Background
Demyelination and axonal degeneration, the hallmarks of multiple sclerosis (MS), have long been associated with the central nervous system (CNS) inflammation [
1]. Chemokines are important mediators of the inflammatory response, and the chemokine-guided influx of inflammatory cells into the CNS is a major contributor to myelin damage and axonal degeneration. The inflammation observed in the CNS of MS patients and of mice with experimental autoimmune encephalomyelitis (EAE) is associated with increased expression of the C-X-C motif chemokine 12 (CXCL12) [
2‐
5]. As a strong chemoattractant for T- and B-lymphocytes and inflammatory monocytes, CXCL12 is regarded as a mediator of pathogenic inflammation in the CNS. Recently, however, the function of CXCL12 in the CNS of mice with EAE, and apparently also of MS patients, was shown to be more complex; depending on its interaction with receptors CXCR4 or CXCR7 and on its spatial and temporal expression, CXCL12 may be anti-inflammatory [
4], immunomodulatory [
6], and may also promote remyelination [
7‐
10].
CXCL12 is a key pleiotropic chemokine. In addition to its essential function in regulating hematopoiesis and in the patterning of the immune system, it plays an important role in the plasticity and patterning of the CNS [
7,
11,
12]. During embryonic development, CXCL12 is crucial for neural stem cells (NSCs) proliferation, survival and migration, intercellular communication, and neuronal axon guidance [
7,
11,
13,
14]. CXCL12 and its receptors, CXCR4 or CXCR7, are also expressed in the adult brain [
11,
15], where the CXCL12/CXCR4 axis plays a central role in regulating the survival, proliferation, maturation, and migration of NSCs in response to CNS insults, including neuroinflammation-associated disorders, suggesting that CXCL12/CXCR4 axis is important also in adult neuro/oligodendrogenesis following CNS damage [
13,
16‐
20].
Although expression patterns of CXCR4 and its CXCL12 ligand have been extensively investigated during CNS development and in healthy and injured adult CNS [
7,
10,
11,
21], few studies have analyzed functions of CXCL12 in the context of CNS demyelinating diseases such as EAE/MS [
7,
8,
22]. In relapsing/remitting MS, the most common form of the disease, relapse episodes are associated with demyelination and neuronal damage. Albeit limited, spontaneous remyelination may occur in MS [
23] and, more profoundly, in viral model of MS [
24]. Remyelination occurs via oligodendrocyte precursor cells (OPCs), which are derived primarily from NSCs that persist in the adult brain in neurogenic niches lining the subventricular zone of the lateral ventricles and the subgranular layer of the hippocampal dentate gyrus (DG). The NSCs are self-renewing and can differentiate to neuronal progenitor cells (NPCs) and to OPCs [
7,
11]. The OPCs migrate to demyelinated areas in the CNS where they differentiate into mature myelin-producing oligodendrocytes [
7,
25].
It has been proposed that CXCL12 via its CXCR4 receptor plays a role also in regulating the survival and migration of NSCs in response to CNS trauma, such as brain tumors, ischemia, and neuroinflammation-associated disorders [
13,
16‐
20], and in regulating the migration and survival of OPCs in vitro [
26,
27]. In vivo, the role of CXCL12 and its CXCR4 receptor in promoting the differentiation of OPCs and their remyelination has been demonstrated in mice following cuprizone-induced demyelination [
8] and ischemia-induced demyelination [
10]. Yet, the potential effect of CXCL12 on NPCs or OPCs during the clinical course of EAE, a model for MS, and its possible relevance to spontaneous recovery from the disease is yet to be elucidated.
In this study, analysis of the expression dynamics of CXCL12 in the CNS of mice during the progression of clinical EAE and following spontaneous recovery showed that CXCL12 expression, which increased with disease progression, did not decline to basal levels in the CNS of spontaneously recovered mice. In recovered mice, CXCL12 levels remained significantly elevated relative to those in naïve CNS, although CNS inflammation and CXCL12-producing inflammatory cells had been abolished during recovery. The post-inflammatory CXCL12 expression in the CNS of spontaneously recovered mice was associated with a significant increase in numbers of NPCs and OPCs at the DG and corpus callosum (CC), respectively, that expressed the CXCL12. Notably, a significant proportion of the NPCs and OPCs co-expressed CXCL12 and its receptor, CXCR4. The significant increase in numbers of CXCL12+ CXCR4+ NPCs and OPCs, which have an intrinsic potential to proceed towards differentiation to mature neuronal cells or oligodendrocytes, likely via autocrine signaling mechanisms, links the post-inflammatory expression of CXCL12 with post-CNS-inflammation elevation of endogenous myelin/neuronal repair capacity associated with spontaneous recovery from clinical EAE.
