Oligodendrocyte‐specific deletion of FGFR2 ameliorates MOG35‐55‐induced EAE through ERK and Akt signalling

Abstract Fibroblast growth factors (FGFs) and their receptors (FGFRs) are involved in demyelinating pathologies including multiple sclerosis (MS). In our recent study, oligodendrocyte‐specific deletion of FGFR1 resulted in a milder disease course, less inflammation, reduced myelin and axon damage in EAE. The objective of this study was to elucidate the role of oligodendroglial FGFR2 in MOG35‐55‐induced EAE. Oligodendrocyte‐specific knockout of FGFR2 (Fgfr2ind −/−) was achieved by application of tamoxifen; EAE was induced using the MOG35‐55 peptide. EAE symptoms were monitored over 62 days. Spinal cord tissue was analysed by histology, immunohistochemistry and western blot. Fgfr2ind −/− mice revealed a milder disease course, less myelin damage and enhanced axonal density. The number of oligodendrocytes was not affected in demyelinated areas. However, protein expression of FGFR2, FGF2 and FGF9 was downregulated in Fgfr2ind −/− mice. FGF/FGFR dependent signalling proteins were differentially regulated; pAkt was upregulated and pERK was downregulated in Fgfr2ind −/− mice. The number of CD3(+) T cells, Mac3(+) cells and B220(+) B cells was less in demyelinated lesions of Fgfr2ind −/− mice. Furthermore, expression of IL‐1β, TNF‐α and CD200 was less in Fgfr2ind −/− mice than controls. Fgfr2ind −/− mice showed an upregulation of PLP and downregulation of the remyelination inhibitors SEMA3A and TGF‐β expression. These data suggest that cell‐specific deletion of FGFR2 in oligodendrocytes has anti‐inflammatory and neuroprotective effects accompanied by changes in FGF/FGFR dependent signalling, inflammatory cytokines and expression of remyelination inhibitors. Thus, FGFRs in oligodendrocytes may represent potential targets for the treatment of inflammatory and demyelinating diseases including MS.


| I N T RODUC T ION
Multiple sclerosis (MS) is a chronic inflammatory and neurodegenerative disease of the central nervous system (CNS) characterised by immune-mediated demyelination, oligodendrocyte injury and axon degeneration. In an attempt to repair myelin sheaths, oligodendrocyte precursor cells (OPC) migrate to lesion areas and differentiate into mature oligodendrocytes (1). However, this repair mechanism often fails in MS presumably because of impaired migration capacity and differentiation of OPCs into mature oligodendrocytes (2). In fact, recent reports showed that growth factors and inhibitors of remyelination regulate migration and differentiation of OPC (2). In an experimental autoimmune encephalomyelitis (EAE) disease mouse model for MS, deletion of ciliary neurotrophic factor (CNTF) and brain-derived neurotrophic factor (BDNF) caused a more severe disease course and decrease in myelin sheath thickness (3,4). Furthermore, treatment with an antibody against LINGO-1 resulted in a significant decrease of motor deficits, promoted spinal cord remyelination and axonal integrity in EAE (5). These studies indicate that modulation of growth factors and inhibitors of remyelination may result in functional improvement in a model for MS.
Accumulating evidence suggests that fibroblast growth factor (FGF)/FGFR signalling plays a significant role in the pathology of MS. Findings from MS brain tissue suggest that FGF1 promotes remyelination (6) and that FGF2 enhances migration of (FGFR1+) OPCs to lesion areas (7), whereas FGF9 inhibits myelination and remyelination (8). The corresponding receptor FGFR1 is upregulated in an OPC subpopulation in active and around chronic lesions of MS (7). FGF1 is expressed in oligodendrocytes, astrocytes, microglia/macrophages and infiltrating lymphocytes (6), FGF2 is mainly detected in microglia/macrophages in lesion areas of MS (7), whereas FGF9 is expressed in oligodendrocytes and astrocytes (8). The role of FGFs in MS pathology is evident, also since analyses of cerebrospinal fluid (CSF) showed higher levels of FGF2 in patients and a higher expression of FGF2 in relapses (9). However, data on FGFR2 in MS or its model EAE are not available.
In mice, FGFR1 is expressed in both OPCs and differentiated oligodendrocytes, whereas FGFR2 is found only in differentiated oligodendrocytes (10). Findings on the function of FGFR1 in oligodendrocytes are available from two demyelinating disease models (11,12). In a toxic demyelinating mouse model, a cell-specific deletion of FGFR1 in oligodendrocytes caused an increase in myelin thickness and axon diameters (12). To further delineate the cell-specific functions of FGFR1 in oligodendrocytes in MS, we generated an oligodendrocyte cell-specific FGFR1 knockout (Fgfr1 ind−/− mice) and subjected it to MOG 35-55 -induced EAE (11). Our results revealed that Fgfr1 ind−/− mice showed fewer motor deficits, and reduced myelin and axon degeneration (11). Additionally, an increased phosphorylation of the FGFR downstream signalling molecules ERK and Akt, and increased expression of the neurotrophin BDNF and its receptor TrkB were found in Fgfr1 ind−/− mice (11).
Although, both of these receptors are localised in oligodendrocytes, a recent study suggested that FGFR2 is more important for myelin growth than FGFR1 (13). Moreover, systemic ablation of FGF2 in an EAE mouse model showed higher motor deficits, enhanced axonal loss and decreased remyelination (14). However, the function of FGFR2 in oligodendrocytes and its role in de-and remyelination processes in EAE models of MS is unknown. Therefore, we generated a cell-specific deletion of FGFR2 in oligodendrocytes and characterised its functional role in the EAE mouse model of MS. We hypothesised that deletion of oligodendroglial FGFR2 (Fgfr2 ind−/− mice) leads to less motor deficits and protection of myelin and axons in the chronic phase of EAE. Correspondingly, our data revealed less motor deficits, less myelin and axon degeneration in Fgfr2 ind−/− mice in the chronic phase. Moreover, we observed less lymphocyte and macrophage/microglia infiltration in spinal cord lesions in Fgfr2 ind−/− mice. Increased expression of CD200 and less release of the pro-inflammatory cytokines TNF-α and IL-1β were found in Fgfr2 ind−/− mice in the chronic phase of EAE. Taken together, these data suggest that FGF/FGFR pathways are important for inflammation and myelination in MS and demyelinating disease models.

