Published online before print June 7, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Division of Infectious Diseases, Department of Research, University Hospitals, Basel, Switzerland;
Department of Neurosurgery, University Hospital, Zurich, Switzerland;
Departments of Medicine and Pathology and Laboratory Medicine, David Geffen School of Medicine at University of California, Los Angeles; and
Laboratory of Immunopharmacology of Microbial Products, Tokyo University of Pharmacy and Life Science, Japan
1Correspondence: Division of Infectious Diseases, Department of Research, University Hospital, Hebelstrasse 20, CH-4031 Basel, Switzerland. E-mail: Regine.Landmann{at}unibas.ch
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: chemokines • polymorphonuclear cells • meningitis
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50% of the survivors suffer from neurological sequelae [2
]. In the pathogenesis of meningitis, penetration of bacteria through the blood-brain barrier (BBB) or blood-cerebrospinal fluid (CSF) barrier initiates activation of brain endothelia and glia and leads to leukocyte recruitment and detrimental inflammatory mediator release [3
, 4
]. Pattern recognition receptors (PRRs) expressed in infiltrating and resident brain cells are stimulated by pneumococci and their cell wall (PCW) components and mediate the harmful inflammation. CD14 was described initially as a PRR for lipopolysaccharide (LPS), the major component of the outer membrane of Gram-negative bacteria [5 ]. Later, CD14 has been shown to bind Gram-positive [6 ] cell wall components. CD14 is a 55-kDa glycosylphosphatidylinositol (GPI)-anchored glycoprotein expressed strongly in human monocytes and all tissue macrophages and weakly in neutrophils [7 ]. Mice, in contrast, have very low or no basal CD14 expression in macrophages [8 ], and polymorphonuclear leukocytes (PMN) have not been studied. A single mRNA transcript is translated and processed in the endoplasmatic reticulum, where a GPI anchor is attached. As a GPI-linked protein without transmembrane and intracytoplasmic domains, membrane-bound CD14 (mCD14) most likely does not transmit a signal on its own. It was therefore postulated that in the presence of Gram-positive bacteria, CD14 Toll-like receptor 2 (TLR2) complexes function to initiate signal transduction, as cellular responses to heat-killed Staphylococcus aureus and S. pneumoniae [9 ] or purified lipoteichoic acid (LTA) [10 ] in transfected cells have been shown to be mediated by CD14 in conjunction with TLR2.
In addition to mCD14, two soluble isoforms of CD14 (sCD14) are generated. A 55-kDa isoform is liberated by escaping from GPI anchoring, and the 49-kDa isoform is derived from the cell membrane by proteolytic cleavage through a serine protease [11 ]. sCD14 can mediate activation of cells that lack mCD14 expression, such as human endothelial, epithelial, and smooth-muscle cells [12 ]. This suggests that sCD14 binds a transmembrane receptor, although it has not been shown that sCD14 binds TLR2 and thereby transmits signals in the same way as mCD14 does. Our and other studies [13 , 14 ] reported elevated levels of sCD14 in human CSF during neuroinflammatory diseases, such as bacterial meningitis. Our group [13 ] also confirmed this observation in a mouse model of experimental bacterial meningitis. CD14 is known to be crucial in endotoxin-induced shock, as shown by the resistance of CD14-deficient mice to septic shock induced by intravenous (i.v.) or intraperitoneal (i.p.) injection of high doses of Escherichia coli or LPS [15 ]. Opposed to that finding, several in vivo studies have shown a deleterious effect of CD14 blockade in mice or rabbits infected with Klebsiella pneumoniae, Shigella flexneri, or low inocula of E. coli [16 17 18 ]. We described previously a pronounced dose-dependent up-regulation of mCD14 and sCD14 in CSF during murine pneumococcal meningitis, and CD14 expression in blood remained low [13 ]. We postulated a role of CD14 in leukocyte immigration into the CSF. We therefore investigated leukocyte recruitment into the subarachnoid space and its consequences on the course of the disease in mice with a targeted deletion of the CD14 gene and in wild-type (WT) mice treated with neutralizing anti-CD14-specific antibody.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Six- to 8-week-old C57BL/6 WT (RCC, Ittingen, Switzerland) and CD14–/– mice (kindly provided by Mason W. Freeman, Lipid Metabolism Unit, Massachusetts General Hospital, Harvard Medical School, Boston; mice had been back-crossed for six generations on a C57BL/6 background) were kept under specific, pathogen-free conditions in the Animal House of the Department of Research, University Hospitals (Basel, Switzerland), according to the regulations of the Swiss veterinary law. Mice were anesthetized via i.p. injection of 100 mg/kg ketamine (Ketalar©, Warner-Lambert AG, Baar, Switzerland) and 20 mg/kg xylazinum (Xylapan©, Graeub AG, Bern, Switzerland) and subsequently, subarachnoidally inoculated in the left forebrain with 0.9% NaCl, live S. pneumoniae [3x103 colony-forming units (CFU)], or PCW (300 µg/kg).
