| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* INSERM U433, IFR des Neurosciences de Lyon, Faculté de Médecine Laënnec, Lyon, France;
ENS/CNRS 5161, Laboratoires de Biologie Moléculaire de la Cellule, Lyon, France; and
Laboratoire d’Immunologie, Hôpital Neurologique, Lyon, France
2Correspondence: INSERM U433, Faculté de Médecine Laennec, 07 rue Guillaume Paradin, 69372 Lyon, France. E-mail: nataf{at}lyon.inserm.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: glial cell • bone marrow • stem cells • cell transfer
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. As such, microglia have been involved in the pathophysiology of numerous inflammatory or non-inflammatory CNS disorders such as multiple sclerosis, Alzheimer’s disease, and amyotrophic lateral sclerosis (3
4
5)
. The constant renewal of microglia by blood circulating cells is currently accepted, and bone marrow (BM) transfer experiments have allowed the hematopoietic origin of microglial cells to be formally established (6
, 7)
. Moreover, an accelerated turnover of microglia by blood circulating, BM-derived cells was shown to occur under several neuroinflammatory conditions, including experimental allergic encephalomyelitis (EAE) (8
, 9)
. However, the exact identity of these blood cells and the different steps allowing them to migrate and differentiate into microglia have not been determined. Compared with other tissue macrophages, microglia behave as immature myeloid cells displaying unique phenotype and functions such as a high proliferative potential (10
11
12)
. Also, previous studies have shown that, as opposed to most tissue resident macrophages, microglia are unlikely to be renewed through terminal differentiation of mature blood monocytes (13
, 14)
. Thus, microglia might derive from one or several specific cell population(s), distinct from mature monocytes and sharing common traits with immature myeloid cells. To identify such cells, we first analyzed microglia in vitro and in vivo under developmental conditions. The detection of CD34 and B220 antigens on a subset of microglial cells prompted us to examine the behavior of CD34+/B220+ cells under neuroinflammatory conditions. We thus assessed the presence of CD34+/B220+ cells in the blood and brain of mice suffering from EAE and studied the in vivo trafficking of labeled-myeloid progenitors during EAE. Finally, we evaluated the ability of purified CD34+/B220+ cells to differentiate in vitro into microglial-like cells. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Glial cell cultures
For mixed glial cell cultures, microdissected cortices obtained from newborn (24 h–48 h) C57Bl/6 mice (Charles River) were mechanically dissociated, then resuspended in Dulbecco’s modified Eagle essential medium glutamax (DMEM glutamax; Life Technologies, Gaithersburg, MD, USA) containing 25 mM glucose (Glc), supplemented with 20% heat-inactivated fetal calf serum (FCS) and penicillin-streptomycin (1 µg/ml). Cells were seeded in poly-L-lysine precoated 75 cm2 flasks, then incubated at 37°C in a moist 5% CO2-95% air atmosphere. Three days after plating, the medium was changed and the FCS concentration was decreased to 10%. Twelve days after plating, floating microglial cells were harvested and reseeded on uncoated plastic dishes for 30 min as described previously (15)
. Non-adherent cells were then eliminated by two washes with fresh PBS and microglial cells were grown further in medium consisting of 50% DMEM glutamax supplemented with 10% FCS and 50% glial cell conditioned medium. After 4 days culture, microglial cells were fixed with either ethanol or acetone for immunocytofluorescence experiments or fixed with glutaraldehyde for electron microscopy (EM) analysis. Alternatively, cells were detached by trypsin treatment for FACS analysis. For preparation of glial cell conditioned medium, the supernatant from 12-day-old mixed glial cell cultures was harvested, centrifuged at 720 g, sterile filtered, then stored at –20°C until use. In other cases, the same procedure was applied to 30-day-old mixed glial cell cultures whose medium had been replaced on days 12 and 20 after plating.