| Ethics statement
Animal studies were performed according to the guidelines of FELASA. Animal experiments were approved by the local state authorities of Hesse, Giessen, Germany (GI 20/23-Nr. 31/2008) in accordance with the German animal welfare law and the European legislation for the protection of animals used for scientific purposes. Mice were maintained in a 12-h light/dark cycle with a standard pellet diet and water ad libitum. All efforts were made to minimise pain and suffering.

| Histopathology
Fgfr2 ind−/− and control mice were anaesthetised and transcardially perfused with 4% of paraformaldehyde. Spinal cord tissues were collected and embedded in paraffin blocks. A minimum of six spinal cord cross sections were examined per animal. Spinal cord sections were investigated for inflammatory infiltrates (haematoxylin and eosin), myelin loss (Luxol fast blue/periodic acid-Schiff and MBP) and axon degeneration (Bielschowsky silver impregnation). Spinal cord sections were analysed by light microscopy (Olympus BX51, Hamburg, Germany), images were captured using a digital camera (Olympus DP71, Olympus America Inc., Centre Valley PA, USA). The inflammatory index was calculated as the average number of perivascular inflammatory infiltrates in spinal cord white matter lesions. The extent of myelin loss was calculated by calculating the ratio of the areas of myelin loss and the entire total white matter area using the Image J software (Image J 1.47d, NIH, USA). Axonal densities were evaluated in Bielschowsky silver impregnated sections as described earlier (16). Numbers of axons were counted, and axonal density within white matter lesions was compared with axonal density of normal appearing white matter.

| Immunohistochemistry
For immunohistochemistry, deparaffinised spinal cord sections were hydrated and antigens were retrieved by boiling the sections in an appropriate buffer (CD3, Mac-3, B220, MBP and NogoA in citrate buffer (pH 6.0); Olig2 in TE buffer). Sections were incubated overnight with primary antibodies at 4°C. Antibodies are summarised in Table S1 provided as Supporting Information. On the next day, sections were incubated with biotinylated secondary antibodies (goat anti-rat; Mac 3, B220 and goat anti-rabbit; CD3, MBP, Olig2, NogoA), and the antigen-antibody complex signals were detected by incubation with an avidin-biotin complex by DAB. Haematoxylin staining was performed for nuclear staining. Light microscopic images (Olympus BX51, Hamburg, Germany) were captured with a digital camera (Olympus DP71, Olympus America Inc., Centre Valley PA, USA). MBP(+) immunostainings were analysed by calculating the ratio of MBP(+) myelin in lesion areas and total white matter using ImageJ 1.47d. CD3(+), B220(+), Mac3(+), Olig2(+) and NogoA(+) cells were counted within spinal cord white matter lesions with an ocular morphometric grid. The average numbers of positive cells were normalised to an area of 1 mm 2 .

