Microglial Recruitment Activation and Proliferation in Response to Primary Demyelination

【摘要】  We have characterized the cellular response to demyelination/remyelination in the central nervous system using the toxin cuprizone, which causes reproducible demyelination in the corpus callosum. Microglia were distinguished from macrophages by relative CD45 expression (CD45dim) using flow cytometry. Their expansion occurred rapidly and substantially outnumbered infiltrating macrophages and T cells throughout the course of cuprizone treatment. We used bromodeoxyuridine incorporation and bone marrow chimeras to show that both proliferation and immigration from blood accounted for increased microglial numbers. Microglia adopted an activated phenotype during demyelination, up-regulating major histocompatibility class I and B7.2/CD86. A subpopulation of CD45dim-high microglia that expressed reduced levels of CD11b emerged during demyelination. These microglia expressed CD11c and were potent antigen-presenting cells in vitro. T cells were recruited to the demyelinated corpus callosum but did not appear to be activated. Our study highlights the role of microglia as a heterogeneous population of cells in primary demyelination, with the capacity to present antigen, proliferate, and migrate into demyelinated areas.
--------------------------------------------------------------------------------
Microglia, central nervous system (CNS)-resident borne marrow-derived cells, and infiltrating macrophages are thought to play a pivotal role in the response to demyelination and the pathogenesis of multiple sclerosis (MS).1 In the CNS of mice with experimental autoimmune encephalomyelitis (EAE), a widely used mouse model of MS, microglia proliferate and, similar to macrophages, up-regulate expression of molecules needed for efficient antigen presentation, such as major histocompatibility (MHC) class I and II, B7.2/CD86, B7.1/CD80, and CD40.1-5 Although CNS-infiltrating and/or -associated macrophages and dendritic cells have been considered to be more efficient antigen-presenting cells in vitro,6-8 microglia can also be effective as antigen-presenting cells after demyelination in mouse models of MS.2,3 In EAE, microglia were found to be activated in the CNS before the onset of symptoms and before macrophage and lymphocyte entry into the CNS.4 Furthermore, blocking microglial response in the CNS of mice with EAE led to amelioration of symptoms.9
Although MS, like EAE, may be initiated by autoreactive lymphocytes migrating from the periphery, lymphocyte recruitment and activation can also follow primary demyelination in the CNS.10-12 Therefore, it is important to understand the response of microglia, macrophages, and lymphocytes to primary demyelination in the CNS. To accomplish this, we have used a model of primary demyelination, induced in adult mice by the copper chelator cuprizone. Treatment with cuprizone, administered as a feed additive, induces demyelination in the corpus callosum of C57BL/6 mice, and withdrawal of the toxin allows remyelination.13-15 The model is considered immune-independent because RAGC/C mice show no difference in demyelination or remyelination.16 Cuprizone-induced demyelination also does not promote blood brain barrier break down.17-19
Previous immunohistochemical studies of cuprizone-treated mice demonstrated increased response of microglia and/or macrophages in the demyelinated corpus callosum.13,20-22 These studies were unable to differentiate between the two cell types because there is overlap of surface marker expression, and activated cells share morphology. However, microglia can be distinguished from macrophages by their differential expression of CD45.7,23,24 To differentiate microglia from macrophages on the basis of this distinction and to quantify the cell types from the demyelinated corpus callosum, we have used flow cytometry. We show that in response to cuprizone treatment, microglia, macrophages, and T cells accumulate within the demyelinated corpus callosum, with microglia constituting the majority among these cells. Our data indicate that during primary demyelination, microglial cells increase in number through proliferation as well as by recruitment of progenitors from blood. Furthermore, we identify a unique population of activated microglia during cuprizone treatment that expresses CD11c and has antigen-presenting capability.

【关键词】  microglial recruitment activation proliferation response demyelination

Materials and Methods

Cuprizone Treatment

Cuprizone chow was prepared by adding 0.2% cuprizone by weight (Sigma, Oakville, ON, Canada) to milled Purina chow (Harlan Teklad, Madison, WI). To induce demyelination, mice were fed cuprizone chow for up to 6 weeks. Remyelination was induced in mice that had been treated for 6 weeks with 0.2% cuprizone by placing the mice back on normal food for 3 weeks. Unless otherwise specified, 8-week-old female C57BL/6 mice, purchased from The Jackson Laboratory (Bar Harbor, ME), were used in these experiments. To investigate effects of CCL2-CCR2 signaling, CCL2-deficient mice (10 generations C57BL/6; The Jackson Laboratory25 ) and CCR2-deficient mice (F2 C57BL/6 x 129/Ola, bred in our facility26 ) were treated with cuprizone. B6129P F2/J mice purchased from The Jackson Laboratory were used as background controls in studies with CCR2-deficient mice. Mouse handling and experiments were performed in accordance with the guidelines of the Canadian Council on Animal Care as approved by the McGill University Animal Care Committee.

Bone Marrow Chimeras

Twenty 6- to 8-week-old C57BL/6 Ly5.1 female mice (CD45.2) were irradiated and reconstituted with bone marrow from congenic C57BL/6 Ly5.2 mice (CD45.1) (The Jackson Laboratory) as previously described.27 After 7 weeks, mice were placed onto a 0.2% cuprizone diet and sacrificed after 3 or 4.5 weeks of treatment. Chimeric mice that were not treated with cuprizone were also sacrificed at equivalent time points. Chimerism was assessed by flow cytometric analysis of CD45 isoform expression in blood. Flow cytometric analysis of blood showed that between 10 to 11.5 weeks after reconstitution only 3.9 ?? 0.3% (mean ?? SEM) of the circulating blood cells were host derived (CD45.2-positive), indicating successful reconstitution.

