| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Molecular Immunopathology Unit, Bosch Institute, School of Medical Sciences and
School of Molecular and Microbial Biosciences, University of Sydney, New South Wales, Australia;
Mycobacterial Research Division, Centenary Institute, Camperdown, New South Wales, Australia; and
Perlmutter Laboratory, Children’s Hospital and Harvard Medical School, Boston, MA 02115
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
and lymphotoxin-
in the brains of mice protected from FMCM. These data suggest a role for CXCR3 in the pathogenesis of FMCM through the recruitment and activation of pathogenic CD8+ T cells. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Proinflammatory cytokines are known to play an integral role in the pathogenesis of fatal murine CM (FMCM), as IFN-–/– (3) and lymphotoxin (LT)-–/– (4) mice are protected against cerebral symptoms. The activation of the endothelium is another key event in the pathogenesis of FMCM, and mice deficient in adhesion molecules, for example, ICAM-1 (5), also are resistant to FMCM, because these can facilitate the adhesion of monocytes and leukocytes (6, 7). Adhesion molecules also are important for the adherence of parasitized RBCs (PRBCs) to the endothelium during human CM (8, 9). It is thought that these events lead to damage of the blood-brain barrier (BBB) (10, 11, 12), thus contributing to the neurological signs of FMCM, by permitting entry of inflammatory mediators into the CNS.
Recent work has suggested that damage to the endothelium is mediated by CD8+ T cells, which are essential for the development of FMCM (3) and also contribute to circulatory shock during malaria infection (13). CD8+ T cells are likely to be responsible for the early IFN- production associated with protection against FMCM (14). Furthermore, CD8+ T cells are found to sequester within the brain at the onset of cerebral symptoms (11), and these cells are activated and able to secrete proinflammatory cytokines (15). Recent work from our laboratory (16, 17, 18) and others (19, 20) has suggested that CD8+ T cells contribute to FMCM via perforin- and FasL-mediated pathways. Recruitment of these encephalitogenic T cells during FMCM is dependent on the chemokine receptor CCR5 (21) but not CCR2 (22).
CD8+ T cells also can be recruited via CXCR3, a chemokine receptor commonly found on memory and activated T cells, particularly the Th1 subset, as well as a proportion of circulating T cells, NK cells, and B cells (23). Others also report CXCR3 expression on recruited and peripheral blood monocytes (24), as well as plasmacytoid monocytes (25). Previously, CXCR3 has been shown to be important in a variety of diseases including allograft rejection (26) and lymphocytic choriomeningitis virus infection (27) primarily through modulation of leukocyte infiltration.
Chemokine expression during CM has not been extensively studied (as reviewed in Ref. 28). Some chemokines have been shown to be up-regulated in the plasma during severe falciparum malaria (29, 30), as a consequence of macrophage induction by hemozoin (31, 32, 33), or possibly due to activated neutrophils (34). Some chemokines are expressed in the brain during human CM (35) and also during murine CM induced by either Plasmodium berghei ANKA (PbA; Ref. 36) or Plasmodium yoelii 17XL (37), possibly due to astrocyte stimulation (36). In examining all known murine chemokine ligands and receptors using real-time PCR, we found up-regulation of chemokine and chemokine receptor mRNAs associated with inflammatory responses, including CXCR3, within brains during FMCM. Chemokine RNA expression was localized to the cerebral microvessels and surrounding reactive glial cells. Subsequent studies using CXCR3–/– mice revealed that deficiency in this gene conferred protection against FMCM in the majority of mice, and this was associated with reduced localization of CD8+ T cells in deep brain vessels, as evidenced by quantitative flow cytometry, immunohistochemical staining, and real-time PCR. Furthermore, this protection observed in CXCR3–/– mice was abolished following the adoptive transfer of CD8+ cells from C57BL/6 mice with FMCM.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Female mice (6–8 wk) were housed in the Blackburn Animal House (University of Sydney, Sydney, Australia) and given food and water ad libitum. CBA/T6, C57BL/6 (wild-type (WT)) and CXCR3–/– gene knockout (KO) mice (26), backcrossed into C57BL/6 for 11 generations, were used in these experiments. PbA (courtesy of Prof. G. Grau, University of Sydney, Sydney, Australia) inoculation of CBA/T6 or WT mice is a common model of FMCM. Mice inoculated i.p. with 1 x 106 PbA PRBCs display neurological symptoms by days 6–7 postinoculation (p.i.), whereas mice inoculated with P. berghei K173 (PbK) PRBCs (courtesy of Prof. I. Clark, Australian National University, Canberra, Australia), do not show these neurological signs, but instead succumb to acute anemia and hyperparasitemia 
2 wk p.i. (38). Parasite burden, the total number of circulating erythrocytes containing at least one parasite, was calculated as described previously (39). All procedures involving the use of animals were approved by, and conducted according to the regulations of, the Animal Ethics Committee (University of Sydney).
Tissue extraction, sample preparation, and histopathology
Mice were anesthetized by isoflurane inhalation and euthanized at each stated time point. Mice were perfused with 10 ml of cold sterile PBS to clear the vessels of blood and organs were removed. Formalin-fixed, paraffin-embedded sections (7 µm) were used to examine histopathological features, as described previously (17).