| Protein quantification
Spinal cord tissues were homogenised in protein lysis buffer with tissue ruptor (Qiagen Instruments, Hombrechtikon, Switzerland). The amount of protein was quantified (Pierce® BCA Protein Assay Kit, Thermo Scientific, IL, USA), normalised and subjected to SDS-PAGE (miniprotein system, Bio-Rad, Munich, Germany). Proteins were transferred (Trans Blot, semi dry Transfer cell, Bio-Rad, Munich, Germany) to a nitrocellulose membrane (GE Healthcare, AmershamTM Hybond ECL, Buckinghamshire, UK) and blocked with 5% of BSA. Membranes were incubated overnight at 4°C with primary antibodies for different targets. Antibodies are summarised in Table S1 provided as Supporting Information.

| Statistical analysis
All analyses were performed in a blinded fashion. EAE scores from three independent experiments were analysed using a Mann-Whitney U test. For immunohistochemical analyses, positively labelled cells were counted. A minimum of five spinal cord sections per mouse was analysed for each parameter. The number of animals per group is provided in the figure legends. Histological, immunohistochemical and western blot data analyses were evaluated using a t-test. Statistical analysis was performed using SigmaPlot 14 (Systat, San Jose, CA, USA). Graphs were prepared using SigmaPlot 14 (Systat, San Jose, CA, USA). Values are expressed as mean ±standard error of mean. *indicates p ≤ 0.05, **indicates p ≤ 0.01, ***indicates p ≤ 0.001.

| Cell-specific deletion of FGFR2 results in reduced motor deficits in MOG 35-55 -induced EAE
To study the cell-specific functions of FGFR2 in oligodendrocytes, EAE was induced in Fgfr2 ind−/− mice (n = 14) and control mice (n = 13) by s.c. injections of the MOG  peptide. Expression of the mutant genes (Fgfr2 lox/lox and PLP cre) was confirmed by agarose gel electrophoresis ( Figure 1A). The disease course was monitored until day 60 p.i. ( Figure 1B). Motor deficits were observed from day 10.6 ± 0.4 p.i. in Fgfr2 ind−/− mice and from day 11.2 ± 0.4 p.i. in controls (p = 0.292). The peak of the disease was seen on day 14 in Fgfr2 ind−/− mice and on day 15 in controls. From day 24 p.i. oligodendroglial Fgfr2 deficient mice showed less symptoms than controls (p < 0.05). In the remission phase, beginning at day 18 p.i. in Fgfr2 ind−/− mice and controls, both groups showed an improvement in motor deficits, which was more pronounced in Fgfr2 ind−/− mice. These data demonstrate that cell-specific deletion of FGFR2 in oligodendrocytes reduces motor deficits in the MOG 35-55 -induced EAE mouse model.

| Deletion of oligodendroglial FGFR2 regulates FGF and FGFR expression
Since FGF/FGFR expression is regulated in MS and EAE (6,11), we next investigated whether cell-specific deletion of FGFR2 alters the expression of FGF and FGFR-related molecules in spinal cord homogenates. The expression of FGF2 (p = 0.206) and FGF9 (p = 0.570) was not affected by deletion of FGFR2 in the acute phase of EAE ( Figure 1C). In contrast, in the chronic phase of EAE expression of FGF2 (p = 0.003) and FGF9 (p = 0.004) was downregulated in Fgfr2 ind−/− mice ( Figure 1D). FGFR2 expression was less in Fgfr2 ind−/− mice in the acute (p = 0.017; Figure 1C) and chronic (p = 0.002; Figure 1D) phases of EAE. However, FGFR1 protein expression was not affected by deletion of FGFR2 (p = 0.796 for acute phase; p = 0.745 for chronic phase; Figure 1D). These findings indicate that cell-specific deletion of FGFR2 in oligodendrocytes alters molecules of FGF/FGFR pathways.