In vivo CD14 blockade experiments were performed by pretreatment with i.v. injection of 1 mg/kg neutralizing 4C1 [22 , 23 ] or control rat immunoglobulin G (IgG)2b monoclonal antibody (mAb) 5 min prior to subarachnoidal injection of 3 x 103 CFU live S. pneumoniae. In vivo CXC chemokine receptor 2 (CXCR2) blockade experiments were performed by i.p. administration of 0.5 ml goat anti-murine CXCR2 serum or control goat serum 2 h prior to S. pneumoniae infection. The health status of the mice was assessed by using severity scores as described previously [19 , 24 ]. Mice, which presented a score of 5 during infection, were killed by i.p. injection of 100 mg/kg pentobarbital (Abbott Laboratories, North Chicago, IL). Blood was obtained by intracardiac puncture and collected in EDTA. Animals were perfused with Ringer’s solution (Braun Medical AG, Emmenbrücke, Switzerland) into the left cardiac ventricle. Brains were collected on dry ice for frozen sections and in 4% paraformaldehyde for paraffin sections. For in situ hybridization, mice were perfused with 4% paraformaldehyde (Sigma Chemical Co., St. Louis, MO) in 0.1 M borax (Sigma Chemical Co.). CSF was obtained by puncture of the cisterna magna as described elsewhere [25 ]. For the leukocyte depletion treatment, cyclophosphamide (Sigma Chemical Co.) was reconstituted with sterile phosphate-buffered saline (PBS) and injected i.p. (300 mg/kg in 0.2 ml) 48 h before S. pneumoniae inoculation.
Determination of bacterial counts and inflammatory parameters
CSF samples were centrifuged at 800 g for 7 min (4°C) to obtain cell-free CSF. Thereafter, they were stored at –20°C until cytokine and chemokine determinations. The pelleted CSF cells were counted and identified via cytospin or phenotypically analyzed. Brains were removed and homogenized with a Polytron homogenizer in PBS for chemokine determinations. Bacterial titers were determined by plating serial tenfold dilutions in 0.9% saline on blood agar plates. The concentration of the proinflammatory cytokine tumor necrosis factor (TNF) in CSF was determined with a bioassay as described previously [19
]. Soluble intercellular adhesion molecule-1 (sICAM-1), interleukin (IL)-6, and the chemokine macrophage inflammatory protein-2 (MIP-2) and keratinocyte-derived chemokine (KC) levels in CSF or brain were measured by enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems Europe, Abingdon, UK), according to the instructions of the manufacturers. Concentrations of sCD14 in CSF and serum were determined by ELISA, as described previously [13
]. The neutralizing rat anti-mouse CD14 antibody did not interfere with the ELISA, in which a monoclonal hamster and a polyclonal rabbit anti-CD14 antibody were used for coating and detection, respectively.
Fluorescein-activated cell sorter (FACS) analysis of blood and CSF cells
Whole EDTA blood after erythrocyte lysis and PMN from CSF were first treated with rabbit serum to block Fc receptors and were then stained by indirect immunofluorescence with goat anti-CXCR2 antiserum (1:100; kind gift from R. M. Strieter), rat anti-granulocyte antibody GR-1 (Becton Dickinson, San Jose, CA), rat anti-CD14 antibody 4C1, and/or hamster anti-CD14 G5A10 [13
] or the corresponding isotype control antibodies. Secondary antibodies were phycoerythrin- and fluorescein-conjugated rabbit anti-goat, rabbit anti-rat, or rabbit anti-hamster antibodies from Becton Dickinson. Fluorescence was analyzed in a FACScan with the Cellquest software. CD11a/CD18 (Becton Dickinson), CD11b/CD18 (BMA), CD11c/CD18 (kind gift from K. Frei), CD31 (clone MEC13.3, Becton Dickinson), CD54 (clone 3E2, Becton Dickinson), and CD62-L (Becton Dickinson) were used for staining of blood and CSF PMN adherence molecules by FACS.