Immunohistochemical Analysis and Myelin Staining

Mice were anesthetized with sodium pentobarbital (Somnotol) (120 mg/kg) (MTC Pharmaceuticals, Cambridge, ON, Canada) by intraperitoneal injection in 150 µl of (0.9%) saline. Mice were then perfused intracardially with 20 ml of ice-cold phosphate-buffered saline (PBS). The brains were removed and fixed overnight in 4% paraformaldehyde (Sigma) followed by overnight incubation in 20% sucrose and then mounted in OCT (Canemco, Montreal, QC, Canada) and frozen in cryomolds (Canemco). Tissues were stored at C80??C and then sectioned on a cryostat (Microm; Scientific Instruments, Montreal, QC, Canada). Eight-µm coronal sections were taken between bregma 0.38 and 0.98 mm. Slides were stained with Luxol fast blue and cresyl violet to show myelin and nuclei. For immunohistochemistry, slides were fixed in 4% paraformaldehyde and then permeabilized in a PBS/1% Triton X-100 solution for 30 minutes followed by blocking in 20% fetal bovine serum (Invitrogen, Burlington, ON, Canada), 20% normal goat serum (Jackson ImmunoResearch, West Grove, PA), 80% rat anti-mouse FcR (24G2; American Type Culture Collection, Rockville, MD) hybridoma supernatant and 0.01% sodium azide for 1 hour. The primary antibodies used were as follows: monoclonal rat anti-mouse CD11b (clone 5C6; Serotec, Toronto, ON, Canada) at 1/150, polyclonal rabbit anti-mouse laminin at 1/50 (Cedarlane, Hornby, ON, Canada), and fluorescein isothiocyanate-conjugated monoclonal hamster anti-mouse CD11c (BD Pharmingen, Oakville, ON, Canada) at 1/100. Sections were then incubated with secondary antibodies (if appropriate) conjugated to either Alexa Fluor 555 or Alexa Fluor 488 (Invitrogen), stained with Hoechst 33258 stain (Invitrogen) and mounted with Prolong-Gold (Invitrogen). For all staining procedures, controls were performed with no primary antibodies, serum controls (for polyclonal antibodies), and isotype controls for monoclonal antibodies.

Luxol Fast Blue Quantification

To analyze demyelination in the corpus callosum, coronal sections stained with Luxol fast blue were analyzed using an Olympus BH-2 light microscope (Tokyo, Japan). The qualitative scoring scheme assigned 4 for an unmanipulated mouse (having normal myelin levels), whereas a fully demyelinated mouse was given a score of 0. Sections were scored by an independent individual who was blinded to the study design.

Flow Cytometry

For flow cytometry analysis, microdissected corpus callosa were dissociated by forcing through a 70-µm sieve (BD Pharmingen) with a syringe plunger. The cell suspension was then processed as previously described28 and stained with various combinations of antibodies or isotype controls, obtained from BD Pharmingen unless otherwise specified (Table 1) . Cells stained with biotinylated antibodies were incubated with streptavidin-fluorochrome conjugates (BD Pharmingen). Samples were run on a FACScan (BD Pharmingen) or a FACSCalibur (BD Pharmingen). Samples were gated using forward scatter/side scatter to exclude dead cells. Data were analyzed using Cell Quest software (BD Pharmingen).

Table 1. Antibodies Used for Flow Cytometry

T-Cell Proliferation

Eight-week-old female C57BL/6 mice were immunized on each side at the base of the tail with 100 µg of myelin oligodendrocyte glycoprotein (MOG) peptide 35-55 (Sheldon Biotechnology Center, Montreal, QC, Canada) emulsified in complete Freund??s adjuvant (Fisher, Montreal, QC, Canada) containing 500 µg of heat-killed Mycobacterium tuberculosis (H37RA; Difco, Detroit, MI). Mice were sacrificed on day 8 after immunization, and the lymph nodes were collected into ice-cold Hanks?? balanced salt solution. Lymph nodes were dissociated in Hanks?? balanced salt solution by forcing through a 70-µm sieve (BD Pharmingen). Negative selection of T cells was then performed using Dynal mouse T-cell-negative isolation according to the manufacturer??s specifications (Invitrogen). After negative selection, 6 x 105 T cells were cultured in flat-bottomed microwells in a 200-µl final volume with 50 µg/ml MOG p35-55 and 1.2 x 106 CD11c-positive or CD11c-negative microglia (CD45dim, CD11b-positive), which had been sorted (FACSAria; BD Pharmingen) from pooled microdissected corpus callosa of either 6-week cuprizone-treated or unmanipulated mice. T cells or microglia alone in MOG 35-55 or medium were cultured as controls. Cells were maintained at 37??C in 5% CO2 in a humidified atmosphere. Ten µl of thymidine (0.5 µC) (MP Biomedicals, Aurora, OH) was added to the cultures during the last 18 hours, and the cultures were harvested after 72 hours and incorporated radioactivity measured by scintillation counting.29

Microglia Preparation for Cell Sorting

Mice were anesthetized with Somnotol as described above and then perfused intracardially with 20 ml of ice-cold PBS. The corpus callosa were then dissected from each animal, pooled, and dissociated in minimal essential medium with penicillin and streptomycin. The cells were centrifuged at 423 x g for 10 minutes at 4??C. Percoll (37%; GE Healthcare, Baie d??Urfe, QC, Canada) was then layered over the cells, and the microglia were harvested as previously described.30 The cells were then incubated with fluorescence-activated cell sorting (FACS) block , and stained with the appropriate antibodies.

Bromodeoxyuridine (BrdU) Incorporation

Mice undergoing cuprizone treatment or unmanipulated mice received an intraperitoneal injection of 100 µl of BrdU (1 mg/ml) (Sigma) dissolved in sterile saline daily from days 22 to 27. Cuprizone treatment was continued until day 32 (4.5 weeks). Corpus callosa were dissected and analyzed by FACS for BrdU incorporation using a BrdU flow kit as per the manufacturer??s instructions (BD Pharmingen). A second group of mice received daily intraperitoneal injections of BrdU from day 27 of cuprizone treatment until sacrifice on day 32 (4.5 weeks).

Results

Cuprizone Treatment Increases Cellularity in the Demyelinated Corpus Callosum

Demyelination occurred in female C57BL/6 mice treated with 0.2% cuprizone, as previously reported.15,31,32 After 3 weeks, animals showed varying levels of demyelination, as assessed by Luxol fast blue staining, and demyelination was complete after 6 weeks (Figure 1, B and D) . Mice treated for 6 weeks with cuprizone followed by 3 weeks of feeding with normal food (6 + 3*) showed remyelination in the corpus callosum, with myelination scores returning to those of unmanipulated mice (Figure 1, C and D) . Cuprizone treatment has been reported to increase cellularity in the demyelinated corpus callosum, despite a loss of oligodendrocytes.13 We confirmed that cellularity was enhanced in the corpus callosum of cuprizone-treated mice and that cellularity decreased after remyelination (Figure 1, ACC) . To quantify changes in cellularity, we used flow cytometry. Cell suspensions from isolated corpus callosa were acquired and gated on forward scatter/side scatter to exclude dead cells and debris from these analyses. Significant increases in cell numbers were observed in the corpus callosum after 3, 4.5, and 6 weeks of cuprizone treatment (Figure 1E) . This represented an increase from 120,200 ?? 14,900 (mean ?? SEM) cells in unmanipulated corpus callosum to 235,570 ?? 19,560 at 3 weeks, which then increased to 532,140 ?? 93,890 at the time of peak response (4.5 weeks). Cell numbers dropped by half at 6 weeks and then approached basal levels during the remyelination phase. Subsequent analyses showed that 15 to 35% of these cells expressed CD45.