Methods
Reagents and antibodies
The mouse PLP139-151 peptide was synthesized in the laboratory of Prof. M. Fridkin (Department of Organic Chemistry, Weizmann Institute of Science), using the Fmoc technique with an automated peptide synthesizer (AMS422; ABIMED, Langenfeld, Germany).
Antibodies used for immunostaining included the following: mouse anti- CXCL12 (clone K15C kindly provided by Prof. Tzvi Lapidot; Weizmann Institute, Israel); rabbit anti-CD3 (NeoMarkers, Fremont, CA); rat anti-mouse MAC2 (Biolegend); monoclonal mouse anti-GFAP (Sigma); rabbit anti-GFAP (Dako); rat anti-MBP (Chemicon); rabbit anti-NF200 (Sigma); goat anti-DCX (Chemicon); monoclonal mouse anti-β-Tubulin III (Chemicon); rabbit anti NG2 (Chemicon); and rat anti-CXCR4 (eBioscience).
The chemokine receptor antagonist, bicyclam AMD3100, was purchased from Sigma-Aldrich (St. Louis, MO, USA).
Animals
C57Bl/6J and SJL/J mice were purchased from Harlan (Jerusalem, Israel), and (C57Bl/6J × SJL/J) F1 mice were bred and maintained at the Weizmann Institute Animal Facility under specific pathogen-free conditions. All animal procedures and experimental protocols were approved by the IACUC of the Weizmann Institute (permit number 02820711-3) and were performed in compliance with its relevant guidelines and regulations.
Induction of EAE and tissue processing
For our experiments, PLP/EAE in (C57BL × SJL/J) F1 mice was found to be more suitable model than PLP/EAE in SJL/J mice. In our animal facilities, the disease in SJL/J mice was more severe (and with morbidity) than the disease in the F1 mice. PLP/EAE in (C57BL × SJL/J) F1 mice was with less morbidity and with more mice that recover from the disease spontaneously, mice that were necessary for the experiments in this study. (C57BL × SJL/J) F1 mice (females, 7–8 weeks old) were injected subcutaneously at one site in the flank with 200 μl of emulsion containing PLP139-151 (100 μg) in CFA containing 300 μg mycobacterium tuberculosis H37Ra (Difco). Mice received 300 ng pertussis toxin in 500 μl PBS in the tail vein immediately and 48 h after immunization. Following the encephalitogenic challenge, mice were observed and scored as previously described [
28]. For further tissue processing, mice were anesthetized with a xylazine/ketamine mixure and perfused intracardially with cold 4 % paraformaldehyde (PFA). Spinal cords and brains were post-fixed in 2 % PFA and cryoprotected in a 15 % sucrose solution. Free-floating sections (16 μm thick for the spinal cord and 30 μm thick for the brain were cut coronally with a sliding microtome (Leica SM 2000r; Leica, Nussloch, Germany) and stored at 4 °C prior to immunolabeling.
Immunolabeling and quantification
Immunocytochemistry
Coverslips were washed with PBS, fixed with 2 % PFA, treated with a permeabilization/blocking solution [(10 % FCS, 2 % bovine serum albumin, 1 % glycine, and 0.1 % Triton X-100 (Sigma-Aldrich, Rehovot, Israel)] and stained with either of the following antibodies (diluted in the permeabilization/blocking solution): monoclonal anti-mouse GFAP (1:400); rabbit anti-NG2 (1:200); rat anti-MBP (1:300); goat anti-DCX (1:200); monoclonal anti-β-tubulin III (1:500); rabbit anti-NF200 (1:500); and monoclonal anti-CXCL12 (1:84). Cover slips were exposed to primary antibodies for 1 h in a humidified chamber at room temperature. Secondary anti-IgG antibodies used included: Cy3-conjugated goat anti-mouse, Cy3-conjugated goat anti-rabbit, Alexa488-conjugated donkey anti-rabbit, Cy2-conjugated donkey anti-rat, Rhodamine Red™-x-conjugated donkey anti-goat, Cy3-conjugated goat anti-mouse, Cy5-conjugated donkey anti-mouse, and Cy5-conjugated donkey anti-rat. All secondary antibodies were purchased from Jackson ImmunoResearch Laboratories Inc. and used at a dilution of 1:250, except for Rhodamine Red™-x-conjugated donkey anti-goat (1:100). Cover slips were exposed to secondary antibodies for 1 h in a humidified chamber at room temperature. Following the immunostaining, cells were counterstained with 4′,6′-diamidino-3-phenylindole (DAPI) to visualize the nuclei. Control coverslips (not treated with primary antibody) were used to distinguish specific staining from non-specific staining or autofluorescent components.