Histology and immunohistochemistry of brain
Histopathological paraffin sections were stained with haematoxylin-eosin (H&E). Immunohistochemical staining was performed from sequential 8 µm coronal cryosections of brain after fixation with polylysine-paraformaldehyde with anti-CD54, -CD106 (Becton Dickinson), and -CD62-E (clone 10E9.6, Becton Dickinson) primary antibodies and then with biotinylated secondary hamster or rat anti-mouse antibodies and developed with the Vectastain Elite ABC reagent (Vector Laboratories, Burlingame, CA) and 3-amino-9-ethyl carbazole chromogen as described previously [13
].
TNF cRNA probes and in situ hybridization
Plasmids were linearized, and the sense and antisense riboprobes were synthesized as described [26
]. Radioactive cRNA copies were synthesized by incubation of 250 ng linearized plasmid in MgCl2 6 mM, Tris (pH 7.9) 40 mM, spermidine 2 mM, NaCl 10 mM, dithiothreitol (DTT) 10 mM, adenosine 5'-triphosphate/guanosine 5'-triphosphate/cytidine 5'-triphosphate 0.2 mM, 100 µCi 
-35S-uridine 5'-triphosphate (Dupont NEN, Boston, MA), 20 U RNasin (Promega, Catalys, San Luis Obispo, CA), and 10 U T7, SP6, or T3 RNA polymerase for 60 min at 37°C. Unincorporated nucleotides were removed using the ammonium-acetate precipitation method, 100 µl DNase solution was added, and 10 min later, a phenol-chloroform extraction was performed. The cRNA was precipitated with 80 µl 5 M ammonium acetate and 500 µl 100% ethanol for 20 min on dry ice. The pellet was dried and resuspended in 50 µl 10 mM Tris/1 mM EDTA. A concentration of 107 counts per minute/probe was mixed into 1 ml hybridization solution (500 µl formamide, 60 µl 5 M NaCl, 10 µl 1 M Tris [27
], 2 µl 0.5 M EDTA [27
], 50 µl 20x Denhart’s solution, 200 µl 50% dextran sulfate, 50 µl 10 mg/ml tRNA, 10 µl 1 M DTT). This solution was mixed and heated for 10 min at 65°C before being spotted on slides.
Hybridization histochemical localization of TNF mRNA was carried out on every 12th section of the whole rostro-caudal extent of each brain using 35S-labeled cRNA probes as described previously [26 ]. The sections were exposed at 4°C to X-ray films (Biomax, Kodak, Rochester, NY) for 1–3 days. Thereafter, the slides were defatted in xylene, dipped in NTB-2 nuclear emulsion (Kodak), and exposed for 10 days. The slides were then developed in a D19 developer (Kodak) for 3.5 min at 14–15°C, washed 15 s in water, and fixed in rapid fixer (Kodak) for 5 min. Thereafter, tissues were rinsed in running, distilled water for 1 h, counterstained with thionin (0.25%), dehydrated through graded concentrations of alcohol, cleared in xylene, and coverslipped with distrene plasticizer xylene-mounting medium (Electron Microscopy Science, Washington, PA). Immunohistochemistry was combined with in situ hybridization to determine the cell type, which expresses TNF transcripts and CD45 used as a marker for infiltrating leukocytes. Sections were processed by using the avidin-biotin amplification bridge method with peroxidase as a substrate, as described previously [26 ].
Preparation of peripheral blood PMN cells for chemotaxis
Mouse blood was harvested by intracardiac puncture and collected in EDTA, and PMN were isolated as described previously for human PMN [28
], but using a modified density gradient centrifugation on a discontinuous Percoll gradient with 59% and 67% Percoll (Percoll, Amersham Pharmacia Biotech, UK) in PBS. The interface of the two Percoll layers, which contain the PMN, was collected, and contaminating erythrocytes were removed by hypotonic lysis in water. By morphologic criteria, the final cell preparation contained >97% PMN, and the viability of the cells, as assessed by trypan blue exclusion, was >95%.
Chemotaxis assay
Migration of mouse PMN was assessed in vitro using 48-well microchemotaxis chambers as described previously [29
]. Dilutions of samples or CSF diluted 1:2 (v/v) in 28 µl Hanks’ Hepes buffer were added to the lower wells in triplicates. A polyvinylpyrrolidone-free polycarbonate filter membrane (3 µm pore size, Nuclepore, Sterico AG, Dietikon, Switzerland) was placed between the lower wells and the upper wells containing 45 µl cell suspensions (6.9x105 cells/ml in Hanks’ Hepes buffer). The chemotaxis chambers were incubated for 60 min at 37°C in 5% CO2 in humidified air. After incubation, the filters were removed, fixed, and stained with Wright-Giemsa. The number of cells that migrated through the pores of the filter was counted in six high-powered fields (x63). Cell migration was expressed as the mean percentage of PMN that migrated per well. Hanks’ Hepes buffer alone was used as a control for random migration and N-formylmethionyl-leucyl-phenylalanine (Sigma Chemical Co.) as a positive control at a concentration of 10–6 M.