Figure 1. Cuprizone treatment induces demyelination and increases cellularity in the corpus callosum. Luxol fast blue staining of coronal brain sections from unmanipulated mice (A), 6-week cuprizone-treated mice (B), and mice treated with cuprizone for 6 weeks followed by 3 weeks of recovery (6 + 3* weeks) (C) revealed demyelination in 6-week treated mice and remyelination in recovered animals. D: Quantification of the Luxol fast blue stains showed that animals treated for 3 weeks with cuprizone had incomplete demyelination, whereas after 6 weeks of treatment, demyelination was more complete. Six weeks of cuprizone treatment followed by 3 weeks of recovery showed myelination scores returning to those found in unmanipulated animals. E: Flow cytometry of corpus callosa from mice during cuprizone treatment showed changes in the number of live-gated cells recovered, which peaked at 4.5 weeks. Significance of data was analyzed using a one-way analysis of variance with a Dunnett posthoc test. *P < 0.05, **P < 0.01.

Macrophage/Microglial Response in the Demyelinated Corpus Callosum

Relative CD45 levels can be used to distinguish infiltrating macrophages (CD11b-positive, CD45high) from parenchymal microglia (CD11b-positive, CD45dim).7,23,24 We evaluated the relative dynamics of the microglial and macrophage response by assessing the total numbers of CD11b-positive cells that expressed low or high levels of CD45, respectively, using flow cytometry.

The majority of CD11b-positive cells expressed low levels of CD45 (CD45dim), which identifies parenchymal microglia. The number of microglia increased during cuprizone treatment, peaking at 4.5 weeks when there were 109,590 ?? 6770 microglia per corpus callosum (Figure 2A) . Microglia comprised 99.5% of the CD11b-positive cells at 4.5 weeks. During recovery (6 + 3*), the number of microglia subsided to levels observed in unmanipulated mice.

Figure 2. Flow cytometric analysis of CD45- and CD11b-positive cells isolated from corpus callosa. Corpus callosa were microdissected from animals treated with cuprizone at 0, 1, 3, 4.5, and 6 weeks as well as from mice undergoing recovery (6 + 3*). Dissociated corpus callosa were analyzed by FACS to quantify the number of microglia and macrophages. Both microglia (A) and macrophages (B) significantly increased in number during cuprizone treatment compared with the number found in unmanipulated mice. Each data point represents the number of cells from one mouse. Data were analyzed using a one-way analysis of variance with a Dunnett posthoc test. **P < 0.01.

Very few (less than 100) CD11b-positive cells expressed the high levels of CD45 (CD45high) that identify blood-derived macrophages in the corpus callosum of unmanipulated mice (Figure 2B) . There was a significant increase in the number of macrophages by 4.5 weeks, when there were 520 ?? 120 macrophages in the demyelinated corpus callosum. The number of macrophages then declined by 6 weeks but remained elevated compared with untreated mice. Macrophage numbers decreased to lower levels during remyelination.

Cuprizone Treatment Induces Microglial Activation

At 4.5 weeks, the peak of microglial expansion, we found a significant increase in microglial expression of B7.2/CD86 and MHC class I (filled black histograms) compared with unmanipulated mice (gray histograms) (Figure 3, A and C) . At 4.5 weeks, the mean fluorescent intensity of B7.2/CD86 expression on microglia was 103.7 ?? 1.5 compared with 57.5 ?? 0.5 at 0 weeks. For MHC class I, the mean fluorescent intensity at 4.5 weeks was 240.5 ?? 15.3 compared with 73.5 ?? 4.6 at 0 weeks. Mean fluorescent intensity for B7.1/CD80 increased slightly but was not statistically significant (Figure 3B) . However, the proportions of microglia expressing B7.2/CD86, B7.1/CD80, and MHC class I were significantly greater after 4.5 weeks of cuprizone treatment than in unmanipulated mice (0 weeks) (Figure 3, DCF) . There was no change in basal MHC class II or CD40 expression, compared with that on microglial cells from unmanipulated mice (Figure 3, G and H) .

Figure 3. Expression of markers of activation on microglia at 4.5 weeks of cuprizone treatment. Microglia were analyzed for expression of activation markers using flow cytometry at either 0 weeks (gray histograms) or 4.5 weeks (black histograms) after cuprizone treatment. For analysis, samples were gated on CD11b-positive CD45dim cells for microglia. The mean fluorescence intensity was examined for B7.2/CD86 (A), B7.1/CD80 (B), MHC class I (C), MHC class II (G), and CD40 (H). Isotype controls for each histogram are shown in white. The percentage of microglia expressing each marker was also examined for B7.2/CD86 (D), B7.1/CD80 (E), and MHC class I (F). Significance of data was analyzed using a two-tailed unpaired Student??s t-test. *P < 0.05, ***P < 0.001.

A Subpopulation of Microglia Express CD11c

Throughout the course of cuprizone treatment, a subpopulation of microglia emerged that expressed CD11c (Figure 4A , black dots). The proportion of microglia expressing CD11c was significantly increased during cuprizone treatment, from 1.1 ?? 0.1% in controls to 4.9 ?? 0.8% at 3 weeks (Figure 4B) . At 4.5 weeks, the peak of total microglial expansion, 14.3 ?? 0.8% of microglia were CD11c-positive, whereas at 6 weeks 39.0 ?? 2.3% were positive. After 6 weeks of cuprizone treatment, the CD11c-positive microglia fell within a subpopulation of microglia that had significantly reduced CD11b expression and elevated CD45 expression (although remaining CD45dim), as compared with CD11c-negative microglia (Figure 4, A and C) . Some macrophages also expressed CD11c. A higher proportion of CD11c-positive microglia expressed either MHC class I, B7.2/CD86, or B7.1/CD80 at 4.5 weeks compared with CD11c-negative microglia at this time point (Figure 4D) . As with the total microglial population, MHC II and CD40 were not detectably elevated above basal levels on the CD11c-expressing microglia (data not shown).