Generation of BM-derived myeloid progenitors
Two- to 5-month-old female C57Bl/6 mice (Charles River) were sacrificed by halothane inhalation and bone marrow cells were harvested by flushing out tibiae and femurs. Total bone marrow cells were then seeded on uncoated flasks at 5 x 105 cells/ml in Iscove’s modified Dulbecco’s medium (IMDM, Invitrogen, Carlsbad, CA, USA) supplemented with 15% fetal calf serum (Fetal Clone II, Hyclone, Logan, Utah), penicillin-streptomycin (1 µg/ml), and 10 ng/ml human macrophage-colony stimulating factor (hM-CSF, PreproTech Inc., London, UK). Cultures were incubated at 37°C in a moist 5% CO2-95% air atmosphere. Three days after plating, nonadherent cells were harvested and poorly adherent cells were recovered by flushing with fresh PBS. The cells collected were then washed once in PBS and analyzed by flow cytometry or replated for the generation of microglial-like cells. In some experiments, BM-derived myeloid progenitors were obtained by M-CSF stimulation of bone marrow cells derived from transgenic enhanced green fluorescent protein mice (16)
.
Differentiation into microglial-like cells
After replating, BM-derived myeloid progenitors were differentiated in medium consisting of 50% DMEM glutamax supplemented with 10% FCS and 50% glial cell conditioned medium (GCCM). After 4 days culture, adherent cells were then either fixed with ethanol or acetone for immunocytofluorescence experiments or fixed with glutaraldehyde for electron microscopic analysis. Alternatively, cells were detached by trypsin treatment for FACS analysis.
Experimental allergic encephalomyelitis (EAE)
For EAE induction, mice (n=44) were immunized with myelin oligodendrocyte glycoprotein (MOG) peptide 35–55 as described previously (17)
. Briefly, MOG peptide was synthesized by standard 9-fluorenyl-methoxy-carbonyl chemistry and shown to be 95% pure as determined by reversed phase-HPLC (Institut de Biologie et Chimie des Protéines, IBCP, Lyon, France). Mice were immunized on days 1 and 7 by s.c. injection of 150 µg peptide emulsified in complete Freund’s adjuvant. In addition, on days 0 and 2 post-immunization (p.i.), mice were given 500 ng pertussis toxin (Sigma, St. Louis, MO, USA) intraperitonally. Clinical scores were monitored daily using the following scale: 0: lack of clinical signs; 1: tail weakness or tail paralysis; 2: hind leg paraparesis, hemiparesis or ataxia; 3: hind leg paralysis or hemiparalysis; 4: complete paralysis; 5: moribund; 6: death. In EAE mice, clinical signs initially occurred on day 14 (14±2.4 days) p.i. and reached a first peak on day 22 (22±2.6 days) p.i. (mean clinical score: 3.9±0.6). Then, mice partially recovered and entered a chronic clinical phase starting on day 28 (28±3.1 days) p.i. (mean clinical score: 2.7±0.4). Results presented in this paper were obtained from EAE mice that were sampled or sacrificed during the chronic phase of the disease (between day 28 and day 78 p.i). For cell transfer experiments, EAE mice were injected intravenously (i.v.) on day 22 ± 2.6 days p.i. (first clinical peak) and sacrificed 10 days later for immunohistochemical or FACS analysis of brains and spinal cords.
Cell transfer experiments
In a first set of experiments, BM-derived myeloid progenitors were generated from wild-type (WT) C57/Bl6 mice and labeled with 5-(and-6)-carboxyfluorescein diacetate, succinimidyl ester (CFSE, Molecular Probes, Eugene, OR, USA) before transfer. Briefly, BM-derived myeloid progenitors were rinsed in PBS, centrifuged at 520 g for 5 min at room temperature, resuspended in 5 ml PBS, and incubated for 5 min at 37°C with 1 µM CFSE. Then 250 µl of FCS was added and cells were further incubated for 5 min at 37°C before being washed in PBS and resuspended at a dilution of 107 cells/100 µl in phenol red-free DMEM. EAE mice were then injected i.v. on day 22 ± 2 days p.i. (first clinical peak) with 5 to 7.106 labeled cells (n=17) or vehicle alone (n=13).
In a second set of experiments, BM-derived myeloid progenitors were generated from enhanced GFP (EGFP) transgenic mice that had been backcrossed for at least 10 generations into a C57/Bl6 background. In these mice, we repeatedly found that GFP was expressed by <50% of M-CSF expanded CD34+ myeloid progenitors (data not shown). Such a low rate of expression led us to inject 3-fold more myeloid progenitors (2.107 cells) than in transfer experiments of CFSE-labeled myeloid progenitors. Otherwise, timing of injections and sacrifice were identical.