For neutralization experiments, CSF samples were preincubated with 30 µg/ml anti-MIP-2 (R&D Systems) or control rat IgG2b mAb for 120 min at 37°C.
Microvascular endothelial cell (MVEC) cultures
Brain-derived MVEC were isolated from 6- to 12-week-old C57BL/6 or CD14-deficient mice of either sex, according to published methods [30
31
32
]. Cells were grown at 37°C in a humidified 8% CO2 atmosphere with medium replacement every 2 days. MVEC were grown to confluence and passaged between day 6 and 9. MVEC cultures were at least 90% pure, as determined with the expression of von Willebrand factor (Dako, Glostrop, Denmark) and uptake of Dil acetylated low-density lipoprotein (Molecular Probes, Yuro Supply, Lucerne, Switzerland). Cells were trypsinized (0.25% trypsin/EDTA, Biological Industries, Kibbutz Beit Haemek, Israel) and seeded in collagen I-coated culture dishes (2x105 cells/dish, 35 mm diameter, BioCoat, Fort Washington, PA). After overnight culture, the medium was replaced with fresh culture medium without gentamycin, and viable S. pneumoniae were added at a multiplicity of infection (MOI) of 2 and 20. After 2 h, the bacteria-containing medium was removed, and the cells were washed twice with PBS and replaced with fresh Dulbecco’s modified Eagle’s medium culture medium supplemented with 100 µg/ml gentamycin to prevent extracellular bacterial growth. Supernatants were harvested 8 and 24 h later and stored at –20°C until ELISA analysis.
Statistical analysis
Differences in survival between WT and CD14–/– mice were tested using the log rank test and Kaplan-Meier analysis. Results of the CFU in brain homogenates, leukocyte numbers in blood and CSF, as well as the chemotactic activity of CSF samples were compared with the Mann-Whitney U-test.
The relationship between two parameters (parametric variables) was assessed in a linear regression model using Spearman rank correlation test. In all statistical tests, a P < 0.05 was considered statistically significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (37K): [in a new window] |
Figure 1. (A) CD14 expression in blood resting PMN and CSF PMN 24 h after infection with 3 x 103 CFU S. pneumoniae (dashed line, isotype control; thin line, blood PMN; thick line, CSF PMN). One out of three experiments with three mice is represented. (B) Time course of pleocytosis into the CSF of WT and CD14–/– mice after infection with S. pneumoniae. CSF was collected at indicated times, and infiltrating PMN were counted at 6 h (n=15), 12 h (n=11), and 24 h (n=11). Results are represented as mean ± SD. *, P < 0.05. (C) H&E stains of WT (a) and CD14–/– (b) brain 24 h after infection. One representative example out of four experiments is shown. Infiltrating cells are accumulated in the ventricles and subventricular region.
|
 View larger version (24K): [in a new window] |
Figure 2. Leukocyte numbers in (A) blood (n=15 or 17) and (B) CSF (n=9 or 11) in nontreated and leukocyte-depleted WT and CD14–/– mice 24 h after infection with 3 x 103 CFU of S. pneumoniae. Cyclophosphamide (300 mg/kg) was injected i.p. 48 h prior to S. pneumoniae inoculation. Mean ± SD of four independent experiments. *, P < 0.05; **, P < 0.01.
|
Alternatively, chemokine levels in the CSF or brain could affect neutrophil transmigration. Therefore, the neutrophil chemoattractants MIP-2 and KC were measured. MIP-2 and KC levels in CSF increased during infection but were not significantly higher in CD14–/– than in WT mice (Table 1 ), and ICAM-1, which could also play a role in neutrophil transmigration [33 ], did not change during infection (Table 1) . For IL-8, the potent PMN chemoattractant in humans, a strong association with PMN was shown to occur [34 ]. Furthermore, microglial cells, astrocytes, and endothelial cells are known to be a source of chemokines during inflammation or after bacterial infection [35 36 37 ]. Therefore, MIP-2 was also measured in brain lysates, which contain adherent PMN and resident cells. MIP-2 was significantly higher in brains of CD14–/– than of WT mice 12 h after infection (Fig. 3A ). As it was suggested earlier that IL-6 contributes to the control of MIP-2 [38 ], IL-6 concentrations were measured in plasma and CSF but did not differ in the two mouse strains between 12 and 48 h (data not shown).
