Oligodendrocyte-Specific Deletion of FGFR1 Reduces Cerebellar Inflammation and Neurodegeneration in MOG35-55-Induced EAE

Multiple sclerosis (MS) is a chronic inflammatory and degenerative disease of the central nervous system (CNS). MS commonly affects the cerebellum causing acute and chronic symptoms. Cerebellar signs significantly contribute to clinical disability, and symptoms such as tremor, ataxia, and dysarthria are difficult to treat. Fibroblast growth factors (FGFs) and their receptors (FGFRs) are involved in demyelinating pathologies such as MS. In autopsy tissue from patients with MS, increased expression of FGF1, FGF2, FGF9, and FGFR1 was found in lesion areas. Recent research using mouse models has focused on regions such as the spinal cord, and data on the expression of FGF/FGFR in the cerebellum are not available. In recent EAE studies, we detected that oligodendrocyte-specific deletion of FGFRs results in a milder disease course, less cellular infiltrates, and reduced neurodegeneration in the spinal cord. The objective of this study was to characterize the role of FGFR1 in oligodendrocytes in the cerebellum. Conditional deletion of FGFR1 in oligodendrocytes (Fgfr1ind−/−) was achieved by tamoxifen application, EAE was induced using the MOG35-55 peptide. The cerebellum was analyzed by histology, immunohistochemistry, and western blot. At day 62 p.i., Fgfr1ind−/− mice showed less myelin and axonal degeneration compared to FGFR1-competent mice. Infiltration of CD3(+) T cells, Mac3(+) cells, B220(+) B cells and IgG(+) plasma cells in cerebellar white matter lesions (WML) was less in Fgfr1ind−/−mice. There were no effects on the number of OPC or mature oligodendrocytes in white matter lesion (WML). Expression of FGF2 and FGF9 associated with less myelin and axonal degeneration, and of the pro-inflammatory cytokines IL-1β, IL-6, and CD200 was downregulated in Fgfr1ind−/− mice. The FGF/FGFR signaling protein pAkt, BDNF, and TrkB were increased in Fgfr1ind−/− mice. These data suggest that cell-specific deletion of FGFR1 in oligodendrocytes has anti-inflammatory and neuroprotective effects in the cerebellum in the EAE disease model of MS.