Fibrinogen and CD8 immunohistochemistry (IHC)
Fresh frozen sections (7 µm) were cut onto Superfrost Plus slides (Menzel-Glaser) and IHC was performed as described previously (18), except that sections were fixed in acetone for 10 min before staining. For fibrinogen immunostaining, sections were washed in TNT buffer (0.1 M Tris-HCl (pH 7.5) containing 0.3 M NaCl, 0.05% (v/v) Tween 20) and blocked with 0.05% (v/v) H2O2/methanol. The primary Ab used was rabbit anti-human fibrinogen primary Ab (1/50,000; in 1% (v/v) normal horse serum (NHS)/TNT; DakoCytomation) and the secondary Ab used was a biotinylated goat anti-rabbit secondary Ab (1/300; in 1% (v/v) NHS/TNT; Vector Laboratories). For CD8 immunostaining, sections were washed in TBS buffer (0.1 M Tris-HCl (pH 7.5) containing 0.3 M NaCl) and blocked using 0.3% (v/v) H2O2/PBS. The primary Ab used was rat anti-mouse CD8a (1/100; in 5% (v/v) NHS/TBS; BD Pharmingen) and the secondary Ab used was a biotinylated goat anti-rat Ab (1/300; in 5% (v/v) NHS/TBS; Caltag Laboratories). Sections were visualized using ABC peroxidase (Vector Laboratories) and diaminobenzidine (DakoCytomation).
Adoptive transfer of CD8+ cells
Splenocytes from WT donors were prepared 6 days following inoculation with PbA as described previously (19). Briefly, CD8+ cells were stained with an allophycocyanin-conjugated CD8 Ab (eBioscience) and subsequently separated using a FACSAria (BD Biosciences) with a purity of >98%. Alternatively, CD8+ cells were stained with FITC-conjugated CD8 Ab (BD Biosciences) before being washed, resuspended in MACS buffer, and stained with anti-FITC microbeads (Miltenyi Biotec), as per the manufacturer’s instructions. CD8+ cells were purified by separation on an AutoMACS (Miltenyi Biotec) and purity was evaluated by FACS analysis. After three rounds of separation, >96% CD8+ T lymphocyte enrichment was obtained. CD8– cells contained <0.8% CD8+ cells. Recipient mice (WT or CXCR3–/–) were administered i.v. with either 5 x 106 purified CD8+ or CD8– <30 min before inoculation with PbA. Parallel inoculation of WT and CXCR3–/– mice with PbA, without supplementation of CD8+ or CD8– cells, served as further controls.
Reagents for flow cytometric analysis
FITC-anti-CD3 (clone: 145-2C11) and PE-anti-CD8 (clone: 53-6.7) were sourced from eBioscience. PercP-anti-CD4 (clone: RM4-5), PeCy7-anti-CD45 (clone: 30-F11), PE-anti-NK1.1 (clone: PK136), and PE-anti-CD11b (clone: M1/70) were obtained from BD Pharmingen. Allophycocyanin-anti-CXCR3 (clone: 220803) was from R&D Systems.
Isolation of brain-sequestered leukocytes (BSLs)
Sacrificed mice were perfused and the meninges dissected away from the isolated brain, which was gently crushed between the frosted ends of two glass slides. The resulting mixture was resuspended in 10 ml of PBS-5% (v/v) FCS containing 5 mM MgCl2, DNase-I (28 IU/ml; Sigma-Aldrich) and collagenase (0.5 mg/ml; Sigma-Aldrich). The cell suspension then was incubated at 37°C for 30 min with occasional agitation, then triturated several times to break up tissue clumps and returned to incubate at 37°C for a further 30 min. Following enzymatic digestion, cells were pelleted and resuspended in 7 ml of 30% (v/v) Percoll (Pharmacia) in PBS-5% (v/v) FCS, then centrifuged at 1600 x g for 20 min. The upper Percoll layers containing myelin debris and cells other than leukocytes were carefully removed and the cell pellet resuspended in PBS-5% (v/v) FCS. After washing cells in PBS-5% (v/v) FCS, any residual PRBCs were removed by water lysis and washed again in 15 ml of PBS-5% (v/v) FCS before staining.
Flow cytometric analysis
Single-cell suspensions of splenocytes and BSLs were stained according to standard protocols. Where populations could be identified based on clear separation of staining intensity, regions were set around population clusters. For Ags where continuous staining patterns were seen, fluorescence minus one controls were used to set positive and negative regions (40). Cells were initially gated on forward and side scatter, then on forward scatter vs CD45 to separate CD45high leukocytes from resident microglia (41, 42). Cell populations were identified as follows: CD4+ T cells (CD45high, CD3+, CD4+), CD8+ T cells (CD45high, CD3+, CD8+), NK cells (CD45high, CD3–, NK1.1+), NKT cells (CD45high, CD3int, NK1.1+), and myeloid lineage cells (CD45high, CD3–, CD11b+). Quantitation of BSLs was accomplished by spiking the samples with fluorescent microspheres (Invitrogen Life Technologies), at a concentration of 2.5 x 104/ml (43). Samples were analyzed using a FACSAria (BD Biosciences) and data analysis was performed using FlowJo.