|
View this table: [in a new window] |
Table 1. MIP-2, KC, and sICAM-1 Levels in CSF of CD14–/– and WT Mice 6 and 12 h after Subarachnoidal Infection with 3 x 103 CFU of S. pneumoniae
|
 View larger version (17K): [in a new window] |
Figure 3. (A) Time course of MIP-2 levels in brain homogenates from CD14–/– (n=3) and WT (n=3) mice after infection with 3 x 103 CFU of S. pneumoniae. Mean ± SD of three experiments is shown. **, P < 0.01. (B) Time- and MOI-dependent MIP-2 production by CD14-deficient and WT brain-derived MVEC. Mean of triplicate cultures is shown.
|
In summary, increased brain-derived chemokine levels but not altered adherence molecules were found associated with enhanced PMN migration in CD14–/– mice during meningitis.
CSF, MIP-2, and PCW show a stronger chemotactic activity on PMN from CD14–/– mice, and CXCR2 is increased in CD14–/– mice
CSF of patients with bacterial meningitis is known to induce chemotactic activity in human PMN in vitro [27
]; therefore, chemotaxis of mouse blood PMN toward CSF from infected CD14–/– and WT mice was tested in vitro. The percentage of PMN, which migrated toward CSF was significantly higher if cells were derived from CD14–/– than from WT mice (Fig. 4A
). The origin of CSF did not influence the migration capacity of PMN cells of either mouse strain (data not shown). Then, the effect of chemokines in CSF on PMN migration was assessed, and chemotaxis was performed with infected CSF, which was pretreated with neutralizing anti-MIP-2 alone or combined with anti-KC mAb. As shown in Figure 4A , preincubation of CSF with anti-MIP-2 mAb reduced the number of migrated PMN, as compared with CSF preincubated with control rat IgG, to the level of the value that was observed with WT PMN. No additional blockade of chemotaxis was observed with neutralization of MIP-2 and KC. A stronger migration was also observed in CD14–/– PMN when recombinant murine (rmu)MIP-2 was used as chemoattractant, and this effect was neutralized by anti-MIP-2 mAb (Fig. 4B)
. Thus, despite similar MIP-2 levels in CSF of both mouse strains, CD14–/– PMN were more attracted toward MIP-2 than WT PMN. These results suggested that the MIP-2 receptor CXCR2 was implicated in the enhanced migration found in CD14–/– mice [39
]. Therefore, CXCR2 expression levels in blood and PMN from infected CSF were investigated (Fig. 5
). CD14–/– mice presented with a similar fraction of CXCR2-positive cells as WT mice in resting blood and in CSF cells 12 h after infection (Fig. 5A)
. However, 24 h after infection, 76% of CD14–/– PMN in CSF but only 40% of WT cells expressed CXCR2 (Fig. 5A)
. The intensity of CXCR2 expression, as measured by MFI, was similar in blood PMN of the two mouse strains, but MFI was twofold higher in CSF cells of CD14–/– than WT mice (Fig. 5B)
. Finally, the chemokine MIP-2 and its receptor CXCR2 were not the only factors contributing to the increased chemotaxis in CD14–/– mice. PCW also caused a stronger migration of CD14–/– than of WT PMN (Fig. 4B)
, suggesting that other receptors such as platelet-activating factor (PAF) receptors, which interact with PCW phosphorylcholine, may also be under CD14 control.
 View larger version (18K): [in a new window] |
Figure 4. Chemotactic activity of PMN from CD14–/– and WT mice toward CSF and MIP-2 in a 48-well microchemotaxis assay. (A) Percentage migrated PMN toward CSF (50% final concentration in chemotaxis chamber), collected and filtered 24 h after subarachnoidal infection with S. pneumoniae. The CSF was pretreated with anti-MIP-2, anti-KC, or rat IgG. (B) Percentage migrated PMN toward MIP-2 ± anti-MIP-2 or PCW. The percentage of migrated neutrophils was determined by counting 600–900 cells per well. Mean ± SD of three independent experiments. *, P < 0.05; **, P < 0.01.
|
 View larger version (22K): [in a new window] |
Figure 5. CXCR2 expression in blood and CSF PMN during meningitis with 3 x 103 CFU of S. pneumoniae. (A) Percentage CXCR2-positive PMN in blood (0 h) and CSF (12 and 24 h) of WT and CD14–/– mice. Mean values of three experiments are shown. (B) Mean fluorescence intensity (MFI) of CXCR2-positive PMN in blood (0 h) and CSF PMN (12 h after infection). Dashed line, Isotype control; solid line, CXCR2 fluorescence. One representative out of three experiments is shown.