Figure 4. A distinct population of microglia express CD11c. A: Throughout the course of cuprizone treatment, the proportion of CD11c-positive microglia increased in the corpus callosum. These cells formed a distinct cluster defined by CD45 and CD11b staining. FACS plots in A show only the CD45-positive (dim + high) cells. CD11c-positive cells are represented by black dots. B: The percentage of microglia expressing CD11c increased during cuprizone treatment with a maximum at 6 weeks. C: At 6 weeks of treatment, CD11c-positive microglia had lower CD11b levels and higher CD45 levels than CD11c-negative microglia. D: CD11c-positive microglia from cuprizone-treated mice at 4.5 weeks had significantly higher expression of MHC class I, B7.2/CD86, and B7.1/CD80 as compared with CD11c-negative microglia. Significance of data in B was analyzed using a one-way analysis of variance with a Dunnett posthoc test. Other panels were analyzed using a two-tailed unpaired Student??s t-test. *P < 0.05, **P < 0.01, ***P < 0.001.

Histological staining showed that CD11c-positive cells were detected throughout the corpus callosum of mice fed cuprizone for 6 weeks (Figure 5, A and B) . CD11c-positive cells were rarely detected in sections from untreated mice (Figure 5C) . CD11c-positive cells were restricted to the demyelinated corpus callosum and were not detected in other, nondemyelinated regions of the brain (data not shown). Immunohistochemistry for CD11b showed co-localization with CD11c-positive cells, confirming the flow cytometry results (Figure 5, DCF) . Double staining for laminin and CD11c showed that the CD11c-positive cells were within the parenchyma of the CNS and not associated with blood vessels (Figure 5G) .

Figure 5. CD11c-positive microglia accumulate in the corpus callosum after cuprizone treatment. Immunohistochemistry for CD11c (green) counterstained with Hoechst (blue) in coronal sections showed CD11c-positive cells in 6-week cuprizone-treated mice (A) but not in unmanipulated animals (C). B: Higher magnification of the CD11c-positive cells in the corpus callosum of a cuprizone-treated mouse. Double staining for CD11c (D, green) and CD11b (E, red) showing co-localization (yellow, F). Double staining for laminin (red) to visualize blood vessels (G) and CD11c (green) showed that the CD11c-positive cells are not blood vessel-associated. Original magnifications: x400 (A, C); x1000 (B).

Recruitment from the Blood and Proliferation Contribute to Microglial Expansion during Cuprizone Treatment

We created bone marrow chimeras to investigate the source of the microglia that accumulated during cuprizone-induced demyelination. Lethally irradiated C57BL/6 (CD45.2) mice were reconstituted with bone marrow from congenic C57BL/6 (CD45.1) mice. Bone marrow-derived CD45.1-positive cells were quantified in the corpus callosum of untreated mice or mice fed cuprizone for 3 or 4.5 weeks. A small increase in bone marrow-derived immigrant microglial cells occurred after 3 weeks when 2.5 ?? 1.0% of the CD45dim microglial population were CD45.1-positive (Figure 6A) . There was a significant increase by 4.5 weeks, with 29.2 ?? 4.8% of the CD45dim microglial population expressing CD45.1. Thus, immigration from blood contributed to the microglial population in the demyelinating corpus callosum.

Figure 6. Microglia increase in the corpus callosum during cuprizone treatment because of recruitment from the blood as well as proliferation. Flow cytometry was used to enumerate blood-derived (CD45.1-positive) microglia in corpus callosa from bone marrow chimeras treated with cuprizone. A: The percentage of CD45.1-positive microglia in untreated chimeras and in chimeras treated with cuprizone for 3 and 4.5 weeks. B: BrdU treatment of cuprizone-treated mice showed that at 4.5 weeks microglia were proliferating. Corpus callosa from cuprizone-treated mice injected with BrdU for 5 days between either 22 to 27 days or 27 to 32 days were isolated for flow cytometry. The graph shows the number of microglia that had incorporated BrdU at either 4.5 weeks of treatments or 0 weeks. Data were analyzed using a one-way analysis of variance with a Bonferroni posthoc test. **P < 0.01, ***P < 0.001.

Only a minor subset of CD11c-positive microglia expressed CD45.1 at 3 weeks (1.7 ?? 0.9%). By 4.5 weeks, a slightly greater proportion of CD11c-expressing microglia were CD45.1-positive (8.2 ?? 1.5%), indicating that they had recently immigrated from blood. In parallel stainings, most CD11c-positive microglia were found to express CD45.2 at 3 and 4.5 weeks (88.5 ?? 8.5% and 94.1 ?? 0.9%, respectively), consistent with their being resident microglial cells. By contrast, more than 70% of CD11c-positive CD45high macrophages were CD45.1-positive, indicating their recent immigration from blood. The need for down-regulation of CD45 in transition to CD45dim microglia as well as the relative paucity of CD11c-positive compared with CD11c-negative microglia leaves open that immigrant cells expressing this marker could be underestimated.

We also investigated whether proliferation contributed to microglial expansion between 3 and 4.5 weeks. Mice were injected with BrdU for 5 days starting at day 22 or day 27 of cuprizone treatment. Mice were then sacrificed at day 32 (4.5 weeks), and cells that had incorporated BrdU were detected by flow cytometry. Close to 8% of microglia from cuprizone-treated mice had incorporated BrdU at 4.5 weeks (7.5 ?? 1.1% and 7.7 ?? 2.1%, respectively) (Figure 6B) . A greater proportion (11.1 ?? 3.3%) of CD11c-positive microglia were BrdU-positive in mice treated days 22 to 27, whereas 6.4 ?? 1.4% of CD11c-positive microglia were BrdU-positive after treatment on days 22 to 32. The kinetics of the BrdU experiments probably underestimates total proliferation, and these data indicate that proliferation contributes significantly to microglial expansion in the demyelinated corpus callosum. Expansion of CD11c-positive microglia also involved proliferation.

CCR2 Contributes to the Increase in CD11c-Positive Microglia during Cuprizone Treatment

mRNA for a number of chemokines, including CCL2 was rapidly up-regulated in the CNS of mice treated with cuprizone (data not shown). It has been proposed that CCL2-CCR2 signaling recruits microglial progenitors to the CNS.33,34 Total numbers of microglial cells were similar in CCL2-deficient mice and C57BL/6 control mice at 3 and 6 weeks after cuprizone treatment, whereas macrophage numbers were significantly reduced (Figure 7, B and D) . This selective impairment of macrophage recruitment but not microglial expansion was also seen in CCR2-deficient mice. Indeed, macrophage recruitment was significantly blocked in CCR2-deficient mice after 3, 4.5, 6, and 6 + 3* weeks of cuprizone treatment, compared with background matched B6.129 controls (F2), whereas no significant changes in total microglial numbers were observed (Figure 7, A and C) . We confirmed that mice lacking CCR2 and B6.129 controls responded similarly to cuprizone by demyelination at 6 weeks (data not shown).

