriptional level, and all of these proinflammatory cytokines have been shown to be regulated by the NF-␬B via binding
of NF-␬B family members to their individual promoters (27, 60).
CpG ODN and bacterial DNA cause nuclear translocation of the
transcription factor NF-␬B, including p65 and p50 (12). Ordinarily, NF-␬B is sequestered in the cytoplasm by the inhibitory
protein I-␬B, and controls the promoter activity of gene of proinflammatory cytokines in macrophages (26, 27). TNF-␣ and IL-1␤
have the ability to mediate NF-␬B activity in cells of the bloodbrain barrier in response to localized inflammation (51, 52). Our
data that local administration of p65 antisense to NF-␬B abolished
meningitis triggered by CpG ODN show a predominant role of the
p65 subunit of NF-␬B in this condition. This antisense effect was
previously shown to be highly specific. The most likely mechanism by which p65 antisense oligonucleotides influence CpG
ODN-induced meningitis is by reduced local production of cytokines. This is supported by in vitro experiments demonstrating that
mononuclear cells exposed to CpG ODN produced less TNF-␣ if
the culture contained p65 antisense, and in vivo experiments demonstrating that mRNA expression of TNF-␣ and IL-1␤ was significantly reduced in brain of mice treated with p65 antisense. In
further support of this suggestion, it has been shown that local
administration of antisense to the p65 subunit of NF-␬B abrogated
established experimental colitis in mice (22). Overall, these data
suggest that NF-␬B plays a regulatory role in CpG ODN-triggered
inflammatory disorders such as meningitis, possibly providing an
attractive target for future therapeutic interventions in bacterial
meningitis.
Our observations provide the first evidence that bacterial DNA
containing unmethylated CpG motifs induces meningitis, and suggest an important pathogenic role for bacterial DNA in bacterial
meningitis. Macrophages and their products such as TNF-␣ exert
an important role in development of inflammatory disease triggered by bacterial DNA; this type of inflammation can be regu-
4625
4626
23.
24.
25.
26.
27.
28.
29.
30.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
in mycobacterial infection in the central nervous system. Proc. Natl. Acad. Sci.
USA 96:5657.
Jeren, T., and I. Beus. 1982. Characteristics of cerebrospinal fluid in tuberculous
meningitis. Acta Cytol. 26:678.
Benimetskaya, L., J. D. Loike, Z. Khaled, G. Loike, S. C. Silverstein, L. Cao,
J. el Khoury, T. Q. Cai, and C. A. Stein. 1997. Mac-1 (CD11b/CD18) is an
oligodeoxynucleotide-binding protein. Nat. Med. 3:414.
Geselowitz, D. A., and L. M. Neckers. 1992. Analysis of oligonucleotide binding,
internalization, and intracellular trafficking utilizing a novel radiolabeled
crosslinker. Antisense Res. Dev. 2:17.
Zhao, Q., T. Waldschmidt, E. Fisher, C. J. Herrera, and A. M. Krieg. 1994.
Stage-specific oligonucleotide uptake in murine bone marrow B-cell precursors.
Blood 84:3660.
Bennett, R. M., G. T. Gabor, and M. M. Merritt. 1985. DNA binding to human
leukocytes: evidence for a receptor-mediated association, internalization, and
degradation of DNA. J. Clin. Invest. 76:2182.
Hacker, H., H. Mischak, T. Miethke, S. Liptay, R. Schmid, T. Sparwasser,
K. Heeg, G. B. Lipford, and H. Wagner. 1998. CpG-DNA-specific activation of
antigen-presenting cells requires stress kinase activity and is preceded by nonspecific endocytosis and endosomal maturation. EMBO J. 17:6230.
Ueda, A., Y. Ishigatsubo, T. Okubo, and T. Yoshimura. 1997. Transcriptional
regulation of the human monocyte chemoattractant protein-1 gene: cooperation of
two NF-␬B sites and NF-␬B/Rel subunit specificity. J. Biol. Chem. 272:31092.
Harari, O. A., J. F. McHale, D. Marshall, S. Ahmed, D. Brown, P. W. Askenase,
and D. O. Haskard. 1999. Endothelial cell E- and P-selectin up-regulation in
murine contact sensitivity is prolonged by distinct mechanisms occurring in sequence. J. Immunol. 163:6860.