| Cell-specific deletion of FGFR2 in oligodendrocytes reduces inflammation, myelin and axonal degeneration
Demyelination and axonal degeneration are key features of MS and EAE (17). Since cell-specific deletion of FGFR2 caused a reduction of motor deficits in EAE, therefore, we investigated the underlying pathology in the spinal cord. To address this question, spinal cord white matter lesions from control and Fgfr2 ind−/− mice were quantified for the degree of inflammation. Our results revealed no differences in the inflammatory index between

| Cell-specific deletion of FGFR2 alters the composition of inflammatory infiltrates
It is known that peripheral activation of myelin-specific T cells and subsequent reactivation of these cells in the CNS are key mechanisms of MS and EAE (1,18). Inflammatory cells found in lesion areas are macrophages, microglia, T cells, B cells and plasma cells (18). The composition of cellular infiltrates was investigated to depict the effect of  Figure 4P-R) was significantly lower in Fgfr2 ind−/− compared to controls. These data suggest that cell-specific deletion of FGFR2 has an anti-inflammatory effect in EAE.

| Deletion of oligodendroglial FGFR2 alters cytokine expression
It is known that cytokines regulate inflammatory processes in MS (18) and EAE (19). To investigate the effect of oligodendroglial FGFR2 deletion on pro-inflammatory cytokine regulation, spinal cord lysates were analysed by western blot. Indeed, our results revealed that the proinflammatory cytokines TNF-α (p = 0.044 for acute; p = 0.038 for chronic) and IL-1β (p = 0.047 for acute; p = 0.003 for chronic) were significantly less induced in Fgfr2 ind−/− mice with EAE compared to controls ( Figure 5A,B). However, there were no differences in the expression of IFN-γ (p = 0.974 for acute; p = 0.759 for chronic), iNOS (p = 0.202 for acute; p = 0.090 for chronic) or IL-6 (p = 0.562 for acute; p = 0.318 for chronic) between Fgfr2 ind−/− mice and controls ( Figure 5A and B). To study the anti-inflammatory role of FGFR2 in this EAE model, spinal cord lysates were analysed for CD200 expression, a marker for anti-inflammatory activity. Interestingly, in Fgfr2 ind−/− mice, CD200 abundance was significantly downregulated in the acute phase (p = 0.017) and upregulated in the chronic phase of EAE (p = 0.003) ( Figure 5C). Taken together, these data indicate that deletion of oligodendroglial FGFR2 reduces key pro-inflammatory markers in EAE.

| Deletion of oligodendroglial FGFR2 does not alter oligodendrocyte populations
Next, we asked whether the differences in myelin loss between Fgfr2 ind−/− mice and controls were caused by differences in the number of oligodendrocytes. Therefore, we quantified Olig2(+) and NogoA(+) oligodendrocytes in spinal cord white matter lesions. In fact, Olig2 is a nuclear marker for OPCs, and NogoA is a cytoplasmic protein expressed in mature oligodendrocytes. The quantification revealed no differences in the number of Olig2(+) (p = 0.590) or NogoA(+) (p = 0.386) oligodendrocytes between Fgfr2 ind−/− mice and controls in the acute phase of EAE ( Figure 6A-F). Similarly, no differences in the number of Olig2(+) (p = 0.203) or NogoA(+) (p = 0.251) oligodendrocytes were observed in the chronic phase of EAE ( Figure 6G-L). These data reveal that deletion of FGFR2 in oligodendrocytes does not affect the number of Olig2(+) OPCs or NogoA(+) oligodendrocytes in lesion areas.