For BM reconstitution experiments, 10- to 12-wk-old female C57Bl/6 mice (n=13) were exposed to a 10 gray total body irradiation, administered 3 h apart in two equal fractions, with a cobalt 60 source. Approximately 3 h later, irradiated animals were injected via the retro-orbital vein with 5 to 7 x 106 GFP+ BM-derived myeloid progenitors supplemented with 5 to 10 x 106 GFP– BM cells derived from WT C57Bl/6 mice.
Isolation of CNS-associated leukocytes
When needed, brain and spinal cord were dissected, homogenized in ice-cold PBS/2% FCS, and centrifuged at 460 g for 5 min at 4°C. The pellet was resuspended in 4 ml Percoll 70% and overlaid with 4 ml 37% Percoll and 4 ml 30% Percoll as described previously (18)
. The percoll gradient was centrifuged at 460 g for 20 min at 20°C, without braking, then cells were collected from the 37%/70% Percoll interface, washed once in PBS, and resuspended in PBS/2%FCS before flow cytometry analysis.
Flow cytometry
Cells were incubated for 30 min at 4°C with blocking CD16/CD32 monoclonal antibody (mAb) (BD PharMingen, San Diego, CA, USA). After washing, cells were incubated for 30 min at 4°C with the following conjugated mAbs: fluorescein isothiocyanate (FITC) -labeled anti-B220 (clone RA3–6B2), phycoerythrin (PE)-labeled anti-CD11b (clone M1/70), PE-labeled anti-CD45 (clone 30-F11), PE-labeled anti-CD86 (clone GL1), PE-labeled antic-Kit (clone 2B8), PE-labeled anti-Flt3 (clone A2F10.1), biotinylated or PE-conjugated anti-CD34 (clone RAM34), and biotinylated anti-MHCII (clone 2G9) (BD PharMingen). Cells were then washed once in PBS, incubated with streptavidin-Alexafluor (488 or 506, Molecular probes) when needed, and fixed with 1% paraformaldehyde before analysis by flow cytometry on a Coulter EPICS XL (Beckman Coulter Inc., Fullerton, CA, USA). Cell sorting was performed on a FACS Vantage TM SE. cytometer (Becton Dickinson, Franklin Lakes, NJ, USA), at the IFR 128 flow cytometry core facility (CERVI, Lyon, France).
Immunocytofluorescence, immunohistofluorescence, and immunohistochemistry
Cells cultured in LabTek chambers (Nalge Nunc International Inc., Rochester, NY, USA) were fixed in acetone for 5 min at room temperature for CD11b staining or in absolute ethanol for 10 min at room temperature for nestin or B220 staining. In other experiments, brain or spinal cord isolated from P14 mice were frozen on dry ice and immunohistofluorescence analysis of B220 expression was performed on ethanol-fixed 14 µm-thick horizontal brain cryostat sections. Brain sections or cells were rinsed three times in PBS and incubated for 30 min at room temperature with a blocking solution consisting of 4% BSA diluted in PBS and supplemented with 10% normal goat serum. Cells or brain sections were then incubated overnight at 4°C with a mouse anti-rat nestin antibody (Ab) (clone Rat 401, BD PharMingen), a rat anti-mouse CD11b Ab (clone M1/70, BD PharMingen), a PE-conjugated rat anti-mouse CD11b Ab (clone M1/70, BD PharMingen), a biotinylated rat anti-mouse IgD, a biotinylated rat anti-CD34 Ab (clone RAM34, BD PharMingen), a rat anti-mouse F4/80 Ab (clone), or a rat anti-mouse B220 Ab (clone RA3–6B2, BD PharMingen). Antibodies were diluted 1:100 (B220, CD11b, nestin), 1:50 (F4/80), or 1.25 (CD34, IgD) in blocking solution. Brain sections or cells were then rinsed three times in PBS and incubated with a biotinylated goat anti-mouse (Molecular Probes), a biotinylated goat anti-rat Ab (Molecular Probes) or an Alexafluor-conjugated goat anti-rat Ab (Molecular Probes) for 1 h at room temperature. After three washes, brain sections or cells were finally incubated with streptavidin-Alexafluor (488 or 506, Molecular probes) diluted 1:100 in PBS. 4',6'-Diamidino-2-phenylidole (DAPI), Roche, Nutley, NJ, USA) staining of nucleus was performed and slides were mounted using Fluoroprep (BioMérieux). In several experiments, sections were sequentially incubated with anti-B220 Ab, biotinylated goat anti-rat Ab, and an avidin-biotin-peroxidase complex (ABC reagent, Vectastain ABC) Elite kit, Vector Laboratories, Burlingame, CA, USA). Peroxydase activity was then visualized using diaminobenzidine (3,3'-diaminobenzidine), following the manufacturer’s instructions.