Real-time PCR
Real-time PCR was performed as described previously (44), except that PCRs were performed with the Corbett Rotor-Gene (Corbett Research) and using Platinum SYBR Green qPCR SuperMix UDG (Invitrogen Life Technologies). mRNA levels were normalized to a housekeeping gene, hypoxanthine guanine phosphoribosyltransferase (HPRT), and expressed relative to the mean of uninfected control samples using the 
cycle threshold (CT) method. The specificity of primers (Table I) was checked using melt curve analyses and the primer sets used had similar amplification efficiencies.
|
LCM was performed based on a published method from our laboratory (45), with a few modifications. Fresh frozen samples were collected as described above, cut into 7-µm sections onto Superfrost Plus slides (Menzel-Glaser), and stained in a similar fashion to a previously described protocol (46). Briefly, these were quickly fixed in 70% (v/v) ethanol, washed in PBS, stained with a fluorescent lectin (fluorescein Griffonia (Bandeiraea) simplicifolia lectin; 1/20 dilution in PBS; Vector Laboratories) for 2 min in the dark before washing again in PBS (5 x 3 s). Sections were then dehydrated quickly through a series of graded alcohols (70, 96, 100% (v/v); 30 s each) and xylene for 5 min. After air drying, sections were ready for capture.
Laser capture was performed on a PixCell II microscope (Arcturus) using high-sensitivity caps (CapSure HSLCM; Arcturus). Cells were captured using the 7.5-µm laser spot with 1-ms laser pulse with a low setting (80 mW), from four different areas of the brain section. Vessels and parenchymal cells were captured onto separate caps, the RNA isolated using columns as described in the RNeasy MicroKit (Qiagen), and eluted in 12 µl of RNase-free water. The integrity of the RNA samples was tested using Bioanalyzer Pico chips (Agilent) according to the manufacturer’s instructions. RNA was amplified according to a previously described protocol for degraded RNA (47). Reverse transcription proceeded as above, except that 10 µl of RNA was primed with 1 µl of random hexamers (Geneworks), before real-time PCR analyses were conducted as described above.
Dual-label in situ hybridization (ISH) and IHC
Paraffin-embedded sections (8 µm) were cut onto Superfrost Plus slides (Menzel-Glaser), incubated with 33P-labeled cRNA probes transcribed from linearized CXCL9, CXCL10, and CCL5 riboplasmids, and processed for combined ISH/IHC as described previously (48, 49). Sections were processed for IHC to detect astrocytes (rabbit anti-glial fibrillary acidic protein; DakoCytomation), microglia (biotinylated lectin from Lycopersicon esculentum; Sigma-Aldrich), neurons (mouse anti-neuron-specific nuclear protein; Chemicon International), and T cells (rabbit anti-CD3; DakoCytomation). Bound Ab was detected using a Vectastain ABC kit (Vector Laboratories), and diaminobenzidine/H2O2 reagent (Vector Laboratories) was used as the immunoperoxidase substrate as described previously (48, 49).
Statistical analyses
Statistical analyses for gene expression data were performed using the Kruskal-Wallis test. Where statistical differences were observed, probabilities were then calculated using the Mann-Whitney U test on selected comparisons. Statistical analyses for flow cytometry data were performed using a one-way ANOVA with Tukey’s post test. These analyses were conducted using GraphPad Prism version 4.00 for Windows (GraphPad Software).
Photography
Photographs of brain sections were taken using an Olympus BX51 microscope and DP70 camera.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The expression of all known murine chemokine and chemokine receptor mRNAs was examined in the brains of CBA/T6 mice inoculated with PbA or PbK. Many mRNAs for chemokines and their receptors classified as "inflammatory" were found to be strongly up-regulated during PbA infection (FMCM), but not PbK infection (non-NCM (NCM)), while there were no striking differences in chemokines classified as "homeostatic" (Table II). This is supportive of a generalized inflammatory reaction within the brain during FMCM.
|
|
Using a combination of LCM and ISH/IHC techniques, the expression of certain chemokine and chemokine receptor RNAs was localized to the cerebral microvessels during PbA infection (Table III, Figs. 2 and 3). Using LCM, endothelial enrichment of vessel mRNA was confirmed using cell specific markers for endothelial cells, astrocytes, microglial and neuronal cells (data not shown), although some degree of contamination by interdigitating astrocyte foot processes and adherent leukocytes cannot be discounted. Parenchymal mRNA consisted of neuronal and glial cells. The mRNAs for the chemokine receptors CCR2 and CCR5 as well as the chemokines CXCL9, CXCL10, and CCL5 were found to be more strongly expressed in the vessels of PbA-infected WT mice compared with the parenchymal cells (Table III). However, while CCL5 and CCR2 mRNA were not detectable in the parenchyma, mRNA transcripts for CXCL9, CXCL10, and CCR5 were found in parenchymal cells, albeit in lower amounts, suggesting that cells within the brain parenchyma are capable of expressing these chemokines and chemokine receptors. CXCR3 mRNA was not detectable in vessel or parenchymal cell populations, possibly due to the low levels of leukocytes captured (data not shown).