|
CD14–/– mice show earlier death during pneumococcal meningitis
To investigate the in vivo consequence of excess neutrophil infiltration in CSF of CD14–/– mice, the severity of disease and the outcome were analyzed in WT and CD14–/– mice with S. pneumoniae meningitis; as all untreated mice die in this model, a delay or acceleration in the disease severity and mortality indicates an altered host response. S. pneumoniae meningitis was associated with a gradual appearance of lethargy and seizures in WT and CD14–/– mice, but the percentage of severely sick mice was significantly higher among CD14–/– than among WT mice from 48 h after infection, and this difference was maintained during infection (Fig. 6A
). Higher disease severity was followed by earlier mortality in CD14–/– mice with a median survival time of 84 h as compared with 120 h in WT mice (Fig. 6B
, P<0.01). In vivo blockade of CD14 with anti-CD14 mAb also accelerated the mortality after S. pneumoniae infection (data not shown). CD14-deficient and antibody-treated mice differ by the presence of sCD14, which was reported to increase in CSF of patients with meningitis and in experimental meningitis [13
]. Accordingly, in the present study, we also found increased levels of sCD14 in CSF. As values did not differ in anti-CD14 and isotype control mAb-treated mice (3627.5±1862.6 vs 3209.0±3540.4 ng/ml), an effect of antibody treatment on sCD14 function appears unlikely. In summary, the survival studies indicate that CD14 is implicated directly in the host response to pneumococcal meningitis.
 View larger version (21K): [in a new window] |
Figure 6. (A) Percentage of CD14–/– (n=22) and WT (n=32) mice showing high severity score (including score 4 and score 5) after infection with 3 x 103 CFU of S. pneumoniae. Mean ± SD of three independent experiments (*, P<0.02). (B) Survival of CD14–/– and WT mice after infection with S. pneumoniae. CD14–/– mice (thin, gray line; n=22) showed earlier death as compared with WT mice (thick, black line; n=32). P < 0.01.
|
 View larger version (17K): [in a new window] |
Figure 7. (A) Bacterial load in brain homogenates from CD14–/– and WT mice 24 h (n=17 and 18, respectively) and 48 h (n=8 and 11, respectively) after infection with 3 x 103 CFU of S. pneumoniae. Results from individual mice are shown. CD14–/– brains showed more CFU counts than WT brains 24 h after infection (**, P<0.01). (B) Time course of TNF release into the CSF of CD14–/– and WT mice after infection with S. pneumoniae. CSF was collected at indicated times, and TNF was measured at 6 h (n=15), 12 h (n=11), and 24 h (n=11). Results are represented as mean ± SD. *, P < 0.05.
|
 View larger version (41K): [in a new window] |
Figure 8. (A) In situ hybridization of TNF mRNA in brains of immunocompetent control (upper brains) and leukocyte-depleted (lower brains) WT (left) and CD14–/– (right) mice 24 h after infection of S. pneumoniae. Combined bright/dark field photomicrographs of coronal brain cryosections (20 µm) hybridized as described in Materials and Methods. (B) Immunohistochemistry for CD45 combined with in situ hybridization for TNF in WT (left) and CD14–/– (right) mice 24 h after infection. One representative out of three experiments is shown.
|
 View larger version (24K): [in a new window] |
Figure 9. (A) Survival of CD14–/– and WT mice without and with anti-CXCR2 antibody pretreatment. Untreated CD14–/– (thick, gray line, n=22) and WT (thick, black line, n=21) mice; treated CD14–/– (thin, gray line, n=7) and WT (thin, black line, n=7) mice. (B) CSF pleocytosis and MIP-2 levels in brain homogenates 12 h after infection in isotype control-treated (n=3) and anti-CXCR2 antibody-treated (n=3) mice. Mean ± SD of three independent experiments is shown. *, P < 0.05.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The lack of CD14 did not impair systemic host defense, as bacterial counts in blood and serum IL-6 levels of CD14–/– and WT mice were similar. CFU numbers in CSF did not differ at any time-point; this suggests that CD14 did not directly affect clearance of S. pneumoniae by host cells and supports earlier observations, where E. coli binding and uptake did not differ in CD14–/– and WT peritoneal macrophages [42 ]. We have previously observed a slower bacterial clearing in TLR2–/– than WT mice during meningitis in vivo [19 ] and in granulocytes in vitro (unpublished observations). Our data in CD14–/– mice show that CD14 is not involved in the TLR2-modulated bacterial phagocytosis and killing. They suggest that for this function, CD14 is not linked to TLR2.