Slides were examined using an LSM 510 laser scanning confocal microscope (Carl Zeiss, Jena, Germany). Digital images were acquired using the Zeiss LSM 510 software (magnifications ×10, ×25, ×40, or ×63). Images were processed using Photoshop software.
Immunofluorescent staining of tissue sections
Free-floating sections of the spinal cord or brain were washed twice in PBS followed by blocking in 3 % rabbit or goat serum and 0.1 % Triton-X-100 in PBS for 1 h at RT. Samples were stained with the following antibodies in 1 % of the appropriate serum and 0.1 % Triton-X-100 in PBS at 4 °C for overnight: monoclonal anti CXCL12 (1:84) with rabbit anti-GFAP (1:200), rabbit polyclonal anti NG2 (1:200), or with goat anti DCX (1:200). In some cases, samples were stained with rat anti MAC2 (1:200), rabbit anti CD3 (1:100), or with rat anti CXCR4 (1:200). For detection of primary antibodies, samples were stained with either of the secondary antibodies described above. Images were acquired as described above.
Quantification
Quantification of CXCL12 and the number of MAC2+/CD3+, GFAP+, DCX+, and NG2+ alone or with CXCL12 at the different stages of EAE was performed on the same groups of animals. For analysis of the brain tissues, CD3+ and MAC2+ cells were quantified in the forebrain and midbrain and averaged; MAC2+ cells were also quantified in the dentate gyrus. Astrocytes were scored in the dentate gyrus by counting GFAP+ and GFAP+/CXCL12+ cells. Neuronal progenitor cells were scored by counting DCX+ cells in the subgranular zone (SGZ) of the dentate gyrus as well as DCX+/CXCL12+ cells. Oligodendrocyte progenitor cells were scored by counting NG2+ cells in the corpus callosum as well as NG2+/CXCL12+ cells. All cell lineages are expressed as mean number of cells per square millimeter. For quantification, 5–6 coronal consecutive sections (at 300 μm intervals) per mouse brain were stained, counted, and calculated for square millimeter. The nuclei of cells were visualized by DAPI counterstaining.
For in vitro analysis, the percentage of antibody-labeled cells was determined by evaluating 800–1000 cells in at least six randomly chosen fields of view (under ×20 objective). All analyses were performed by an observer blinded to identity of the examined slides.
Image analysis
Quantitative colocalization analysis has been extensively used to reliably determine the colocalization of proteins. The method used in this study has been previously described [
29]. The weighted colocalization coefficient (WCC), which is the sum of intensities of colocalizing pixels relative to the overall sum of pixel intensities above the threshold (or background), was used to determine the relative levels of CXCL12 that colocalized with NG2 immunoreactive cells. This was calculated using the colocalization module according to Manders et al. [
30]. The advantage of a weighted colocalization coefficient is that differences in pixel intensity are taken into account (i.e., not all pixels contribute equally to the final colocalization coefficient value).
Adult neural stem cell culture
The neural stem cells were generated from the subventricular zone (SVZ) of the lateral ventricle from brains of C57Bl/6 mice (5–7 weeks old). Briefly, SVZ was isolated following coronal sectioning and cut into 1 mm3 pieces. The tissue was minced and incubated for digestion at 37 °C, 5 % CO2 for 45 min in 0.25 % Trypsin-EDTA (Biological industries, Beit-Haemek, Israel). Following centrifugation at 110g for 10 min at room temperature, the tissue was further digested in Earle’s balanced salt solution containing 0.94 mg/ml papain (Sigma-Aldrich, Rehovot, Israel) and 0.01 % DNase (Sigma-Aldrich, Rehovot, Israel) for 30 min at 37 °C, 5 % CO2. Then, the tissue was mechanically dissociated by pipette trituration. Single-cell suspension were plated (3500 cells/cm2) in 75 cm2 Falcon tissue culture flasks (BD Biosciences, Franklin Lakes, NJ, USA), in Neurospheres medium [Dulbecco’s modified Eagles’ medium (DMEM):F12 medium (Invitrogen Corp.) supplemented with B27 supplement (Invitrogen Corp.), glucose, Hepes, bFGF (human recombinant, 20 ng/ml) and EGF (mouse recombinant, 20 ng/ml); both from PeproTech, (Rocky hill, NJ, USA)].