|
|
|
To detect the cellular sources of CXCL9, CXCL10, and CCL5 gene expression, the LCM data were complemented by combining ISH and IHC, which revealed that, in addition to endothelial cells, glia were a predominant source of CXCL9 and CXCL10 RNA during PbA infection. In PbA-infected WT mice, CXCL9 RNA colocalized with microglia that were in close proximity to small vessels (Fig. 3A); both ramified and highly activated microglial cells were found to be CXCL9 RNA positive. In addition, a few intravascular, but not parenchymal, lymphocytes were found to express CXCL9 RNA (Fig. 3A, inlay). Astrocytes and neurons did not express CXCL9 RNA at a detectable level (data not shown). In PbA-infected CXCR3–/– mice, the same CXCL9 RNA expression pattern (i.e., in vessels and in microglia) was observed in mice that either developed FMCM, or were protected. Overall, when compared with PbA-infected WT mice, the expression appeared to be higher in mice that succumbed to FMCM (PbA KO A) (Fig. 3D), but lower in protected mice (PbA KO B) (Fig. 3G). In PbA-infected WT mice, vessels also showed prominent localization of CXCL10 RNA (Fig. 3B) but, in contrast to CXCL9, CXCL10 RNA was observed in astrocytes. These CXCL10 RNA-positive astrocytes were often found in clusters orientated around CXCL10 RNA-positive vessels (Fig. 3B). Less frequently, CXCL10 RNA-positive astrocyte clusters were found without a central CXCL10 RNA-positive vessel. Microglial cells, neurons, and CD3+ lymphocytes did not express a detectable amount of CXCL10 RNA (data not shown). Similar to CXCL9, the distribution pattern of CXCL10 in PbA-infected CXCR3–/– mice was unchanged compared with PbA-infected WT mice and, again, the apparent level of expression was higher in mice with FMCM (PbA KO A) (Fig. 3E) and lower in protected mice (PbA KO B) (Fig. 3H). In striking contrast to CXCL9 and CXCL10 RNA expression during PbA infection, CCL5 RNA was predominantly localized in intravascular and parenchymal CD3+ lymphocytes, whereas endothelial cells displayed a lower CCL5 expression (Fig. 3C). The accumulation of intravascular CCL5 RNA-positive lymphocytes was observed in PbA-infected WT and CXCR3–/– mice with FMCM (PbA KO A) (Fig. 3, C and F). In PbA-infected CXCR3–/– mice that did not develop FMCM (PbA KO B), only a few parenchymal CCL5 RNA-positive lymphocytes could be detected (Fig. 3I).
The majority of CXCR3–/– mice are protected from FMCM
Survival rates were followed in C57BL/6 (WT) and CXCR3–/– (KO) mice after inoculation with PbA. Although 65% of WT mice developed FMCM by day 6 p.i., and the remainder succumbed to FMCM by day 8 p.i., only 19% of CXCR3–/– mice developed FMCM by day 6 p.i., with a total of 32% CXCR3–/– mice progressing to FMCM by day 8 p.i. The remaining 68% of CXCR3–/– mice did not develop FMCM, succumbing instead to anemia and hyperparasitemia from approximately day 14 p.i., the majority without cerebral symptoms (Fig. 4A). This was not due to different parasite kinetics, as both WT and CXCR3–/– mice showed similar parasite burdens (Fig. 4B). Furthermore, on close monitoring of clinical signs, it was observed that a small but significant proportion of CXCR3–/– mice were able to recover from cerebral manifestations including hind limb paralysis and seizures, before succumbing to anemia and high levels of parasitemia typical of noncerebral forms of malaria (Table IV). For further analyses, PbA-infected CXCR3–/– (KO) mice were then separated into the following groups: PbA KO A (mice with FMCM, euthanized day 6 p.i.), PbA KO B (asymptomatic mice, euthanized day 6 p.i.), PbA KO C (mice recovering from cerebral signs, euthanized day 8 p.i.), PbA KO D (asymptomatic mice, euthanized day 14 p.i.), PbA KO E (mice fully recovered from cerebral signs, euthanized day 14 p.i.).
|
|
To assess the degree of cerebral involvement, histopathological parameters were examined in the brains of WT and CXCR3–/– mice inoculated with PbA. In a blinded study, CXCR3–/– mice that developed FMCM were found to show typical histopathological features of FMCM such as hemorrhages, edema, and monocyte adherence, while these were absent in CXCR3–/– mice that did not develop FMCM, nor were they observed in CXCR3–/– mice that fully recovered from cerebral signs (data not shown). However, some degree of brain pathology was evident in the group of mice that were recovering from the cerebral symptoms, similar to a previously described resolving model of murine malaria (38). This also is observed during PbA infection of mice deficient in either Fas or FasL (17).
Diffuse fibrinogen immunostaining within the parenchyma, used to demonstrate increased permeability of the BBB to protein (50), was observed only in mice that progressed to FMCM. In the brains of mice protected from FMCM, fibrinogen staining was evident but restricted within the microvasculature, suggesting that endothelial activation was present without compromise of the BBB (data not shown).