Figure 7. CCR2 contributes to the increase in CD11c-positive microglia during cuprizone treatment. Total microglial (A, B) and macrophage (C, D) numbers were assessed in CCR2-deficient mice, CCL2-deficient mice, and wild-type controls by FACS. C57BL/6 mice were used as controls in experiments with CCL2-deficient mice, whereas B6.129 mice (F2) were used to control for genetic background in studies with CCR2-deficient mice. E: CCR2C/C mice (open circles) had fewer CD11c-positive microglia at 4.5 weeks of cuprizone treatment and after recovery (6 + 3*) compared with B6.129 controls (filled circles) at the same time points. Significance of data at each time point was analyzed using a two-tailed unpaired Student??s t-test. *P < 0.05, **P < 0.01.

The fold increase in CD11c-expressing microglia was transiently reduced at 4.5 weeks of cuprizone treatment in CCR2C/C mice compared with B6.129 controls (Figure 7E) . This effect was no longer seen at 6 weeks of cuprizone treatment, but after 3 weeks of remyelination, CCR2-deficient mice had fewer CD11c-positive microglia. No changes in CD11c-expressing microglia were noted in CCL2-deficient mice at either 3 weeks or 6 weeks (data not shown). Therefore, chemokine signaling may also play a role in generation or maintenance of CD11c-positive microglia after 4.5 weeks of cuprizone treatment and during recovery.

CD11c-Positive Microglia Are Potent Antigen-Presenting Cells

Previous reports have indicated that CD11c-expressing cells in the CNS are potent antigen-presenting cells.8,35 To investigate the antigen-presenting capacity of the CD11c-positive microglia, we performed proliferation assays using FACS-sorted microglia as antigen-presenting cells. MOG-reactive T cells isolated from the lymph nodes of MOG/complete Freund??s adjuvant-primed mice were co-cultured either with CD11c-positive microglia that were sorted from pooled corpus callosa taken after 6 weeks of cuprizone treatment or with CD11c-negative microglia sorted from unmanipulated or cuprizone-treated mice. No T-cell responses were observed without addition of MOG peptide, indicating that there was no presentation of endogenously acquired MOG antigen. However, when co-cultured with CD11c-positive or CD11c-negative microglia and with MOG peptide, T-cell proliferation was induced (Figure 8) . CD11c-positive microglia stimulated approximately three times greater T-cell response than CD11c-negative cells, showing enhanced capacity for antigen presentation.

Figure 8. Antigen-presenting capacity of CD11c-positive microglia versus CD11c-negative microglia. Proliferative response of MOG-specific T cells to MOG p35-55 presented by FACS-sorted microglial subpopulations was measured by thymidine incorporation. Microglia were plated in a 1:2 ratio to T cells. Results are shown as mean stimulation index in response to CD11c-positive (white) and CD11c-negative (black) microglial cells from 6-week cuprizone-treated corpus callosum and to CD11c-negative microglia from unmanipulated corpus callosum (black). Significance of differences was analyzed using a one-way analysis of variance with a Bonferroni posthoc test. *P < 0.05, **P < 0.01.

T Cells Are Recruited to the Demyelinated Corpus Callosum

We assessed TCRß-positive T-cell entry to the corpus callosum by flow cytometry (Figure 9) . T-cell numbers increased during cuprizone treatment, with levels being highest after 6 weeks of cuprizone treatment when there were 1630 ?? 290 T cells. After 3 weeks of recovery, T-cell numbers declined to lower levels. Analysis of T-cell subsets demonstrated that the proportion of CD4 T cells steadily declined throughout cuprizone treatment. CD4-positive T cells accounted for 48.4 ?? 6.7% of the T-cell population at 3 weeks. This dropped to 41.4 ?? 3.2% by 4.5 weeks. By 6 weeks, the CD4/CD8 ratio showed a predominance of CD8 T cells with 63.5 ?? 7.7% of the TCRß-positive cells being CD8-positive. The activation markers CD44 and CD69 were not up-regulated on TCRß-positive T cells (data not shown). There were almost no B cells present in the corpus callosum at 6 weeks using FACS (data not shown).

Figure 9. T cells infiltrate the demyelinated corpus callosum. The T-cell response to demyelination and remyelination was measured in 0-, 1-, 3-, 4.5-, and 6-week treated mice as well as from mice treated for 6 weeks with cuprizone followed by 3 weeks of recovery (6 + 3*). T cells were identified as TCRß-positive CD45high cells. Data were analyzed using a one-way analysis of variance with a Dunnett posthoc test *P < 0.05, **P < 0.01.

Discussion

We demonstrate that in response to primary demyelination induced by the toxin cuprizone, microglia become activated, migrate to, and proliferate within the demyelinated corpus callosum. The microglial population became heterogeneous, and a CD11c-positive subpopulation emerged. T cells also infiltrated, and there was a minor contribution to gross cellularity from macrophages. These proinflammatory changes tended to resolve on withdrawal of toxin.

The microglial and macrophage response was maximal at 4.5 weeks, with fewer of either cell type at 1, 3, and 6 weeks of treatment as well as 3 weeks after cuprizone withdrawal. In contrast to an earlier report in which macrophages constituted up to 20% of the CD11b-positive population at 6 weeks,18 our data indicate that at 6 weeks macrophages constituted only 0.5% of the CD11b population. The aforementioned study used pooled whole brains for flow cytometry analysis, whereas our study used individual microdissected corpus callosa, and it may be that macrophages deriving from other brain regions contributed to their results. The macrophage response to cuprizone-induced demyelination therefore resembles the macrophage infiltration seen after axonal injury28,36 rather than the massive macrophage infiltration seen in MOG-induced EAE in C57BL/6 mice.1,29 The large increase in microglial numbers between 3 to 4.5 weeks of cuprizone treatment was in part attributable to recruitment of microglial progenitors from blood as well as from microglial proliferation. Both mechanisms of microglial expansion occur in the unmanipulated mouse CNS, albeit to a lesser extent.37 Between 3 and 4.5 weeks of treatment, 7.6% of the microglia in the corpus callosum had proliferated, which is greater than the rate of microglial proliferation reported in the CNS of mice with EAE35 but less than what was observed in the acutely injured CNS.38 The signals that initiate microglial proliferation are unknown but may involve innate signaling pathways, as we have observed in other systems.36 At 4.5 weeks, although the blood brain barrier remains intact during cuprizone treatment,17-19 30% of the microglia were recent immigrants from the blood. The environmental cues necessary to recruit microglial progenitors from the blood during demyelination are unclear but may include chemokines or cytokines that are produced after cuprizone treatment.39 Our data demonstrate that CCL2-CCR2 signaling is not the critical cue driving this response because microglial numbers were similar in knockout and wild-type animals during demyelination. Microglia originating from bone marrow precursors have been shown to engraft the CNS after injury to the adult mouse CNS.33,38,40,41 Here, we show that influx of microglia progenitors from blood to the corpus callosum is a major contributor to the increase in microglial numbers during primary demyelination.