Fresh media was added every 3–4 days to maintain the cells as proliferating neurospheres, which were then passaged every 4–6 days and re-plated as single cells.
The neurospheres were differentiated towards different neural lineages by plating cells on Poly-d-lysine [PDL (Sigma-Aldrich, Rehovot, Israel)], in growth factor-free neurosphere medium containing 5 % serum (differentiation medium). For immunocytochemistry, cells were plated on coverslips pre-coated with PDL. In some differentiation experiments, NSCs (2 × 104) were cultured in the presence or absence of 10 ng/ml CXCL12 (PeproTech, Rocky hill, NJ, USA) in differentiation medium. To monitor the effect of the CXCR4 antagonist AMD3100 on differentiation of NSCs, AMD3100 (100 ng/ml, Sigma-Aldrich, Rehovot, Israel) was applied with or without CXCL12 (10 ng/ml) for 4 days in differentiation medium. AMD3100 was replaced twice a day.
Densitometry and statistical analysis
ImageJ densitometry software (version 1.36, National Institutes of Health, Bethesda, MD, USA) was used for quantification of CXCL12 intensity from images of brain sections. Results are expressed as mean ± SEM. Statistical significance was assessed with an unpaired two-tail Student’s t test (Excel software). P < 0.05 was considered significant.
Discussion
The data presented in this study highlight the post-inflammation role of CXCL12 in the context of endogenous myelin/neuronal repair associated with the spontaneous recovery from chronic clinical EAE. The association of CXCL12 with a potential repair of EAE-associated CNS tissue damage was initially suggested by the unexpected but consistent detection of about twofold increase in the expression of CXCL12 in the CNS of mice that had spontaneously recovered from clinical EAE (compared to naïve mice). We observed a relatively high expression of CXCL12 in the CC and in the neurogenic DG of recovered mice up to 20 days after these mice were free of clinical symptoms and their CNS inflammation had resolved. Post-inflammatory expression of CXCL12 was associated with an increase in numbers of CXCL12
+ NG2
+ OPCs and CXCL12
+ DCX
+ NPCs in the CC and DG, respectively, of EAE-recovered mice not only compared to naïve mice but also relative to EAE-affected mice at the peak of the disease. These findings suggest the involvement of CXCL12 in the endogenous myelin/neuronal repair of the EAE-associated CNS damage is by increasing endogenous neuro/oligodendrogenesis and possibly also the differentiation of NPCs and OPCs. Prior in vitro studies [
13,
42] and previous reports of the various functions of CXCL12 during CNS development and following injuries to the adult CNS, including demyelination models, suggest the effect of CXCL12 on the proliferation of neural progenitor cells [
8,
10,
27,
43].
Strikingly, in the context of the beneficial post-inflammatory role of CXCL12, we found that in mice with EAE, a significant proportion of neural progenitor cells co-express both CXCL12 and its CXCR4 receptor (Fig.
9). Moreover, the proportions of these CXCL12
+ CXCR4
+ DCX
+ NPCs and CXCL12
+ CXCR4
+ NG2
+ OPCs, which increased with disease progression further increased in recovered mice (2.12- and 2.75-fold, respectively, compared to naïve mice;
p ≤ 0.0002). These findings are consistent with potential endogenous neuro/oligodendrogenesis and neural differentiation that can contribute to myelin/neural repair via autocrine signaling mechanisms, independently of inflammation-induced CXCL12-producing cells (activated microglia/macrophages and astrocytes) which are cleared from the CNS during spontaneous recovery from clinical EAE.
The expression patterns and potential functions of CXCR4 and its ligand CXCL12 have been widely investigated during CNS development and in healthy or injured adult CNS [
7,
10,
11,
21,
44]. However, few studies have analyzed the potential function of CXCL12 in the context of myelin or neuronal repair in CNS inflammatory and demyelinating diseases, such as EAE/MS [
7,
8,
22,
25]. Our analysis of the expression dynamics of CXCL12 with progression of EAE and following recovery consistently indicate that CXCL12 expression is greater in the DG and CC of mice which spontaneously recovered from severe clinical EAE than in naïve mice. This was somewhat surprising since the activated macrophages/microglial cells and astrocytes, which were reported to be as the major source of CXCL12 in the inflamed CNS of MS patients [
3,
11,
45] and in the cuprizone-induced model of demyelination [
8,
9,
46] are abolished during the EAE recovery phase. Such an observation, therefore, raised the question not only of the potential function for elevated levels of CXCL12 in the DG and CC of EAE-recovered mice but also of the origin of such elevated levels of CXCL12 at the DG and the CC of EAE-recovered mice.