Expression of CXCR3 on brain-sequestered leukocytes during FMCM
The expression of CXCR3 on BSL subpopulations in both uninfected and PbA-infected WT mice was examined by flow cytometry (Fig. 5). A total of 5–10% of CD4+ and CD8+ T cells expressed CXCR3, but there were no obvious differences in the percentages of CXCR3+ cells between uninfected and PbA-infected WT mice. However, total numbers of CXCR3+ cells increased in the brain 6 days following PbA infection as a result of the striking increase in the numbers of these cells (see below). In contrast, very few NK cells expressed CXCR3 in uninfected or PbA-infected WT mice, whereas a relatively high proportion of NKT cells expressed CXCR3. Again, there were no obvious changes in the expression levels of CXCR3 on NK or NKT cells between uninfected and PbA-infected mice.
|
To investigate the role of different subpopulations of BSLs in CM, these were isolated from the brains of PbA-infected WT and CXCR3–/– mice that either displayed signs of CM (PbA KO A), or were asymptomatic at this time (PbA KO B). Numbers of sequestered T cells, NK cells, NKT cells, and myeloid lineage cells were determined by flow cytometry (Fig. 6). All cell subtypes increased in number in the brains of PbA-infected mice compared with uninfected control mice. As has been previously reported (15, 21, 22), there was an 70-fold increase in the number of CD8+ T cells in the brains of PbA-infected WT mice (2.1 x 105 ± 2.8 x 104) compared with uninfected control mice. A substantial increase in the number of CD8+ T cells also was observed in the brains of PbA-infected CXCR3–/– mice compared with uninfected controls, but the absolute numbers of CD8+ T cells in the brains of these mice were significantly lower (50% less) than in WT mice (Fig. 6). There was no significant difference in the numbers of brain-sequestered CD8+ T cells in CXCR3–/– mice categorized as symptomatic or asymptomatic at day 6 p.i. There also was 9-fold increase in the number of CD4+ T cells in the brains of all mice 6 days following PbA inoculation, although the total number of CD4+ T cells per brain was 10-fold less than that of CD8+ T cells (2.2 x 104 ± 2.3 x 103; PbA-infected WT). The numbers of CD4+ T cells also were decreased (50% less) in PbA-infected CXCR3–/– mice compared with PbA-infected WT mice, and there were no significant differences in CD4+ T cell numbers between asymptomatic and symptomatic CXCR3–/– mice (Fig. 6). Similar to T cell populations, the numbers of brain-sequestered NK and NKT cells were raised in PbA-infected mice compared with uninfected control mice (Fig. 6); however, there were no significant decreases in the numbers of sequestered cells between PbA-infected WT or CXCR3–/– mice.
|
5-fold in the brains of PbA-infected WT and CXCR3–/– mice compared with uninfected control mice. There were no differences in the number of sequestered myeloid lineage cells between any of the PbA-infected groups (Fig. 6). Localization of CD8+ T cells by immunohistochemistry during FMCM
Immunohistochemical techniques were used to further assess the involvement of CD8+ T cells in FMCM. In WT and CXCR3–/– mice that succumbed to FMCM, there was significant localization of CD8+ cells within brain vessels. These intravascular CD8+ cells were found within the meningeal vessels and ventricles, and also were sequestered deep within vessels of the cerebrum and cerebellum (Fig. 7A). In CXCR3–/– mice that did not develop FMCM, the presence of CD8+ cells was limited to the meningeal vessels and the ventricles (data not shown), and singular cells were observed in and around the vessels without massive accumulation within the microvasculature (Fig. 7B), the pattern observed in FMCM cases (18). There was no staining in the uninfected controls (data not shown). Thus, the localization and position of these CD8+ cells appear to be important for the development of FMCM.
|
The requirement of CXCR3+ CD8+ cells for the development of FMCM was emphasized when CD8+ cells or CD8– cells were adoptively transferred from WT mice with FMCM into CXCR3–/– mice that were then inoculated with PbA (Table V). This led to the induction of FMCM in almost all (15 of 22) CXCR3–/– mice that previously had shown a high level of protection (40–70%) (Fig. 4A and Table V). In mice without supplementation of CD8+ cells, almost all PbA-infected WT mice (17 of 20), but only a fraction of PbA-infected CXCR3–/– mice (6 of 28), developed FMCM by day 7 p.i. The adoptive transfer of CD8– cells had no effect on the incidence of FMCM in either WT or CXCR3–/– mice.
|
Expression of genes implicated in the pathogenesis of FMCM was compared between WT and CXCR3–/– mice infected with PbA, with several genes found to be highly up-regulated in the brains of mice that succumbed to FMCM when compared with mice that were protected. Perforin mRNA (Fig. 8A) expression was more strongly up-regulated in FMCM (PbA WT; 18.6 ± 3.8-fold) than in protected mice, a pattern reflected in the expression of granzyme B mRNA (data not shown). Both Fas (data not shown) and FasL mRNAs (Fig. 8B) were slightly, but significantly, up-regulated in FMCM (PbA WT; 3.3 ± 0.5-fold) but not in mice protected from FMCM. IFN- mRNA (Fig. 8C) (PbA WT; 227.9 ± 51.4-fold) and LT- mRNA (Fig. 8D) (PbA WT; 5.9 ± 0.4-fold) were both strongly expressed in the brains of WT and CXCR3–/– mice that developed FMCM, while expression was significantly lower in protected mice. ICAM-1 mRNA (Fig. 8E) expression was up-regulated during FMCM (PbA WT; 10.7 ± 1.4-fold) and was similarly expressed in different groups of CXCR3–/– mice. mRNAs for CXCL10 (Fig. 8F) (PbA WT; 304.7 ± 38.5-fold), CXCL9, and CCL5 (data not shown) also were up-regulated in the brains of CXCR3–/– mice with and without FMCM.