Scanning EM
Plated cells were sequentially fixed in situ with 2% glutaraldehyde in culture medium for 15 min, then in 2% glutaraldehyde-O.1M Na cacodylate/HCl pH 7.4, for 15 min, at room temperature. Cells were then postfixed in 1% osmium tetroxide-0.15 M Na cacodylate/HCl pH 7.4 for 30 mn at room temperature and dehydrated in graded ethanol and critical point dried. After gold sputtering, samples were examined on a JEOL JSM 5300 scanning electron microscope equipped with MEGAVIEW II camera and analysis software (Electron microscopy core facility, CECIL, Faculté Laennec, Lyon, France).
RT-polymerase chain reaction (RT-PCR)
RNA was isolated (RNA-NOW, Biogentex, Ozyme) from three independent cultures of microglia cells and reverse transcribed using Oligo-dT primers (reverse transcription kit, Promega, Charbonnières, France). Complementary DNA was amplified by polymerase chain reaction (PCR) using Nestin (NM 016701; 5'-GGGAGGATGGAGAATGGACT-3' and 5'-AATCTTCCCCTGAGGACCAG-3') and B220 (NM 011210; 5'-GGGTTGTTCTGTGCCTTGTTC-3' and 5'-CTTGCCTCCATCCACTTCAT-3')-specific primers. Samples were amplified for 30 cycles after an initial denaturation cycle for 7 min at 94°C; each amplification cycle consisted of denaturation for 1 min at 94°C, annealing for 1 min at 60°C, and extension for 1 min at 72°C. The last cycle was followed by an extension cycle of 10 min at 72°C. The quality of the cDNA was monitored using cyclophilin-specific primers (5'-CAAGACTGAATGGCTGGATGGC-3' and 5'-CTTCAGTGAGAGCAGAGATTA CAGG-3').
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and cultured for 4 days in the presence of glial cell conditioned medium. Cells were then analyzed by flow cytometry, immunocytofluorescence, or RT-PCR (Fig. 1
A, B). The purity of the microglial cell cultures was shown to reach >99% based on FACS and immunocytofluorescence assessment of CD11b expression (Fig. 1A
and data not shown) and on a phagocytic assay using fluorescent microspheres (data not shown). As reported for microglia, and as opposed to dendritic cells, MHC class II molecules were not detectable and CD86 level was low (Fig. 1A
). However, FACS and immunocytofluorescence analyzes showed that microglial cells constitutively express B220 antigen, an isoform of CD45 found predominantly on B cells but also on a subset of dendritic cells and their progenitors (upper right panel in Fig. 1A
and photomicrograph in Fig. 1B
) (19
, 20)
. The hematopoietic progenitor cell marker CD34 was detected by FACS on a subpopulation of microglial cells (lower right panel in Fig. 1A
). In contrast, c-kit and flt3, previously reported on myeloid progenitors, were not found on cultured microglia (Fig. 1A
). Finally, the intermediate filament and neural stem cell marker nestin was readily shown by immunocytofluorescence in 50% to 70% of native microglial cells (photomicrograph in Fig. 1B
). Results obtained by FACS analysis and immunocytofluorescence were confirmed at the mRNA level by RT-PCR (Fig. 1B
, right panel). Finally, when analyzing anterior brain sections from 2-day-old (P2) mice, we could detect ramified B220+ cells in the corpus callosum (Fig. 1C
, left and middle panels). B220+ cells harboring a rounded morphology were seen in the lateral ventricles and in periventricular areas (data not shown). In parallel experiments we did not observe cells expressing immunoglobulin D (IgD), a marker for mature B cells, in the developing brain (Fig. 1C
, right panel).