Our immunohistochemistry analysis suggested that NPCs and OPCs that co-express CXCL12 could also be the source of the CXCL12 that was detected at the DG and CC of EAE-recovered mice. The expression of CXCL12 by neuronal progenitors or mature neurons has been reported [
33,
38,
47]. In contrast, the expression of CXCL12 by OPCs or by oligodendrocytes had not been described before either during development or in the adult CNS, despite intensive studies on the involvement of CXCL12 in the proliferation [
9,
27,
39], differentiation, and maturation of OPCs [
8,
27] and their CNS migration [
27,
39]. We have, therefore, confirmed that NG2+ OPCs in the CC and the spinal cord of EAE-recovered mice do express CXCL12, using confocal three-dimensional reconstruction and colocalization analyses. A similar confocal microscopy analyses confirmed that also in vitro differentiating NG2
+ OPCs co-express CXCL12, independent of inflammatory factors present in the CNS during active EAE, suggesting a vital function for a basal expression of CXCL12 by OPCs.
Numerous studies have reported the expression of the CXCR4 and CXCR7 receptors by NSCs, OPCs, and NPCs, and the functional responses to activation by their only ligand, CXCL12, during development [
48] and in inflamed adult CNS, particularly in MS/EAE [
7,
8,
11,
22,
49]. While our detection of CXCR4 expression by NG2
+ OPCs and DCX
+ NPCs in CC and DG, respectively, of EAE-recovered mice was not surprising in view of these studies and of similar observations for in vitro differentiated NPCs and OPCs [
8,
11,
27,
36,
39] (and Figs.
8 and
9), the finding that a significant proportion of these CXCR4
+ DCX
+ NPCs and CXCR4
+ NG2
+ OPCs co-expressed also the CXCR4 ligand, CXCL12, was quite unexpected. Most importantly, the frequencies of CXCL12
+ CXCR4
+ DCX
+ NPCs and CXCL12
+ CXCR4
+ NG2
+ OPCs in the DG and CC not only increased during disease progression but also during recovery with numbers significantly greater by about 1.5-fold than those at disease peak. These findings further support potential involvement of CXCL12 in promoting endogenous neuro/oligodendrogenesis and the differentiation/maturation of NPCs and OPCs towards myelin and neuronal repair following EAE-associated CNS tissue damage.
NG2
+ OPCs and DCX
+ NPCs that co-express CXCR4, and its ligand, CXCL12, have not been previously reported in EAE models or in patients with MS. Whether such CXCL12
+ CXCR4
+ neural progenitor cells can be intrinsically activated to differentiate via autocrine signaling mechanisms in MS/EAE has not been determined. Published data, however, suggest that CXCL12 can act as an autocrine modulator of neuronal activity [
40,
47,
50]. Thus, it is plausible that the elevated numbers of CXCL12
+ CXCR4
+ DCX
+ NPCs and of CXCL12
+ CXCR4
+ NG2
+ OPCs that persist in the CNS of EAE-recovered mice are intrinsically activated via autocrine signaling mechanisms to differentiate into neural cells that would eventually mediate the repair of inflammation-associated myelin/neuronal damage. Such a self-sufficient mechanism (autocrine signaling) of NPCs and OPCs activation is likely to be essential for continuing the repair of the myelin/neural damage after the deleterious inflammation has subsided, and with it, the CXCL12 secreted by the inflammatory cells, particularly, when the small amounts of CXCL12 secreted by the vascular endothelia cells may not be sufficient for the intensive myelin/neuronal repair required after the deleterious inflammation.