|
To ascertain that protection from FMCM in PbA-infected CXCR3–/– mice was not simply a result of systemic changes in immune responsiveness, we investigated the induction of cytokine gene expression and the trafficking of immune cells into the spleen. There was no difference in splenic IFN- mRNA expression between PbA-infected WT and PbA-infected CXCR3–/– mice (symptomatic and asymptomatic) (Fig. 9). Moreover, flow cytometric analysis of leukocyte subsets showed similar patterns of recruitment to the spleen (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
RNA transcripts for CCR2, CCR5, CXCL9, CXCL10, and CCL5 were found to be localized to cells of the BBB during FMCM using LCM (Table III) and ISH (Fig. 3). Endothelial cells were a prominent source of CXCL9 and CXCL10 RNA but expressed CCL5 RNA at much lower levels. Chemokine gene expression was associated with particular cell types, with CXCL9 RNA also expressed by microglia, and CXCL10 RNA also expressed by astrocytes. In contrast, CCL5 RNA was found in lymphocytes. Overall, chemokine expression was limited to mice that developed FMCM, with little expression in PbA-infected CXCR3–/– mice that were protected, and no expression in PbK-infected mice (Fig. 2). These findings support the hypothesis that changes to the BBB occur during FMCM through activation of endothelial and glial cells, and that this contributes to the pathogenesis of FMCM (7).
Changes to the integrity of the BBB may occur through the actions of chemokines, which can cause alterations in tight junction proteins in endothelial cell cultures, with a loss of occludin, claudin-5, zonula occulens-1, and zonula occulens-2 staining (60, 61) leading to an increase in BBB permeability to protein, which is a feature of human and murine CM (10, 38, 62, 63). The expression of CCR2 and CCR5 mRNA in the vessels during FMCM (Table III) suggests that endothelial cells could be responsive to chemokines, which might affect the permeability of the BBB. This may occur in either an autocrine or a paracrine manner, because chemokines are expressed by endothelial cells as well as adjacent glia and leukocytes (Fig. 3).
Chemokines are produced by activated glia in various CNS disease states (64, 65), and are likely to be critical for their immune functions (66, 67). However, in FMCM, glial activation occurs early and is thought to contribute to the increased permeability of the BBB because this occurs well before the onset of symptoms (7). Both astrocytes and microglia become activated and lose their even distribution around cerebral microvessels (68, 69), and may contribute to FMCM through the production of proinflammatory mediators (70). The specific pattern of chemokine expression during FMCM (CXCL9 by microglia, CXCL10 by astrocytes) supports previous findings that these are differentially expressed by CNS cells during inflammatory diseases (71, 72, 73) and may not share redundant functions (74, 75, 76). In particular, CXCL10-expressing astrocytes could attract and retain T cells during FMCM, as has been described in multiple sclerosis lesions (77). Chemokine production by cells at the BBB could lead to inadvertent activation of, and damage to, neurons and other CNS cells. Neurons and glia can up-regulate chemokine receptor expression (65, 78, 79, 80) and nonimmune functions of chemokines are well-documented (81, 82, 83). Neuronal dysfunction could well be a consequence of chemokine production in the CNS during FMCM, particularly because the observed chemokine expression does not specifically recruit leukocytes into the CNS parenchyma (Fig. 7A) (38, 62). Rather, leukocytes remain sequestered within the vessels, possibly due to the intensity of chemokine expression at the cerebral microvessels, which can further concentrate the production of chemokines by leukocytes.
FMCM is predominantly, but not exclusively, associated with a Th1-type response, characterized by the sequestration of monocytes and T cells in cerebral microvessels (2, 7, 84), and chemokine expression at the BBB is responsible for this recruitment. Brain endothelial cells are an important source of chemokines, particularly during inflammatory situations (79, 85) where they are essential for the recruitment and transmigration of leukocytes (86). Chemokines can enhance the adhesion of monocytes (87) and CD4+ T cells (88) and, in particular, CXCR3 ligands can mediate the adhesion of activated T cells (89) and stimulate T cell proliferation and effector function (90). Therefore, it is likely that the endothelial expression of CXCL9, CXCL10, and, to a lesser extent, CCL5 during FMCM, together with the glial expression of CXCL9 and CXCL10, is important for the recruitment of monocytes and T cells to the brain microvasculature without subsequent transmigration, and these cells can themselves be an important source of chemokines, such as CCL5 (Fig. 3). T cells can be recruited through a number of chemokine receptors but Th1 cells are predominantly recruited via CXCR3 and CCR5 (23). In murine CM, CXCR3-mediated recruitment of T cells can be influenced by NK cells (91). Although there was not a significant increase in the percentage of CXCR3+ cells within the brain during FMCM (Fig. 5), absolute numbers of CXCR3+ cells localized there increased dramatically. A low percentage of localized CXCR3-expressing cells similarly has been reported during dengue encephalitis (92). This may result from fluctuating CXCR3 expression during T cell activation and migration (93), or from CXCR3 down-regulation following T cell activation (94).