|
To better analyze the phenotype and fate of the B220+ cells detected in P2 brains, we performed ex vivo FACS analysis on CNS-associated leukocytes isolated from 2-wk-old (P14) (n=7) or adult mice (n=6) (Fig. 2
). When analyzing CNS-associated CD11b+ cells at P14, we observed that 92.46 ± 2.3% were B220+ and 23 ± 5.6% expressed CD34 (Fig. 2A, B
). Thus, in microglia isolated ex vivo from the P14 brain, a great majority of cells expresses B220 constitutively and a significant proportion of cells bears the CD34 antigen. When analyzing CD34+ leukocytes deriving from the P14 brain, B220 was detected on 97.03 ± 1% of cells while only 4.69 ± 1.2% of cells expressed CD11b. This suggests that the immature CNS hosts a population of CD34+/B220+ progenitors that do not express the myeloid marker CD11b. Contrasting with the situation observed in the immature brain, we found that <10% of CNS-associated leukocytes expressed B220 in the normal adult CNS (Fig. 2C
). In addition, preliminary data indicate that in the adult CNS, CD34+ antigen is found on only 10 to 12% of CD11b+ microglial cells (data not shown). Altogether, these results demonstrate that 1) microglial cells form a heterogeneous population with regard to B220 and CD34 expression; and 2) the immature brain hosts a subpopulation of CD34+/B220+/CD11b– progenitors.
|
CD34+ progenitors expand in the brain of EAE mice
Several studies have demonstrated that the renewal of microglia from blood-derived elements is accelerated under neuroinflammatory conditions, including EAE (8
, 9)
. Recent data from our laboratory showed that during the chronic phase of EAE (as defined in Materials and Methods), infiltration by macrophages/microglia increases over time (G. Androdias et al., unpublished report). We thus tested whether CD34 progenitors expanded in the brain and/or in the blood of EAE mice during the chronic phase of the disease. In EAE mice sacrificed on day 45 p.i., CD11b+ cells were found to massively invade CNS parenchyma, as reported (21)
. Using immunohistofluorescence or immunohistochemistry, we detected B220 antigen on a subpopulation of ramified parenchymal cells (Fig. 3
A, B, D–F). In parallel experiments, no IgD+ cells were noted in the CNS parenchyma of EAE mice (Fig. 3C
), indicating that parenchymal B220+ cells were not mature B cells. In addition, double staining experiments showed that these B220+ ramified cells co-expressed CD11b (Fig. 3D-F
). Finally, it has to be noticed that B220+ ramified cells were not demonstrated in the spinal cord parenchyma of normal mice (data not shown). Similar to B220, CD34 antigen was noted on a subpopulation of CD11b+ ramified cells, which localized in the CNS parenchyma of EAE but not control mice (data not shown). These data were confirmed when performing double staining experiments with CD34 and F4/80, another marker for macrophages/microglia (Fig. 3G-I
). Finally, we could detect a population of rounded CD34+/CD11b+ cells that localized in the meninges and ventricles of EAE but not control mice (Fig. 3J-L
and data not shown). Such a localization suggests that, during EAE, the cerebrospinal fluid might contain a subpopulation of CD34+ cells migrating from blood to brain.
|
To accurately assess the phenotype of CNS-infiltrating CD34+ cells, FACS analysis was then performed on leukocytes isolated from the CNS of EAE mice (n=5) (Fig. 4
). Triple-staining experiments showed that nearly 90% (89.2±1.75% cells) of CD34+ progenitors coexpressed CD11b and B220 (Fig. 4
, upper panels). In addition, a significant subpopulation of CD34+ cells (6.18±1.14%) expressed B220 but not CD11b (Fig. 4
, upper panels). Accordingly, the percentage of CD11b+ cells co-expressing CD34 and/or B220 reached 34.58 ± 4.83% (Fig. 4
, middle panels). However, 48.85 ± 3.7% of CD11b cells did not express CD34 nor B220 (Fig. 4
, middle panels). Finally, when gating on B220+ cells we found that <30% of cells (28.6±3.3%) presented a B220+/CD11b–/CD34– phenotype and could thus be identified as putative B-lymphocytes (Fig. 4
, lower panels). Altogether these results show that during EAE, CNS infiltration by CD11b+ cells is accompanied by a parallel increase in CNS-infiltrating CD34+ cells. Moreover, our data raise the possibility that CNS-infiltrating CD34+ progenitors might derive from blood. Based on these findings, we then assessed the behavior of blood CD34+ cells during EAE.