At the peak of gliosis at 4.5 weeks, microglia from cuprizone-treated mice expressed increased levels of MHC class I, CD11c, and B7.2/CD86, indicating activation in response to cuprizone treatment. The expression of such activation-associated molecules by microglia is similar to the expression profile of microglia seen in other demyelinating models.1,2,4,7,42 Interestingly, microglia did not up-regulate MHC class II. This may reflect lack of specific cytokine production such as interferon-, which is known to up-regulate MHC class II expression on microglia.43-45 Activated microglia in mixed glial cultures have also been shown to down-regulate MHC class II expression while up-regulating levels of co-stimulatory molecules such as B7.2/CD86.46 Unlike Mana and colleagues47 who detected interferon- mRNA in microdissected cuprizone-demyelinated corpus callosum from C57BL/6 mice, we did not detect interferon- mRNA by real-time polymerase chain reaction (PCR) analysis at any time (not shown). Although basal MHC class II levels were not up-regulated on microglia, CD11c-positive and CD11c-negative microglia presented antigen to T cells isolated from mice immunized with MOG in complete Freund??s adjuvant, usually considered a MHC II-dependent response. This probably reflects that baseline MHC II levels were sufficient for T-cell recognition??very few MHC class II molecules are required to initiate T-cell activation.48,49

The presence of CD11c-expressing microglia in the CNS after primary demyelination complements studies from mice with EAE,4 in the ischemic CNS,50 as well as in canine distemper lesions42 and Toxoplasma encephalitis,51 in which a proportion of microglia were found to be CD11c-positive. In our study, CD11c-positive microglia expressed higher levels of MHC class I, B7.2/CD86, and B7.1/CD80 than CD11c-negative microglia, in accordance with the notion that CD11c-expressing cells in the CNS are potent antigen-presenting cells. We find that CD11c-expressing microglia are better antigen-presenting cells in vitro compared with CD11c-negative microglia. Previous studies have shown that CNS-associated CD45high CD11b-positive cells expressing CD11c were potent antigen-presenting cells,8,52 but these studies did not specifically examine the antigen-presenting capacity of CD11c-expressing microglia. Other studies examining the antigen-presenting capacity of CD11c-positive cells in the CNS examined CD11b-positive populations in the CNS regardless of their CD45 levels.35,51

The origin of the CD11c subpopulation of microglia is of interest. Migration of microglia expressing CD11c into the demyelinated corpus callosum from other areas of the CNS seems unlikely since there were few CD11c-positive cells outside the corpus callosum (data not shown). Nevertheless, our data do not exclude that CD11c-positive microglia arise through proliferative expansion of a small pre-existing pool, either in or entering the corpus callosum. Indeed, a significant proportion of the CD11c-positive cells in the corpus callosum had incorporated BrdU at 4.5 weeks, supporting that proliferation may contribute to the expansion of this subpopulation. Fischer and colleagues51 reported that CD11c-negative microglia from adult mice could be induced in vitro to differentiate into CD11c-positive microglia with potent antigen-presenting capability, in the absence of proliferation??these cells did not express the plasmacytoid dendritic cell markers CD8 or DEC205.35 It is possible that in our in vivo system, CD11c-negative microglia give rise to CD11c-positive cells, which then expand through proliferation. Our data also suggest that microglia arise by immigration from blood, as has been shown by others. CD11c has been shown to be expressed by microglia that had recently immigrated from the blood.34 In our system, only a minority of CD11c-positive microglia were recent immigrants from blood. We did find immigrating cells that were CD11c-positive, but most of these cells expressed high levels of CD45, consistent with their being macrophages and/or dendritic cells, not CD45dim microglial cells. Ponoromev and colleagues4 found that in the CNS of mice with EAE there were two different populations of CD11c-positive cells, one that was CNS resident and the other bone marrow-derived. Our results also suggest a role for chemokine signaling in the generation and/or maintenance of the CD11c-positive microglia. The fold increase in CD11c-expressing microglia was transiently reduced in CCR2C/C mice at 4.5 weeks as well as after 3 weeks of remyelination.

The contribution of activated microglia to demyelination and remyelination is unclear. Activated microglia may phagocytose debris and produce cytokines/growth factors that stimulate oligodendrocyte precursor cells but can also release cytotoxins such as reactive oxygen species that can damage oligodendrocytes. Pasquini and colleagues53 investigated the role of the microglial response by administering minocycline to mice undergoing cuprizone treatment. This decreased the number of microglia and prevented demyelination. The role of specific microglial subsets needs further examination. When microglia activated by aggregated ß-amyloid were induced to express CD11c, microglial-mediated inhibition of neurogenesis was counteracted in vitro.54 Recently immigrating microglia were found to be better phagocytes than resident microglial cells in a mouse model of Alzheimer??s disease.55

We consider it unlikely that the antigen-presenting capability that CD11c-positive microglia show in vitro plays any role in cuprizone demyelination. RAGC/C mice showed no differences in either the extent or kinetics of cuprizone-induced demyelination or remyelination,16 indicating no role for T cells. Consistent with this, we found that T cells that infiltrated the demyelinated corpus callosum did not show an activated phenotype. The conditions required for activation of T cells in this system remain to be identified.

The response to primary demyelination in the CNS induced by cuprizone involves the recruitment of T cells, macrophages, and microglia to the demyelinated tissue. The CNS microenvironment likely controls interactions between these cell types. Our study has implications for diseases like MS in which activated microglia and T cells are associated to demyelinating lesions.

Acknowledgements

We thank Lyne Bourbonnire and Rachel Wheeler for help with bone marrow chimeras and IFN- PCR (R.W.); Dr. Tanja Kuhlmann, Dr. Sylvie Fournier, and Dominique Martel for helpful discussions; Dr. Alan Peterson and France Bourdeau for assistance with the cuprizone protocol; Marie-H?lne Lacombe for assistance with cell sorting; and Lily Li for help with histology.