Mice genetically deficient in CXCR3 were used to examine the role of CXCR3 in FMCM. CXCR3–/– mice showed increased survival compared with their WT counterparts, and this protection was associated with an absence of histopathology and breakdown of the BBB but systemic immune system changes were not observed (Fig. 9). PbA-infected CXCR3–/– mice showed a significant decrease in the absolute numbers of brain-sequestered CD8+ and CD4+ T cells compared with their WT counterparts (Fig. 6). However, these were not significantly different between CXCR3–/– mice that were showing signs of CM or were asymptomatic, though cellular localization may be critical as discussed below. CXCR3 was observed to be specific for the recruitment of T cells because PbA-infected CXCR3–/– mice did not show decreases in the numbers of NK cells, NKT cells, macrophages, or neutrophils (Fig. 6). Importantly, there was a predominance of CD8+ T cells in the brains of PbA-infected mice, supporting a critical role for cytotoxic CD8+ T cells in FMCM. These sequester within brain vessels at the onset of cerebral symptoms (11), and their importance in the pathogenesis of FMCM is indicated by previous studies in which mice depleted of, or deficient in, CD8+ T cells are protected from FMCM (3, 11, 95). Moreover, adoptive transfer of CD8+ cells from PbA-infected WT mice abolished the protection against FMCM seen in CXCR3–/– mice (Table V). A role for cytotoxic CD4+ T cells receiving "help" from CD8+ T cells cannot be completely ruled out but is much less likely given the evidence discussed above and the demonstration that CD4+ T cells only constitute a small fraction (10-fold less than CD8+ T cells) of BSLs.
IHC revealed some differences in the localization of CD8+ cells in the brains of PbA-infected CXCR3–/– mice that did not display signs of CM. Although increased numbers of CD8+ cells were found to sequester within cerebral microvessels during FMCM, in both WT and CXCR3–/– mice (Fig. 7A), CXCR3–/– mice that were protected from FMCM did not show massive accumulation of CD8+ cells within cerebral microvessels. Instead, these were limited to the meningeal vessels and ventricles. The localization and positioning of these CD8+ T cells appears to be important for FMCM, and similarly has been described in CXCR3–/– mice during intracerebral lymphocytic choriomeningitis virus infection (27). Thus, taken together, the correlation of CD8+ T cell sequestration with the development of FMCM suggests that this process is critical for the development of pathology, along with the local production of cytokines as discussed below.
Previous studies have found an essential role for CCR5 during FMCM, with a reported 80% (21) or 50% (19) survival of CCR5–/– mice, comparable to our observed 70% survival of CXCR3–/– mice. Protection in CCR5–/– mice was similarly associated with a decrease in the numbers of CD8+ T cells in the brain (21). It is conceivable that while deficiency in either receptor results in defective trafficking of CD8+ T cells to the cerebral microvasculature, this is insufficient to lead to complete protection against FMCM. Small numbers of leukocytes could still be recruited to the brain during FMCM despite CCR5 or CXCR3 deficiency, due to the expression of other chemokine receptors and in the presence of other chemokine gradients, and this may be sufficient to initiate an inflammatory cascade. We found that infiltrating lymphocytes were a major source of CCL5 RNA (Fig. 3) and, uniquely, appeared to be transporting CCL5 into the CNS. This may be a critical autocrine mechanism through which the inflammatory process is amplified, particularly because the majority of CCL5+ lymphocytes were sequestered within vessels. However, singular CCL5+ lymphocytes also were detected in the parenchyma and a small percentage (15%) of infiltrating T cells were found in the parenchyma of PbA-infected WT and PbA-infected CXCR3–/– mice that succumbed to FMCM. CCL5 has been localized to infiltrating lymphocytes in other CNS inflammatory diseases (71, 96, 97), and is most likely involved in the recruitment of other immune cells, including monocytes, during FMCM. In FMCM, CD8+ T cells are preferentially recruited and can secrete proinflammatory cytokines (15). Moreover, observations that deficiencies in either CCR2 (22) or CCR1 (our unpublished findings) do not protect against FMCM highlights the complex and highly redundant nature of the chemokine system. This can be contrasted with other inflammatory CNS diseases, such as experimental autoimmune encephalomyelitis, where CCR1 (98) and CCR2 (99) play important roles in the pathogenesis, but CXCR3 (100) and CCR5 (101) do not.
Induction of perforin (Fig. 8A), granzyme B, FasL (Fig. 8B), and Fas mRNA was seen in mice that developed FMCM compared with mice that were protected. We previously have shown that perforin (16, 18) and FasL/Fas (16, 17) are important factors in FMCM. The induction of these genes is likely to be a consequence of high levels of cytokines during FMCM, which are absent in mice without CM. Most strikingly, IFN- mRNA was strongly up-regulated in the brains of mice that developed FMCM, regardless of whether they were WT or CXCR3–/–, but not in CXCR3–/– mice that were protected (Fig. 8C). LT- mRNA expression followed a similar pattern (Fig. 8D), but was not as highly expressed. Both cytokines are critical in the pathogenesis of FMCM (3, 4, 102). The unchanged expression patterns of ICAM-1 (Fig. 8E), CXCL9, and CCL5 mRNA during PbA infection suggest that prolonged activation of the endothelium, leading to continued chemokine gradients, occurs despite CXCR3 deficiency. Comparable results were observed using ISH (Fig. 3).