We first detected an increase in the number of NPCs and OPCs at early clinical onset, during early inflammatory phase when the number of GFAP
+ CXCL12
+ astrocytes and the level CXCL12 also increase in the CNS ([
2‐
4,
35], and this study). This suggests that the CXCL12, a major inflammatory cell chemoattractant [
11], both facilitates inflammatory damage to CNS tissue during disease progression and promotes proliferation of NPCs and OPCs via paracrine signaling mechanisms, as many of the NPCs and OPCs express the CXCL12 receptor, CXCR4 [
7,
8,
11,
27,
36]. The significant increase in neural progenitors during early EAE onset suggests that the endogenous neural repair mechanisms are being activated and built up already with development of the inflammation in CNS, and that the CXCR4-dependent activation by CXCL12 secreted by inflammatory cells may further drive the NPCs and OPCs towards differentiation/maturation and remyelination and neuronal repair [
8,
10,
16]. Apparently, such myelin/neuronal protection/repair mechanisms are not sufficient to compete with the CNS damage incurred by inflammation. However, the endogenous repair process becomes more effective during the recovery phase when the CNS inflammation/damage is arrested and the CXCL12-expressing neural progenitors can increase in numbers and differentiate to mature neurons and oligodendrocytes without being perturbed by the deleterious inflammation.
The studies by McCandless et al. [
4] and Meiron et al. [
51] support the role of CXCL12 in neural protection following CNS inflammation in EAE. They showed in C57BL/6 mice that blocking CXCR4 activation by administering the CXCR4 antagonist, AMD3100 [
4], or by anti-CXCL12 antibody [
51] significantly increases the clinical severity of EAE induced by myelin oligodendrocyte glycoprotein which, unlike untreated mice, reach a plateau phase without significant remission. Different mechanisms were suggested in theses studies: aggravation of disease in the AMD3100-treated mice was attributed to elevated CNS inflammation associated with the disruption of CXCR4’s function in limiting the infiltration of autoreactive effector cells and migration of leukocytes into the parenchyma [
4]. The disease aggravation by anti-CXCL12 antibodies was attributed to the role of CXCL12 in redirecting the polarization of effector Th1 cells into antigen-specific regulatory T cells in a CXCR4-dependent manner. However, the increase in the number of CXCR4
+ NPCs and OPCs and of CXCL12 expressing CXCR4
+ NPCs/OPCs already at early disease onset (shown above) suggests that the blockade of the CXCR4-dependent endogenous potential for neural protection/repair of CXCR4-expressing NPCs and OPCs through paracrine or autocrine CXCL12/CXCR4 signaling mechanisms may also be at play in disease aggravation and lack of remission following treatment with these agents. Therefore, it is likely that blocking the CXCR4-dependent promotion of NPC and OPC proliferation at an early phase of the disease can also contribute to aggravation of the disease by perturbing the remission mechanisms (myelin/neuronal protection) following treatment with AMD3100 or with anti-CXCL12 antibodies.
Conclusions
This study highlights the post-inflammatory beneficial role of CXCL12 in adult CNS following MS-like inflammatory damage. In mice with EAE, the CXCL12/CXCR4 axis appears to be involved in promoting the endogenous myelin and neuronal repair capacity, particularly in mice undergoing spontaneous recovery. The spontaneous recovery from severe clinical EAE was associated with elevated CXCL12 expression in the DG and CC of spontaneously recovered mice, even though the CNS inflammation has subsided. The data presented above link CXCL12 expression in the DG and CC of EAE-recovering mice to promotion of the endogenous neuro/oligodendrogenesis. A significant proportion of the newly generated progenitor cells are CXCR4+ CXCL12+ NPCs and OPCs, endowed with intrinsic potential of neuro/oligodendroglial differentiation and myelin/neuronal repair via autocrine signaling mechanisms. The detection of CXCL12+ CXCR4+ DCX+ NPCs and CXCL12+ CXCR4+ NG2+ OPCs in the DG and CC, respectively, of mice with EAE and their significant increase in numbers in spontaneously recovered mice are consistent with a post-CNS-inflammation role of CXCL12 in promoting the endogenous potential of repairing myelin and neuronal damage, likely via self-sufficient (autocrine) signaling mechanism which is not dependent on CXCL12 secreted during inflammation.
Notably, the detection at the DG and CC of naïve mice of relatively small, but consistent, numbers of CXCL12+ CXCR4+ NPCs and CXCL12+ CXCR4+ OPCs, respectively, may also link the CXCL12/CXCR4 axis to the maintenance of neural homeostasis in normal adult CNS through autocrine signaling mechanisms. This intriguing possibility warrants further investigation.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
RZF, Nathali K, and AB-N conceived the study and designed the experiments. RZF and Nathali K performed the experiments. Naoto K assisted with confocal imaging and analysis, and RZF and AB-N wrote the manuscript. All authors discussed the results and conclusions and reviewed the manuscript. All authors read and approved the final manuscript.