|
CD34+ myeloid progenitors expand in the blood of EAE mice and target the inflamed CNS
We assessed the expression of CD34+ antigen on peripheral blood mononuclear cells (PBMC) isolated from control or EAE mice. Although in control mice (n=13), CD34+ cells were inconstantly detected, we found increased percentages of CD34+ cells in the PBMC of EAE mice sampled during the chronic phase of the disease (n=13) (Fig. 5
A, B). Such an increase could also be seen when assessing the percentage of CD34+ cells in B220+ cells or the percentage of CD34+ cells in CD11b+ cells (Fig. 5A, B
). In EAE mice, triple staining experiments showed that a great majority of blood circulating CD34+ cells harbored a B220+/CD11b+ or a B220+/CD11b– phenotype (Fig. 5A, B
). Overall, our results suggested thus that BM-derived CD34+ microglial progenitors are mobilized in the blood of EAE mice and might target the inflamed CNS.
|
We thus tested whether CD34+ myeloid progenitors delivered exogenously could migrate from blood to CNS under neuroinflammatory conditions. For this purpose, we initiated M-CSF expanded BM cultures, allowing CD34+ myeloid progenitors, which comprised CD34+/B220+ cells, to be generated (data supplement, panel A). M-CSF expanded BM cultures were then labeled with CFSE and injected i.v. to EAE mice (EAE-injected mice, n=15). In parallel experiments, EAE mice received i.v. injections of vehicle alone (EAE control mice, n=5). The presence of CFSE+/CD34+ cells was then assessed in the CNS and spleen of diseased animals, 10 days after the injections. In EAE-injected mice but not in EAE controls, a significant percentage of CFSE+ cells (6±0.4% of cells) was detected in CNS-derived leukocytes (Fig. 6
A, B). The presence of CFSEhigh or CFSElow cells in the CNS of recipient mice suggested that CFSE had diluted as a result of cell proliferation (Fig. 6A
). However, additional studies are needed to establish that cell proliferation actually occurred. FACS analysis showed that CFSE+ cells expressed CD45, CD11b, B220, or CD34 antigen (Fig. 6A, B
). When comparing percentages of CFSE+ cells in CNS-derived leukocytes vs. splenocytes, we observed that CFSE+ cells had preferentially targeted the CNS (Fig. 6B
). In particular, we found that CFSE+/CD34+ cells could be readily detected in the CNS but not in the spleen of recipient mice (Fig. 6B
). These results suggest that, in EAE animals, i.v. injected CD34+ progenitors are able to migrate from blood to CNS and keep their immature CD34+ phenotype while targeting the CNS. To better analyze the morphology and location of these CNS targeting cells, EAE mice were injected i.v. with BM myeloid progenitors deriving from EGFP transgenic mice (n=3). Free-floating sections of the whole CNS were performed in these mice as well as in EAE mice having received vehicle alone (n=3). This approach allowed GFP+ cells to be detected in the spinal cord (20 to 25 cells per section) and cerebellum (10 to 15 cells per section) of EAE mice (Fig. 6C
and data not shown). Most GFP+ cells were found in parenchymal rather than perivascular locations. Moreover, a great majority of them exhibited morphological features reminiscent to those of reactive microglia. These include a rod-shaped cellular body or the presence of short ramifications (Fig. 6C
).