【参考文献】
  Raivich G, Banati R: Brain microglia and blood-derived macrophages: molecular profiles and functional roles in multiple sclerosis and animal models of autoimmune demyelinating disease. Brain Res Brain Res Rev 2004, 46:261-281

Juedes AE, Ruddle NH: Resident and infiltrating central nervous system APCs regulate the emergence and resolution of experimental autoimmune encephalomyelitis. J Immunol 2001, 166:5168-5175

Mack CL, Vanderlugt-Castaneda CL, Neville KL, Miller SD: Microglia are activated to become competent antigen presenting and effector cells in the inflammatory environment of the Theiler??s virus model of multiple sclerosis. J Neuroimmunol 2003, 144:68-79

Ponomarev ED, Shriver LP, Maresz K, Dittel BN: Microglial cell activation and proliferation precedes the onset of CNS autoimmunity. J Neurosci Res 2005, 81:374-389

Santambrogio L, Belyanskaya SL, Fischer FR, Cipriani B, Brosnan CF, Ricciardi-Castagnoli P, Stern LJ, Strominger JL, Riese R: Developmental plasticity of CNS microglia. Proc Natl Acad Sci USA 2001, 98:6295-6300

Carson MJ, Sutcliffe JG, Campbell IL: Microglia stimulate naive T-cell differentiation without stimulating T-cell proliferation. J Neurosci Res 1999, 55:127-134

Ford AL, Goodsall AL, Hickey WF, Sedgwick JD: Normal adult ramified microglia separated from other central nervous system macrophages by flow cytometric sorting. Phenotypic differences defined and direct ex vivo antigen presentation to myelin basic protein-reactive CD4+ T cells compared. J Immunol 1995, 154:4309-4321

McMahon EJ, Bailey SL, Castenada CV, Waldner H, Miller SD: Epitope spreading initiates in the CNS in two mouse models of multiple sclerosis. Nat Med 2005, 11:335-339

Heppner FL, Greter M, Marino D, Falsig J, Raivich G, Hovelmeyer N, Waisman A, Rulicke T, Prinz M, Priller J, Becher B, Aguzzi A: Experimental autoimmune encephalomyelitis repressed by microglial paralysis. Nat Med 2005, 11:146-152

Barnett MH, Prineas JW: Relapsing and remitting multiple sclerosis: pathology of the newly forming lesion. Ann Neurol 2004, 55:458-468

Lucchinetti CF, Bruck W, Rodriguez M, Lassmann H: Distinct patterns of multiple sclerosis pathology indicates heterogeneity on pathogenesis. Brain Pathol 1996, 6:259-274

Trapp BD: Pathogenesis of multiple sclerosis: the eyes only see what the mind is prepared to comprehend. Ann Neurol 2004, 55:455-457

Hiremath MM, Saito Y, Knapp GW, Ting JP, Suzuki K, Matsushima GK: Microglial/macrophage accumulation during cuprizone-induced demyelination in C57BL/6 mice. J Neuroimmunol 1998, 92:38-49

Morell P, Barrett CV, Mason JL, Toews AD, Hostettler JD, Knapp GW, Matsushima GK: Gene expression in brain during cuprizone-induced demyelination and remyelination. Mol Cell Neurosci 1998, 12:220-227

Matsushima GK, Morell P: The neurotoxicant, cuprizone, as a model to study demyelination and remyelination in the central nervous system. Brain Pathol 2001, 11:107-116

Arnett HA, Mason J, Marino M, Suzuki K, Matsushima GK, Ting JP: TNF alpha promotes proliferation of oligodendrocyte progenitors and remyelination. Nat Neurosci 2001, 4:1116-1122

Kondo A, Nakano T, Suzuki K: Blood-brain barrier permeability to horseradish peroxidase in twitcher and cuprizone-intoxicated mice. Brain Res 1987, 425:186-190

McMahon EJ, Suzuki K, Matsushima GK: Peripheral macrophage recruitment in cuprizone-induced CNS demyelination despite an intact blood-brain barrier. J Neuroimmunol 2002, 130:32-45

Bakker DA, Ludwin SK: Blood-brain barrier permeability during cuprizone-induced demyelination. Implications for the pathogenesis of immune-mediated demyelinating diseases. J Neurol Sci 1987, 78:125-137

Blakemore WF: Observations on oligodendrocyte degeneration, the resolution of status spongiosus and remyelination in cuprizone intoxication in mice. J Neurocytol 1972, 1:413-426

Blakemore WF: Demyelination of the superior cerebellar peduncle in the mouse induced by cuprizone. J Neurol Sci 1973, 20:63-72

Chen MK, Baidoo K, Verina T, Guilarte TR: Peripheral benzodiazepine receptor imaging in CNS demyelination: functional implications of anatomical and cellular localization. Brain 2004, 127:1379-1392

Renno T, Krakowski M, Piccirillo C, Lin JY, Owens T: TNF-alpha expression by resident microglia and infiltrating leukocytes in the central nervous system of mice with experimental allergic encephalomyelitis. Regulation by Th1 cytokines. J Immunol 1995, 154:944-953

Sedgwick JD, Schwender S, Imrich H, Dorries R, Butcher GW, ter Meulen V: Isolation and direct characterization of resident microglial cells from the normal and inflamed central nervous system. Proc Natl Acad Sci USA 1991, 88:7438-7442

Lu B, Rutledge BJ, Gu L, Fiorillo J, Lukacs NW, Kunkel SL, North R, Gerard C, Rollins BJ: Abnormalities in monocyte recruitment and cytokine expression in monocyte chemoattractant protein 1-deficient mice. J Exp Med 1998, 187:601-608

Kuziel WA, Morgan SJ, Dawson TC, Griffin S, Smithies O, Ley K, Maeda N: Severe reduction in leukocyte adhesion and monocyte extravasation in mice deficient in CC chemokine receptor 2. Proc Natl Acad Sci USA 1997, 94:12053-12058

Hassan-Zahraee M, Tran EH, Bourbonniere L, Owens T: Elevated interferon-gamma in CNS inflammatory disease: a potential complication for bone marrow reconstitution in MS. J Neuroimmunol 2000, 108:40-44

Babcock AA, Kuziel WA, Rivest S, Owens T: Chemokine expression by glial cells directs leukocytes to sites of axonal injury in the CNS. J Neurosci 2003, 23:7922-7930