Laflamme, N., and S. Rivest. 1999. Effects of systemic immunogenic insults and
circulating proinflammatory cytokines on the transcription of the inhibitory factor
␬B␣ within specific cellular populations of the rat brain. J. Neurochem. 73:309.
Laflamme, N., S. Lacroix, and S. Rivest. 1999. An essential role of interleukin-1␤
in mediating NF-␬B activity and COX-2 transcription in cells of the blood-brain
barrier in response to a systemic and localized inflammation but not during endotoxemia. J. Neurosci. 19:10923.
Tracey, K. J., and A. Cerami. 1994. Tumor necrosis factor: a pleiotropic cytokine
and therapeutic target. Annu. Rev. Med. 45:491.
Dinarello, C. A. 1991. Interleukin-1 and interleukin-1 antagonism. Blood 77:
1627.
Saukkonen, K., S. Sande, C. Cioffe, S. Wolpe, B. Sherry, A. Cerami, and
E. Tuomanen. 1990. The role of cytokines in the generation of inflammation and
tissue damage in experimental Gram-positive meningitis. J. Exp. Med. 171:439.
Tang, T., P. S. Frenette, R. O. Hynes, D. D. Wagner, and T. N. Mayadas. 1996.
Cytokine-induced meningitis is dramatically attenuated in mice deficient in endothelial selectins. J. Clin. Invest. 97:2485.
Bogdan, I., S. L. Leib, M. Bergeron, L. Chow, and M. G. Tauber. 1997. Tumor
necrosis factor-␣ contributes to apoptosis in hippocampal neurons during experimental group B streptococcal meningitis. J. Infect. Dis. 176:693.
Koedel, U., A. Bernatowicz, R. Paul, K. Frei, A. Fontana, and H. W. Pfister.
1995. Experimental pneumococcal meningitis: cerebrovascular alterations, brain
edema, and meningeal inflammation are linked to the production of nitric oxide.
Ann. Neurol. 37:313.
Granert, C., J. Raud, X. Xie, L. Lindquist, and L. Lindbom. 1994. Inhibition of
leukocyte rolling with polysaccharide fucoidin prevents pleocytosis in experimental meningitis in the rabbit. J. Clin. Invest. 93:929.
Libermann, T. A., and D. Baltimore. 1990. Activation of interleukin-6 gene expression through the NF-␬B transcription factor. Mol. Cell. Biol. 10:2327.
Lebel, M. H., B. J. Freij, G. A. Syrogiannopoulos, D. F. Chrane, M. J. Hoyt,
S. M. Stewart, B. D. Kennard, K. D. Olsen, and G. H. McCracken, Jr. 1988.
Dexamethasone therapy for bacterial meningitis: results of two double-blind, placebo-controlled trials. N. Engl. J. Med. 319:964.
Tsenova, L., K. Sokol, V. H. Freedman, and G. Kaplan. 1998. A combination of
thalidomide plus antibiotics protects rabbits from mycobacterial meningitis-associated death. J. Infect. Dis. 177:1563.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
31.
p65 subunit of NF-␬B abrogates established experimental colitis in mice. Nat.
Med. 2:998.
Kastenbauer, S., U. Koedel, and H. W. Pfister. 1999. Role of peroxynitrite as a
mediator of pathophysiological alterations in experimental pneumococcal meningitis. J. Infect. Dis. 180:1164.
Bird, A. P. 1986. CpG-rich islands and the function of DNA methylation. Nature
321:209.
Calame, W., A. E. Douwes-Idema, M. T. van den Barselaar, R. van Furth, and
H. Mattie. 1994. Influence of cytostatic agents on the pulmonary defense of mice
infected with Klebsiella pneumoniae and on the efficacy of treatment with ceftriaxone. J. Infect. 29:53.
Baeuerle, P. A., and T. Henkel. 1994. Function and activation of NF-␬B in the
immune system. Annu. Rev. Immunol. 12:141.