In conclusion, we have shown that a range of inflammatory chemokine and chemokine receptor mRNAs were highly up-regulated in the brain during FMCM compared with NCM, and this was associated with the proinflammatory reaction typical of FMCM. Moreover, CXCR3 was involved in the pathogenesis of FMCM, and this seemed to occur primarily through the recruitment of leukocytes, particularly pathogenic CD8+ T cells to the brain microvasculature. Chemokine expression at the BBB appeared to contribute to leukocyte retention within cerebral microvessels, which are then able to damage endothelial cells. This leads to increased permeability of the BBB to circulating inflammatory mediators or malaria Ags and local induction of cytokine expression, as well as neuronal dysfunction, all of which are features of FMCM (2, 7).
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by grants from the National Health and Medical Research Council of Australia and the Sir Zelman Cowen Universities Fund (to N.H.H.), and National Institute of Health Grant NS044905 (to I.L.C.). J.M. was supported by an Australian Postgraduate Award. S.L.C. is a recipient of an Endeavour International Postgraduate Research Scholarship and was supported by an International Postgraduate Award. M.M. was a postdoctoral fellow from the Deutsche Forschungsgemeinschaft (Mu17-07/3-1) and also received support from the "Innovative Medical Research" fund, University of Münster. H.J.B. is a Rolf Edgar Lake Research Fellow of the Faculty of Medicine, University of Sydney. 
2 Address correspondence and reprint requests to Prof. Nicholas H. Hunt, Molecular Immunopathology Unit, Bosch Institute, Medical Foundation Building (K25), University of Sydney, Sydney, New South Wales, Australia. E-mail address: nhunt{at}med.usyd.edu.au
3 Abbreviations used in this paper: CM, cerebral malaria; FMCM, fatal murine CM; BBB, blood-brain barrier; WT, wild type; KO, knockout; PbA, Plasmodium berghei ANKA; PbK, Plasmodium berghei K173; p.i., postinoculation; IHC, immunohistochemistry; NHS, normal horse serum; PRBC, parasitized RBC; HPRT, hypoxanthine guanine phosphoribosyltransferase; BSL, brain-sequestered leukocyte; NCM, non-CM; LCM, laser capture microdissection; ISH, in situ hybridization; LT-, lymphotoxin-.
Received for publication July 31, 2006. Accepted for publication October 26, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
), not systemic tumor necrosis factor-(), is the principle mediator of murine cerebral malaria. J. Exp. Med. 195: 1371-1377. β CD8+ T cells in experimental cerebral malaria. J. Immunol. 169: 6369-6375. and interleukin-8 in patients suffering from acute Plasmodium falciparum malaria. Clin. Immunol. Immunopathol. 76: 32-36. [Medline], and MIP-1β) in vitro, and altered thermoregulation in vivo. J. Inflamm. 45: 85-96. [Medline] expression in the brain during fatal murine cerebral malaria: evidence for production by microglia and astrocytes. Am. J. Pathol. 150: 1473-1486. [Abstract]- and β-chemokines by cells of diverse lineage in the central nervous system of mice with relapses of chronic experimental autoimmune encephalomyelitis. Am. J. Pathol. 150: 617-630. [Abstract] and IP-10 are expressed in demyelinating brain lesions. Proc. Natl. Acad. Sci. USA 96: 6873-6878. . Acta Neuropathol. 103: 458-468. [Medline] with patterns of tissue expression that suggest nonredundant roles in vivo. J. Immunol. 157: 4511-4520. [Abstract]/CXC chemokine ligand 9 stimulates T lymphocyte proliferation and effector cytokine production. J. Immunol. 172: 7417-7424. chemoattractant (CXCL11). J. Immunol. 167: 7084-7093. ) production in CXCR3–/– mice with experimental autoimmune encephalomyelitis. J. Immunol. 176: 4399-4409. or its CCR5 receptor. Eur. J. Immunol. 30: 1410-1415. [Medline] can prevent experimental cerebral malaria and its associated overproduction of tumor necrosis factor. Proc. Natl. Acad. Sci. USA 86: 5572-5574. This article has been cited by other articles:
|
|
 |
|
  M. Miyakoda, D. Kimura, M. Yuda, Y. Chinzei, Y. Shibata, K. Honma, and K. Yui Malaria-Specific and Nonspecific Activation of CD8+ T Cells during Blood Stage of Plasmodium berghei Infection J. Immunol., July 15, 2008; 181(2): 1420 - 1428. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Randall, F. H. Amante, K. A. McSweeney, Y. Zhou, A. C. Stanley, A. Haque, M. K. Jones, G. R. Hill, G. M. Boyle, and C. R. Engwerda Common Strategies To Prevent and Modulate Experimental Cerebral Malaria in Mouse Strains with Different Susceptibilities Infect. Immun., July 1, 2008; 76(7): 3312 - 3320. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Miu, N. H. Hunt, and H. J. Ball Predominance of Interferon-Related Responses in the Brain during Murine Malaria, as Identified by Microarray Analysis Infect. Immun., May 1, 2008; 76(5): 1812 - 1824. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. S. V. Campanella, A. M. Tager, J. K. El Khoury, S. Y. Thomas, T. A. Abrazinski, L. A. Manice, R. A. Colvin, and A. D. Luster Chemokine receptor CXCR3 and its ligands CXCL9 and CXCL10 are required for the development of murine cerebral malaria PNAS, March 25, 2008; 105(12): 4814 - 4819. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