Tran EH, Prince EN, Owens T: IFN-gamma shapes immune invasion of the central nervous system via regulation of chemokines. J Immunol 2000, 164:2759-2768

Zehntner SP, Bourbonniere L, Hassan-Zahraee M, Tran E, Owens T: Bone marrow-derived versus parenchymal sources of inducible nitric oxide synthase in experimental autoimmune encephalomyelitis. J Neuroimmunol 2004, 150:70-79

Stidworthy MF, Genoud S, Suter U, Mantei N, Franklin RJ: Quantifying the early stages of remyelination following cuprizone-induced demyelination. Brain Pathol 2003, 13:329-339

Selvaraju R, Bernasconi L, Losberger C, Graber P, Kadi L, Avellana-Adalid V, Picard-Riera N, Van Evercooren AB, Cirillo R, Kosco-Vilbois M, Feger G, Papoian R, Boschert U: Osteopontin is upregulated during in vivo demyelination and remyelination and enhances myelin formation in vitro. Mol Cell Neurosci 2004, 25:707-721

Bechmann I, Goldmann J, Kovac AD, Kwidzinski E, Simburger E, Naftolin F, Dirnagl U, Nitsch R, Priller J: Circulating monocytic cells infiltrate layers of anterograde axonal degeneration where they transform into microglia. FASEB J 2005, 19:647-649

Simard AR, Rivest S: Bone marrow stem cells have the ability to populate the entire central nervous system into fully differentiated parenchymal microglia. FASEB J 2004, 18:998-1000

Fischer HG, Reichmann G: Brain dendritic cells and macrophages/microglia in central nervous system inflammation. J Immunol 2001, 166:2717-2726

Babcock AA, Wirenfeldt M, Holm T, Nielsen HH, Dissing-Olesen L, Toft-Hansen H, Millward JM, Landmann R, Rivest S, Finsen B, Owens T: Toll-like receptor 2 signaling in response to brain injury: an innate bridge to neuroinflammation. J Neurosci 2006, 26:12826-12837

Lawson LJ, Perry VH, Gordon S: Turnover of resident microglia in the normal adult mouse brain. Neuroscience 1992, 48:405-415

Ladeby R, Wirenfeldt M, Dalmau I, Gregersen R, Garcia-Ovejero D, Babcock A, Owens T, Finsen B: Proliferating resident microglia express the stem cell antigen CD34 in response to acute neural injury. Glia 2005, 50:121-131

McMahon EJ, Cook DN, Suzuki K, Matsushima GK: Absence of macrophage-inflammatory protein-1alpha delays central nervous system demyelination in the presence of an intact blood-brain barrier. J Immunol 2001, 167:2964-2971

Priller J, Flugel A, Wehner T, Boentert M, Haas CA, Prinz M, Fernandez-Klett F, Prass K, Bechmann I, de Boer BA, Frotscher M, Kreutzberg GW, Persons DA, Dirnagl U: Targeting gene-modified hematopoietic cells to the central nervous system: use of green fluorescent protein uncovers microglial engraftment. Nat Med 2001, 7:1356-1361

Wirenfeldt M, Babcock AA, Ladeby R, Lambertsen KL, Dagnaes-Hansen F, Leslie RG, Owens T, Finsen B: Reactive microgliosis engages distinct responses by microglial subpopulations after minor central nervous system injury. J Neurosci Res 2005, 82:507-514

Stein VM, Czub M, Schreiner N, Moore PF, Vandevelde M, Zurbriggen A, Tipold A: Microglial cell activation in demyelinating canine distemper lesions. J Neuroimmunol 2004, 153:122-131

Dutta T, Spence A, Lampson LA: Robust ability of IFN-gamma to upregulate class II MHC antigen expression in tumor bearing rat brains. J Neurooncol 2003, 64:31-44

Frei K, Siepl C, Groscurth P, Bodmer S, Schwerdel C, Fontana A: Antigen presentation and tumor cytotoxicity by interferon-gamma-treated microglial cells. Eur J Immunol 1987, 17:1271-1278

Liew SC, Penfold PL, Provis JM, Madigan MC, Billson FA: Modulation of MHC class II expression in the absence of lymphocytic infiltrates in Alzheimer??s retinae. J Neuropathol Exp Neurol 1994, 53:150-157

Carson MJ, Reilly CR, Sutcliffe JG, Lo D: Mature microglia resemble immature antigen-presenting cells. Glia 1998, 22:72-85

Maña P, Linares D, Fordham S, Staykova M, Willenborg D: Deleterious role of IFNgamma in a toxic model of central nervous system demyelination. Am J Pathol 2006, 168:1464-1473

Demotz S, Grey HM, Sette A: The minimal number of class II MHC-antigen complexes needed for T cell activation. Science 1990, 249:1028-1030

Valitutti S, Muller S, Cella M, Padovan E, Lanzavecchia A: Serial triggering of many T-cell receptors by a few peptide-MHC complexes. Nature 1995, 375:148-151

Reichmann G, Schroeter M, Jander S, Fischer HG: Dendritic cells and dendritic-like microglia in focal cortical ischemia of the mouse brain. J Neuroimmunol 2002, 129:125-132

Fischer HG, Bonifas U, Reichmann G: Phenotype and functions of brain dendritic cells emerging during chronic infection of mice with Toxoplasma gondii. J Immunol 2000, 164:4826-4834

Greter M, Heppner FL, Lemos MP, Odermatt BM, Goebels N, Laufer T, Noelle RJ, Becher B: Dendritic cells permit immune invasion of the CNS in an animal model of multiple sclerosis. Nat Med 2005, 11:328-334

Pasquini LA, Calatayud CA, Bertone Una AL, Millet V, Pasquini JM, Soto EF: The neurotoxic effect of cuprizone on oligodendrocytes depends on the presence of pro-inflammatory cytokines secreted by microglia. Neurochem Res 2007, 32:279-292

Butovsky O, Koronyo-Hamaoui M, Kunis G, Ophir E, Landa G, Cohen H, Schwartz M: Glatiramer acetate fights against Alzheimer??s disease by inducing dendritic-like microglia expressing insulin-like growth factor 1. Proc Natl Acad Sci USA 2006, 103:11784-11789

Simard AR, Soulet D, Gowing G, Julien JP, Rivest S: Bone marrow-derived microglia play a critical role in restricting senile plaque formation in Alzheimer??s disease. Neuron 2006, 49:489-502

作者单位：From the Neuroimmunology Unit,* Montreal Neurological Institute, McGill University, Montreal, Canada; and the Medical Biotechnology Center, University of Southern Denmark, Odense, Denmark