However, despite an early PMN immigration into CSF of CD14–/– mice, bacteria were accumulating faster in CD14–/– than in WT brain with ongoing disease. This suggests that the enhanced inflammation, which followed the high PMN numbers in CSF, might have prevented efficient bacterial clearing.
In an earlier study, we found CD14 induced in PMN and in soluble form, in CSF of pneumococcus-infected WT mice [13 ]. Here, we found CD14 deficiency associated with excess infiltrating PMN early in infection, which occurred even in the presence of an 80% reduction in blood leukocytes by cyclophosphamide. Our findings are in agreement with a previous study, which described more infiltrating PMN in peritoneum of CD14–/– than of WT mice after i.p. injection of E. coli [43 ]. Analysis of adhesion molecules in PMN and of their counter-ligands in brain endothelia revealed similar increased expression in brains of WT and CD14–/– mice after infection, excluding a CD14-dependent modulation of these adhesion molecules in our model.
In addition to adhesion molecules, chemokines and bacterial components are the best-known candidates for leukocyte recruitment to inflammatory sites [44
]. The human and mouse CXC chemokines growth-related oncogene- (GRO-) and MIP-2 were found in CSF of patients with bacterial meningitis and of mice with experimental Listeria monocytogenes meningitis, respectively, and were shown to mediate chemotaxis in vitro [27
, 45
]. Intracerebral injection of KC (murine homologue of GRO-) and MIP-2 resulted in neutrophil recruitment into the CSF [46
], and blockade of KC as well as of MIP-2 or its receptor CXCR2 [47
] has been demonstrated to reduce neutrophil migration in several mouse-infection models [45
, 48
49
50
]. In the present study, MIP-2 and KC concentrations in CSF were similar in WT and CD14–/– mice, which is consistent with a previous observation excluding any correlation between neutrophil counts and chemokine concentrations in CSF during bacterial meningitis [51
]. However, MIP-2 in brain homogenates was tenfold higher in CD14–/– than in WT mice. It was likely derived from PMN-bound MIP-2 [34
] and from resident cells [35
36
37
]. In this study, we identified a resident cell source: CD14–/– endothelial cells released more MIP-2 upon in vitro S. pneumoniae infection than WT endothelia. Thus, the absence of CD14 led to an accumulation of tissue-derived MIP-2 but not of MIP-2 in CSF. In an earlier study, cerebral MIP-2 mRNA was found higher in IL-6–/– compared with WT mice with pneumococcal meningitis [38
]. This suggests that CD14 might control MIP-2 via the intermediate of IL-6. However, as IL-6 levels did not differ in the two mouse strains, we could exclude this hypothesis. At this time, the mechanism of excess MIP-2 release in CD14–/– endothelial cells is unknown, as WT endothelial cells lack mCD14 (unpublished observations), and sCD14 was supplied in both cultures by adding fetal calf serum. Furthermore, it remains open whether MIP-2 in CD14-deficient brains and endothelia increased because of an enhanced production or a slowed degradation.
CD14–/– PMN showed excess migration in vitro upon stimulation with rmuMIP-2. Furthermore, the increased chemotaxis of CD14–/– PMN toward infected CSF was abolished by anti-MIP-2. Yet, MIP-2 concentrations did not differ in CSF from WT and CD14–/– mice; therefore, it was postulated that CD14-deficient PMN migrated more toward CSF because of increased chemokine receptor expression or a stronger signaling. Indeed, CXCR2 expression was twofold higher early in infection in PMN of CD14–/– than WT mice, and it was detectable on the surface of CD14–/– PMN for a longer time than in WT cells. The phenotype of CD14–/– mice with increased CXCR2 expression had a strong impact on outcome, as anti-CXCR2 treatment abolished the earlier death in CD14–/– as compared with WT mice. In fungal infections, where early leukocyte recruitment is essential, mice that lack the CXCR2 receptor showed a severe impairment of neutrophil recruitment and a bad outcome [47 ]. Here, in meningitis, the blockage of CXCR2 was beneficial early, as it abolished fast recruitment of PMN into CSF of CD14–/– mice and thereby possibly blocked the increased TNF and aggravation of the disease observed in CD14–/– mice.