| Oligodendroglial FGFR2 deletion leads to an increase of PLP and reduces expression of remyelination inhibitors
Since oligodendroglial FGFRs regulate myelin protein expression (11,13,20), next, we investigated whether deletion of FGFR2 has an effect on the regulation of myelin proteins in the EAE model. Therefore, we studied expression of the myelin proteins MBP, PLP and CNPase in spinal cord lysates. Our findings showed no regulation of myelin proteins in Fgfr2 ind−/− mice in the acute phase of EAE (p = 0.130 for MBP; p = 0.247 for PLP; p = 0.707 for CNPase; Figure 6M). In contrast, in the chronic phase of EAE, PLP was significantly upregulated compared to controls (p = 0.049; Figure 6N), however, MBP (p = 0.168; Figure 6N) and CNPase (p = 0.980; Figure 6N) were not altered in Fgfr2 ind−/− mice. To better characterise remyelination in our model, we investigated the expression of the remyelination inhibitors TGF-β, SEMA3A and Lingo-1 in spinal cord lysates. SEMA3A was significantly less in Fgfr2 ind−/− mice in both the acute (p = 0.015; Figure 7A) and chronic (p = 0.004; Figure 7B) phases of EAE. In addition, TGF-β was less in Fgfr2 ind−/− mice in the chronic phase of EAE (p = 0.019; Figure 7B), and it was not changed in the acute phase (p = 0.516; Figure 7A). Expression of Lingo-1 was not changed in Fgfr2 ind−/− mice in the acute (p = 0.126; Figure 7A) and chronic (p = 0.688; Figure 7B) phases of EAE. Taken together, these data suggest that deletion of oligodendroglial FGFR2 favours myelin repair by increased expression of PLP and decreased expression of remyelination inhibitors in the chronic phase of EAE.

| Oligodendroglial FGFR2 regulates ERK/Akt phosphorylation
FGFs bind to FGFRs leading to activation of downstream signalling molecules such as ERK and Akt known to regulate oligodendrocyte development and myelination (10). Therefore, to assess the extent of myelin protection in Fgfr2 ind−/− mice in the chronic phase of EAE, ERK and Akt phosphorylation was measured in spinal cord lysates. Indeed, there were no alterations in phosphorylation observed in Fgfr2 ind−/− mice for ERK (p = 0.390) or Akt (p = 0.691) in the acute phase of EAE ( Figure 8A). However, in the chronic phase of EAE, phosphorylation of ERK was significantly downregulated (p < 0.001) and phosphorylation of Akt was significantly upregulated in Fgfr2 ind−/− mice (p = 0.018) ( Figure 8B). These data indicate a regulation of downstream signalling proteins by oligodendroglial FGFR2 in EAE.

| Deletion of oligodendroglial FGFR2 decreases TrkB expression
We recently showed that BDNF and its receptor TrkB were upregulated by cell-specific deletion of FGFR1 in the chronic phase of EAE (11). Therefore, we asked whether the decrease in axonal degeneration and motor deficits in the chronic phase of EAE in the Fgfr2 ind−/− mice was because of increased axonal preservation through the involvement of the neuronal growth factor BDNF. To address this question, we analysed BDNF and its receptor TrkB in spinal cord lysates by western blot. Interestingly, we did not observe any differences in the expression of BDNF (p = 0.175 for acute; Figure 8A, p = 0.555 for chronic; Figure 8B) between controls and Fgfr2 ind−/− mice. Expression of TrkB was also not regulated in the acute (p = 0.806; Figure 8A), however, TrkB was significantly downregulated in Fgfr2 ind−/− mice in the chronic phase of EAE (p = 0.007; Figure 8B). These data suggest that TrkB is involved in FGFR2-mediated axonal degeneration in the chronic phase of EAE.