|
We then assessed whether BM CD34+ myeloid progenitors cells could similarly target the CNS of irradiated reconstituted mice. M-CSF expanded BM cultures derived from actin-enhanced GFP transgenic mice were injected i.v. into lethally irradiated mice (n=13) that were concomitantly reconstituted with GFP– whole bone marrow cells. When examining the CNS of recipient mice 2 months after cell transfer, we observed GFP+ ramified cells presenting as isolated or clustered cells in the brain parenchyma and choroid plexuses (Fig. 6D
, left panel). At this time, FACS analysis showed that the chimerism rate of CNS-associated CD11b+ cells was below 1%. In the brain of irradiated reconstituted mice, CD34+ cells could be readily detected by FACS analysis and 15 ± 6% of them coexpressed GFP (Fig. 6D
, middle panel). In contrast, <1% of CD34+ splenocytes from recipient mice coexpressed GFP (Fig. 6D
, right panel). These data suggest that in the bone marrow chimera paradigm, M-CSF expanded CD34+ progenitors are able to target the CNS and to maintain CD34 expression, at least temporarily, while differentiating into microglia.
Bone marrow-derived CD34+/B220+ cells differentiate into microglial-like cells in vitro
Based on our findings, we hypothesized that BM-derived CD34+ myeloid progenitors could be driven toward microglial cell differentiation when exposed to a neural environment. To test this hypothesis, we first cultured whole M-CSF-expanded BM cells containing CD34+ progenitors (data supplement, panel A) in the presence of a glial cell conditioned medium. Following this protocol, microglial-like cells (MG-like cells) could be generated which showed the previously described morphological features of native microglia (Fig. 7
A). These included 1) a ramified morphology in >10% of cells as defined by the presence of at least one process 3-fold longer than the cell body diameter (15
, 22)
, 2) the detection of multiple CD11b+ pseudopodia and filament-like structures (23
, 24)
, and 3) the presence of microspikes detectable by scanning EM (22)
. As a negative control, we examined M-CSF expanded BM-derived macrophages and did not observe such morphological features (right panels in Fig. 7A
). Also, nestin expression was shown on a subset of MG-like cells (5 to 10% of cells) as well as on native microglia, but not BM-derived macrophages (bottom panels in Fig. 7A
). Finally, by FACS analysis we observed that BM-derived MG-like cells displayed a comparable phenotype compared with native microglia: MHC class II–/B220low/CD86low/CD11b+ with a subpopulation of cells expressing CD34 (Fig. 7B
). The differentiation of M-CSF expanded BM cells into MG-like cells appears to be related at least in part to the clonal expansion and differentiation of a limited number of cells, as shown by analyzing individual colonies grown in methylcellulose (data supplement, panel B). Accordingly, such a differentiation process was accompanied by robust cell proliferation as assessed by 3H-thymidine incorporation (data supplement, panel C). These data prompted us to examine whether, in our culture system, CD34+/B220+ progenitors were particularly prone to differentiate into microglial-like cells (MG-like cells). For this purpose, M-CSF expanded BM cultures were cell-sorted on the basis of CD34 and/or B220 expression, cultured for 4 days in the presence of glial cell conditioned medium, then examined by immunocytofluorescence (Fig. 8
). In these experimental conditions, CD34+/B220+ cells were able to differentiate into MG-like cells, which showed a ramified morphology (19.5±2.9% of cells harboring at least one process 3-fold longer than the cell body diameter), extended filament-like structures (17.3±2.1% of cells) and expressed nestin (53.7±4.2% of cells) (Fig. 8B
). In contrast, when exposed to glial cell conditioned medium, purified CD34–/B220+ cells adhered poorly to plastic and did not differentiate into MG-like cells (data supplement, panel D). Altogether, these results show that in M-CSF expanded BM cultures, CD34+/B220+ progenitors but not CD34–/B220+ cells display differentiation potential toward microglia in vitro.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, there have been no reports so far of B220 expression by circulating monocytes or resident macrophages (25)
. However, in human adult brain, CDw75, another surface molecule initially regarded as being restricted to B cells, was documented in vivo on microglia (26)
. Moreover, recent studies have shown that macrophages and B cells share a common progenitor (27
, 28)
. Therefore, one may consider that finding B220 on murine micoglia actually reflects a lineage relationship between microglia and B cells. Alternatively, B220 may provide microglia with a specific activation pathway under developmental or neuroinflammatory conditions.