The mechanism by which CD14 affected CXCR2 expression is unknown. Recently, it has been shown that LPS acts through TLR4 to down-regulate expression of two kinases interacting with a G protein-coupled receptor in response to MIP-2. Down-regulation of these kinases resulted in decreased CXCR2 receptor internalization and increased PMN migration [52
]. Furthermore, the TLR2 agonist Pam3CysSerLys4 was found to reduce surface expression of CXCR2 in PMN [53
, 54
]. These findings were correlated with a down-regulation of CXCR2 expression [54
]. A similar, more complex mechanism with bacterial stimulation of CD14 and/or TLR2 could be involved in our model of Gram-positive meningitis and lead to increased chemokine levels, chemokine receptor up-regulation, and migration [10
, 55
, 56
]. Our analysis of CD14/TLR2 expression in blood and CSF PMN from WT mice revealed that in blood, a small proportion (25%) of PMN carries TLR2 and CD14, and the majority (70%) has only TLR2, and in CSF, 90% of the cells carry TLR2 and CD14, and 10% have only TLR2 (unpublished results). These data show that TLR2-positive PMN acquired CD14 during infection and suggest that CD14 may affect TLR2-mediated CXCR2 expression on these cells and thus control their migration. Subpopulations of PMN equipped with different PRRs may thus regulate inflammation by modulating migration. In our model, not only CSF from mice with pneumococcal infection but also subarachnoidal injection of PCW induced a stronger neutrophil migration in CSF of CD14–/– than of WT mice. This suggested the presence of additional bacterial chemotactic factors. Indeed, peptidoglycan, LTA, and teichoic acid, which are the most active components of PCW, induced leukocyte pleocytosis in vivo after intracisternal injection into rabbits [57
]. It is unlikely that these factors were different in WT and CD14–/– mice, as early bacterial load was similar in both strains. However, receptors for PCW possibly differed in WT and CD14–/– mice. PAF receptors are among the candidate receptors that bind phosphoryl-choline in several PCW [58
] proteins and may be regulated by CD14.
In conclusion, the present study shows that the lack or blockade of CD14 is associated with a stronger neutrophil recruitment into CSF, leading to excessive meningeal inflammation and aggravated disease after S. pneumoniae infection. The stronger neutrophil migration correlates with increased chemokine concentrations in brain and with an enhanced migratory capacity of CD14-deficient PMN as a result of increased CXCR2 chemokine receptor expression.
Received February 1, 2005; revised April 28, 2005; accepted April 29, 2005.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
converting enzyme as adjuvant therapy in pneumococcal meningitis Brain 124,1734-1742 and -2 are produced intrathecally and mediate chemotactic activity in cerebrospinal fluid of infected mice J. Immunol. 155,4367-4375[Abstract]
differentially modulates the release of cytokines and chemokines in lipopolysaccharide- and pneumococcal cell wall-stimulated mouse microglia and macrophages Eur. J. Neurosci. 16,2113-2122[CrossRef][Medline] production J. Immunol. 162,4801-4805 attenuates neutrophil recruitment in the central nervous system during experimental bacterial meningitis Infect. Immun. 67,2590-2601This article has been cited by other articles:
 |
|
 |
|
  M. Schmaler, N. J. Jann, F. Ferracin, L. Z. Landolt, L. Biswas, F. Gotz, and R. Landmann Lipoproteins in Staphylococcus aureus Mediate Inflammation by TLR2 and Iron-Dependent Growth In Vivo J. Immunol., June 1, 2009; 182(11): 7110 - 7118. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Farooq, R. Stillie, M. Svensson, C. Svanborg, R. M. Strieter, and A. W. Stadnyk Therapeutic Effect of Blocking CXCR2 on Neutrophil Recruitment and Dextran Sodium Sulfate-Induced Colitis J. Pharmacol. Exp. Ther., April 1, 2009; 329(1): 123 - 129. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Mildner, M. Djukic, D. Garbe, A. Wellmer, W. A. Kuziel, M. Mack, R. Nau, and M. Prinz Ly-6G+CCR2- Myeloid Cells Rather Than Ly-6ChighCCR2+ Monocytes Are Required for the Control of Bacterial Infection in the Central Nervous System J. Immunol., August 15, 2008; 181(4): 2713 - 2722. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Kesteman, G. Vansanten, B. Pajak, S. M. Goyert, and M. Moser Injection of lipopolysaccharide induces the migration of splenic neutrophils to the T cell area of the white pulp: role of CD14 and CXC chemokines J. Leukoc. Biol., March 1, 2008; 83(3): 640 - 647. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Dessing, S. Knapp, S. Florquin, A. F. de Vos, and T. van der Poll CD14 Facilitates Invasive Respiratory Tract Infection by Streptococcus pneumoniae Am. J. Respir. Crit. Care Med., March 15, 2007; 175(6): 604 - 611. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. K. Paterson and T. J. Mitchell Innate immunity and the pneumococcus Microbiology, February 1, 2006; 152(2): 285 - 293. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