| DI SC US SION
Much of the knowledge on cell-specific function of FGFRs in oligodendrocytes is derived from experimental studies on FGFR1 (11,12), however, little information is available on the function of FGFR2 in oligodendrocytes. Therefore, in this study, we investigated the function of FGFR2 by generating an inducible and cellspecific deletion of FGFR2 in oligodendrocytes. These mice were subjected to the MOG 35-55 -induced EAE model of MS. Our results demonstrated a significant reduction of motor deficits accompanied by less myelin and axon degeneration. Furthermore, the number of oligodendrocytes was not altered in demyelinating lesions. However, inflammatory infiltrates and pro-inflammatory cytokines were decreased, whereas an upregulation of CD200 associated with immunosuppression was observed. Analyses of FGF/FGFR2 pathway signalling molecules showed that oligodendroglial FGFR2 modulates pERK and pAkt in EAE.
Oligodendrocyte lineage cells express FGFRs in a developmentally regulated manner (10,21). FGFR1 is expressed in OPCs and mature oligodendrocytes, whereas FGFR2 is expressed solely in mature oligodendrocytes (21,22). Although, recent studies have focused on the function of FGFR1 in different demyelinating mouse models (11,12), the exact functions of FGFRs in oligodendrocytes were not clear yet. Therefore, in an effort to elucidate the function of FGFR, we characterised the FGFR1 in oligodendrocytes in MOG 35-55 -induced EAE (Fgfr1 ind−/− ) (11). Interestingly, Fgfr1 ind−/− mice displayed less motor deficits, and a reduction of myelin and axon degeneration in the chronic phase of EAE. Moreover, no difference in the number of oligodendrocytes in demyelinating lesions was observed between Fgfr1 ind−/− mice and controls. In agreement with our observations, FGFR1 deletion in oligodendrocytes did not change the number of OPCs during demyelination in a toxic demyelination model (12). The functional role of FGFR1 in oligodendrocytes has been studied, however, the role of FGFR2 in oligodendrocytes is not well understood. In fact, there is only a single study describing the physiological function of FGFR2 in oligodendrocytes (13). Furusho and colleagues demonstrated that deletion of FGFR2 in oligodendrocytes exhibited a reduction in myelin thickness. However, FGFR1 deletion in oligodendrocytes did not affect myelin thickness, therefore, the authors suggested that FGFR2 might be a key receptor in oligodendrocytes for transducing extracellular signals for myelin growth (13). Moreover, the function of FGFR2 in oligodendrocytes has not been investigated in disease models, especially in EAE. In agreement to our findings on FGFR1, Fgfr2 ind−/− mice also displayed less severe motor deficits, and decreased myelin and axon degeneration. There were no differences in the number of oligodendrocytes in demyelinating lesions. In contrast to findings in a normal physiological condition, we found an upregulation of the myelin protein PLP expression in Fgfr2 ind−/− mice (13). The effect of cell-specific deletion of FGFR2 does not depend on the number of oligodendrocytes, meaning that other factors are mediating the protection of myelin and axons.
In studies on toxic demyelination, FGF2 deletion enhanced oligodendrocyte repopulation of lesions (23) and increased remyelination (24). In contrast to these findings, deletion of FGF2 caused more severe symptoms in MOG 35-55 -induced EAE (14). In order to circumvent and rescue the deficiency of FGF2, intrathecal injection of a viral vector coding for FGF2 ameliorated MOG 35-55induced EAE symptoms (25). Our findings in Fgfr2 ind−/− mice showed reduced expression of FGF2 and FGF9, which are associated with less myelin and axonal degeneration. Similarly, in Fgfr1 ind−/− mice we observed decreased FGF2 expression (11). These results on FGFs are in agreement with MS patients (6,8) suggesting that FGFs may exert diverse actions in different demyelinating models. Inflammatory infiltrates in lesions cause destruction of oligodendrocytes and myelin sheaths in MS und EAE (26). In EAE, both FGF2 and FGFR1 are increased in activated macrophages/microglia around spinal cord lesions (27,28). It is known that microglia and infiltrating T and B cells express FGF2, which attracts OPCs to lesion areas (7). In MOG 35-55 -induced EAE, FGF2 gene therapy caused a reduction in the number of T lymphocytes and macrophages (25). However, an increase of CD8+ T lymphocytes and macrophages/microglia was found in FGF2 −/− mice (14). FGF expression in demyelinating diseases could serve several functions such as modulating the activity of microglia/macrophages in an autocrine fashion or exerting direct paracrine effects on neighbouring oligodendrocytes mediated by FGFR1 and FGFR2 activation (28). Fgfr1 ind−/− mice showed a reduced number of macrophages/microglia and lymphocytes in demyelinating lesions. In agreement, the present study revealed a decrease of these cells within demyelinating spinal cord lesions of Fgfr2 ind−/− mice. The effect of oligodendroglial FGFR deletion on inflammatory cells was similar in Fgfr1 ind−/− and Fgfr2 ind−/− mice. Moreover, the key pro-inflammatory cytokines TNF-α and IL-1β were downregulated and the anti-inflammatory mediator CD200 upregulated in Fgfr2 ind−/− mice. We were unable to assess the cell-specific expression of cytokines or CD200 in these mice. These data suggest that cell-specific deletion of FGFR2 in oligodendrocytes reduces inflammation.