Shakhov, A. N., M. A. Collart, P. Vassalli, S. A. Nedospasov, and
C. V. Jongeneel. 1990. ␬B-type enhancers are involved in lipopolysaccharidemediated transcriptional activation of the tumor necrosis factor ␣ gene in primary
macrophages. J. Exp. Med. 171:35.
Leist, T. P., K. Frei, S. Kam-Hansen, R. M. Zinkernagel, and A. Fontana. 1988.
Tumor necrosis factor ␣ in cerebrospinal fluid during bacterial, but not viral,
meningitis: evaluation in murine model infections and in patients. J. Exp. Med.
167:1743.
Waage, A., A. Halstensen, R. Shalaby, P. Brandtzaeg, P. Kierulf, and T. Espevik.
1989. Local production of tumor necrosis factor ␣, interleukin 1, and interleukin
6 in meningococcal meningitis: relation to the inflammatory response. J. Exp.
Med. 170:1859.
Rajora, N., G. Boccoli, D. Burns, S. Sharma, A. P. Catania, and J. M. Lipton.
1997. ␣-MSH modulates local and circulating tumor necrosis factor-␣ in experimental brain inflammation. J. Neurosci. 17:2181.
Buchner, R. R., T. E. Hugli, J. A. Ember, and E. L. Morgan. 1995. Expression of
functional receptors for human C5a anaphylatoxin (CD88) on the human hepatocellular carcinoma cell line HepG2: stimulation of acute-phase protein-specific
mRNA and protein synthesis by human C5a anaphylatoxin. J. Immunol. 155:308.
Foreman, K. E., A. A. Vaporciyan, B. K. Bonish, M. L. Jones, K. J. Johnson,
M. M. Glovsky, S. M. Eddy, and P. A. Ward. 1994. C5a-induced expression of
P-selectin in endothelial cells. J. Clin. Invest. 94:1147.
Wetsel, R. A. 1995. Structure, function and cellular expression of complement
anaphylatoxin receptors. Curr. Opin. Immunol. 7:48.
Kornelisse, R. F., K. Hoekman, J. J. Visser, W. C. Hop, J. G. Huijmans,
P. J. van der Straaten, A. J. van der Heijden, R. N. Sukhai, H. J. Neijens, and
R. de Groot. 1996. The role of nitric oxide in bacterial meningitis in children.
J. Infect. Dis. 174:120.
Nathan, C. 1997. Inducible nitric oxide synthase: what difference does it make?
J. Clin. Invest. 100:2417.
Springer, T. A. 1994. Traffic signals for lymphocyte recirculation and leukocyte
emigration: the multistep paradigm. Cell 76:301.
Ley, K., G. Linnemann, M. Meinen, L. M. Stoolman, and P. Gaehtgens. 1993.
Fucoidin, but not yeast polyphosphomannan PPME, inhibits leukocyte rolling in
venules of the rat mesentery. Blood 81:177.
Teixeira, M. M., and P. G. Hellewell. 1997. The effect of the selectin binding
polysaccharide fucoidin on eosinophil recruitment in vivo. Br. J. Pharmacol.
120:1059.
Pugin, J., I. D. Heumann, A. Tomasz, V. V. Kravchenko, Y. Akamatsu,
M. Nishijima, M. P. Glauser, P. S. Tobias, and R. J. Ulevitch. 1994. CD14 is a
pattern recognition receptor. Immunity 1:509.
Wright, S. D., R. A. Ramos, P. S. Tobias, R. J. Ulevitch, and J. C. Mathison.
1990. CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS
binding protein. Science 249:1431.
Zysk, G., W. Bruck, I. Huitinga, F. R. Fischer, F. Flachsbarth, N. van Rooijen,
and R. Nau. 1997. Elimination of blood-derived macrophages inhibits the release
of interleukin-1 and the entry of leukocytes into the cerebrospinal fluid in experimental pneumococcal meningitis. J. Neuroimmunol. 73:77.
Tsenova, L., A. Bergtold, V. H. Freedman, R. A. Young, and G. Kaplan. 1999.
Tumor necrosis factor ␣ is a determinant of pathogenesis and disease progression
BACTERIAL DNA INDUCES MENINGITIS