Our results also demonstrate that murine microglia form a heterogeneous cell population composed of a subpopulation of CD34+/B220+/CD11b+ cells. These data agree with a previous study demonstrating CD34 on a subset of microglia in the developing human brain (29)
. Similarly, CD34 was detected on proliferating resident microglia in a murine model of acute neural injury (30)
. These studies along with our work suggest that, under developmental or inflammatory conditions, CD34+/B220+/CD11b+ microglial cells may proliferate and give rise to CD34–/B220+/CD11b+ activated microglia. However, we also detected CD34 on a subset of CD11b– cells, which cannot be considered microglia per se since they do not express CD11b. These CD34+/B220+/CD11b– progenitors may be related to the CNS-associated CD11b– myeloid progenitors, which were previously evidenced using ex vivo cultures of embryonic or adult CNS cells (10)
. Therefore one may hypothesize that CD34+/B220+/CD11b– cells form a pool of primitive, BM-derived microglial progenitors, which can be readily detected in the developing or inflamed CNS.
Previous studies have shown that BM-derived CD34+ stem cells as well BM-derived CD34+ endothelial progenitors are able to target the CNS (31
, 32)
. Our work extends these data and suggests that CD34+ myeloid progenitors may similarly be able to migrate from blood to brain. Thus, 8 wk after BM transplantation in irradiated mice, M-CSF expanded CD34+ myeloid progenitors were detected in the CNS but not in the spleen of recipient mice. Also, in EAE animals injected with CFSE-labeled myeloid progenitors, CD34+/CFSE+ cells were found in the inflamed CNS but not in the spleen of recipient mice. In both experimental models, we cannot exclude the possibility that CNS-infiltrating CD34+ cells actually derive from blood circulating CD34– cells. However, our transfer experiments and the increased rates of CD34+ PBMC observed in EAE mice strongly suggest that, under neuroinflammatory conditions, CD34+ myeloid progenitors may actually cross the blood-brain barrier and invade the inflamed CNS. Indeed, as recently proposed for CNS tumors (31)
, the inflamed brain might provide ad hoc signals allowing BM-derived CD34+ progenitors to migrate from blood to brain and to maintain their immature phenotype, at least temporarily. Once in the CNS compartment, the fate of BM-derived CD34+ cells is probably dictated by a number of intrinsic and extrinsic factors. In EAE mice, as most CNS-infiltrating CD34+ cells show a CD34+/CD11b+/B220+ microglial-like phenotype, we can postulate that a majority of BM-derived CD34+ myeloid progenitors may follow a similar microglial differentiation pathway. In this pathophysiological scheme, BM would fuel the CNS with blood-derived CD34+/B220+ myeloid progenitors, which would differentiate into activated microglia and perpetuate CNS inflammation. Cell transfer experiments of purified CD34+/B220+ myeloid progenitors in EAE mice are needed to formally establish this point. Should these experiments be conclusive, one may consider designing therapeutic strategies aimed at modulating the migration of CD34+/B220+ myeloid progenitors from blood to brain in order to dampen chronic neuroinflammation. Alternatively, such CD34+ myeloid progenitors might be used as cellular vectors to target the lesioned CNS, as demonstrated in other models using CD34+ hematopoietic stem cells (7
, 33)
.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication December 12, 2005. Accepted for publication May 22, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Vuaillat, G. Androdias, N. Davoust, and S. Nataf About multiple sclerosis, natalizumab, and CD34+ hematopoietic progenitors Blood, July 1, 2008; 112(1): 208 - 209. [Full Text] [PDF]
|
|
|
|
 |
|
 J. Palazuelos, N. Davoust, B. Julien, E. Hatterer, T. Aguado, R. Mechoulam, C. Benito, J. Romero, A. Silva, M. Guzman, et al. The CB2 Cannabinoid Receptor Controls Myeloid Progenitor Trafficking: INVOLVEMENT IN THE PATHOGENESIS OF AN ANIMAL MODEL OF MULTIPLE SCLEROSIS J. Biol. Chem., May 9, 2008; 283(19): 13320 - 13329. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. P. Turrin, M.-M. Plante, M. Lessard, and S. Rivest Irradiation Does Not Compromise or Exacerbate the Innate Immune Response in the Brains of Mice That Were Transplanted with Bone Marrow Stem Cells Stem Cells, December 1, 2007; 25(12): 3165 - 3172. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||