In contrast to our findings on FGFR1, deletion of oligodendroglial FGFR2 resulted in a downregulation of pERK1/2. The differences in findings between Fgfr1 ind−/− (11) and Fgfr2 ind−/− mice may be explained by different signalling potentials of FGFRs (13). There was no evidence for a compensatory upregulation of FGFR1 or TrkB in Fgfr2 ind−/− mice. Treatment with an ERK inhibitor did not modulate the disease course of EAE (29). MEK1/2 is a signalling protein upstream of ERK (30). In contrast to the study using an ERK inhibitor, treatment with a MEK1/2 inhibitor attenuated EAE by inhibition of IL-23 and IL-1 (31). Global knockout of ERK1 resulted in enhanced susceptibility to MOG 35-55 -induced EAE, myelin destruction and an increase of infiltrating cells (32). Moreover, sustained activation of ERK1/2 in oligodendrocytes resulted in accelerated myelin repair (33). Thus, the observations on the function of ERK are inconsistent. Furthermore, ERK activity is essential for the development of T lymphocytes as it is involved in the positive selection (34). The findings indicate that oligodendroglial FGFR2 deletion modulates EAE by reducing inflammation through ERK. Downstream signalling of FGFRs is also mediated through the PI3 K-Akt pathway. Continuous activation of Akt in oligodendrocytes increases myelin synthesis (35). In the present study, cell-specific deletion of FGFR2 resulted in an upregulation of Akt phosphorylation. Findings for pAkt are consistent for FGFR1 (11) and FGFR2 knockout mice in EAE. Indeed, Akt is known to regulate innate and adaptive immune responses in physiological conditions as well as inflammatory and autoimmune disorders (36). PI3 K-Akt signalling in dendritic cells regulates inflammation in part by an increase of anti-inflammatory IL-10 and a decrease of pro-inflammatory IL-12 (36). In agreement with its anti-inflammatory function, Akt3 −/− mice showed an increase in leucocytes and demyelination in MOG 35-55 -induced EAE (37). The data on FGFRs in oligodendrocytes suggest that pAkt is a modulator of myelination and inflammation in EAE.
In agreement with our recent study (11), we did not observe an effect on the number of oligodendrocytes in demyelinating lesions. Similarly, knockout of both FGFR1 and FGFR2 did not cause a reduction in the number of mature oligodendrocytes in a physiological condition (20). In concordance with these findings, continuous activation of Akt in oligodendrocytes did not alter the number of oligodendrocytes (35). In contrast, double deletion of FGFR1 and FGFR2 in oligodendrocytes caused a decrease in the number of differentiated oligodendrocytes in a toxic demyelination model (38). The expression of myelin protein PLP was upregulated in Fgfr2 ind−/− mice in the chronic phase of EAE ( Figure 6N). However, this myelin protein was not regulated in Fgfr2 ind−/− mice by tamoxifen at an age where EAE was induced. Likewise, other myelin proteins such as MBP and CNPase were also not affected by tamoxifen administration in these mice ( Figure S1). Myelin inhibitor expression of TGF-β and SEMA3A was downregulated in Fgfr2 ind−/− mice. In contrast to our previous study in Fgfr1 ind−/− mice, LINGO-1 was not regulated in the present study. These findings suggest that protection of myelin is mediated by different mechanisms. Protection of axons is mediated by BDNF and TrkB. In this context, recently we showed an upregulation of BDNF and TrkB expression in Fgfr1 ind−/− mice in the chronic phase of EAE (26). However, deletion of FGFR2 in oligodendrocytes did not affect BDNF expression; it caused a reduction of TrkB. In MOG 35-55 -induced EAE, cell-specific deletion of TrkB in astrocytes led to less immune cell infiltration of CD4 T cells, B cells and macrophages. Cell-specific deletion of TrkB protected against neurodegeneration possibly mediated by less NO production (39). The lack of change in BDNF expression and decreased TrkB expression suggests that the beneficial effects of FGFRs for axons depends on different mechanisms.
In summary, cell-specific deletion of FGFR2 in oligodendrocytes exhibited a reduction in motor deficits, less myelin and axon degeneration and a decreased inflammation in a mouse model of MS. The beneficial effects of FGFR2 deletion in EAE are associated with regulation of FGF/FGFR signalling downstream proteins such as pERK and pAkt associated with myelination and inflammation. Moreover, no effect of FGFR2 deletion on the number of oligodendrocytes and BDNF was observed. Importantly, Fgfr2 ind−/− mice showed an upregulation of the myelin protein PLP and downregulation of myelin inhibitor expression. Taken together, FGFR2 in oligodendrocytes plays a key role in the pathology of MOG 35-55 -induced EAE. Thus, FGFRs in oligodendrocytes and their downstream signalling proteins are potential targets for the treatment of MS.