Introduction
Multiple sclerosis (MS) is a chronic inflammatory and degenerative disease of the central nervous system (CNS) associated with oligodendrocyte injury, demyelination, and degeneration of axons [1]. As a consequence of demyelination, oligodendrocytes differentiate, proliferate, and migrate to repair damaged myelin sheaths [2][3][4]. This repair mechanism is effective in early stages of MS, but it deteriorates with disease progression [5][6][7][8][9]. The underlying mechanisms of impaired remyelination include failure of oligodendrocyte progenitor cell (OPC) recruitment and differentiation of OPCs into mature oligodendrocytes [8,10], oxidative stress, and mitochondrial dysfunction [11].
FGF/FGFR signaling pathways are involved in the pathogenesis of MS and its disease models. In autopsy tissue, an increased number of FGF1/2+ macrophages and astrocytes were observed in and around active lesions in the cerebral cortex and periventricular white matter; additionally, FGFR1 was upregulated in oligodendrocyte progenitor cells (OPC) in active lesions and around chronic lesions [12,13]. Analysis of the cerebrospinal fluid (CSF) of patients with relapsing-remitting MS showed higher levels of FGF2 expression than healthy controls and revealed that an increased concentration of FGF2 positively correlates with lesion load of disease activity [14]. Human data indicate the implication of FGF2 in the pathogenesis of MS. We have investigated oligodendrocyte-specific deletion of FGFR in experimental autoimmune encephalomyelitis (EAE) [15,16]. Fgfr1 ind−/− and Fgfr2 ind−/− mice showed less motor deficits, reduced inflammation, and demyelination in the spinal cord. Phosphorylation of Akt and ERK was regulated in the spinal cord of both Fgfr1 ind−/− and Fgfr2 ind−/− mice [15,16]. Consequently, data from this disease model of MS suggest that FGFR in oligodendrocytes exert negative effects.
The cerebellum plays a crucial role in coordination and balance [17]. In MS studies one third of subjects have functionally relevant cerebellar deficits [18,19]. Disabling symptoms such as nystagmus, ataxia, dysphagia, or dysarthria are caused by destruction within the cerebellum [20,21]. In fact, the prognosis is worse when patients have early signs of cerebellar deficits [22,23], and symptoms caused by lesions in the cerebellum respond poorly to symptomatic therapies [24,25]. Impairment of cerebellar function may also lead to a lower quality of life [26]. Thus, cerebellar dysfunction in MS has clinically relevant consequences, its underlying mechanism and treatment strategies remains poorly understood. Inflammation, demyelination, and axonal damage can be found in the whole CNS of patients with MS [17]; however, comparative data on the extent of damage in various regions of the CNS are limited. The number of lesions in the cerebellar white matter is lower than the number in the forebrain white matter. Further, axonal loss or remyelination differs between cerebellar and forebrain plaques, the underlying mechanisms are unclear [25]. Magnetic resonance imaging studies have confirmed extensive cerebellar involvement in both relapsing-remitting and progressive MS. Persistent cerebellar dysfunction has been linked with axonal loss in the cerebellum of patients with progressive MS [27].
In EAE, inflammation, demyelination, or neurodegeneration have been found in various regions of the CNS including the cerebrum, corpus callous, hippocampus, and the cerebellum [28]. Several studies have focused on the pathology of cerebellar structures in EAE [29][30][31][32]. In MOG 35-55 -induced EAE, demyelinating lesions and infiltrates of T cells and macrophages were detected in the cerebellum [28]. A reduced number of CD4 and CD8 positive T cells was found in the cerebellum of mice treated with an anti-CX3CL1 antibody [33]. Multilocular neuroinflammation and demyelination was found in the spinal cord and the cerebellum of astrocyte specific IL-23 producing mice [31]. Increased neutrophil infiltrations and demyelination were found in the cerebellum of Socs3 deficient mice [34], and hypoxia and increased variations in oxygenation were found in the grey matter of the cortex and the cerebellum [35]. In another study using MRI, a significant reduction in cerebellar and cerebral cortex volumes was detected [36].
Based on data from previous EAE studies, we hypothesized that oligodendrocytespecific deletion of FGFR1 results in reduced inflammation and degeneration in the cerebellum. In agreement with our hypothesis, we found reduced inflammatory infiltrates as well as less myelin and axonal degeneration in the cerebellum of Fgfr1 ind−/− mice. Akt phosphorylation, TrkB, and BDNF expression were upregulated in Fgfr1 ind−/− mice. In addition, we found reduced IL-6, IL-1β, and CD200 expression in Fgfr1 ind−/− mice. The expression of FGF2 and FGF9 was reduced, and myelin protein PLP expression was increased. These data suggest that FGF/FGFR pathways are associated with proinflammatory mechanisms and myelination, and they may contribute to disability of EAE caused by cerebellar damage.

Reduced Inflammatory Infiltrates in WML of the Cerebellum in Fgfr1 ind−/− Mice
To assess the effects of oligodendrocyte-specific deletion of FGFR1 in the spinal cord, we previously reported a reduction in CD3(+) T cells, B220(+) B cells and Mac3(+) macrophages/activated microglia in Fgfr1 ind−/− mice on day 62 p.i. [16]. In the cerebellum, Fgfr1 ind−/− mice showed less inflammatory infiltrates at day 62 p.i. (p = 0.0251; Figure 1C-E). To study the composition of these infiltrates in WML of the cerebellum,

Oligodendroglia-Specific Deletion of FGFR1 Reduces Myelin Loss and Axonal Damage in the Cerebellum
In our recent study Fgfr1 ind−/− mice showed less myelin and axon degeneration in the spinal cord at day 62 p.i. [16]. In the cerebellum, myelin loss (p = 0.0034; Figure

FGF/FGFR Signaling Proteins and TrkB/BDNF Protein Expression in the Cerebellum of Fgfr1 ind−/− Mice
In the present study, we investigated the role of oligodendroglia-specific deletion of FGFR1 on FGF/FGFR signaling proteins and TrkB/BDNF protein expression in cerebellar lysates by western blot. Fgfr1 ind−/− mice showed a decrease in the expression of FGF2 (p = 0.0453) and FGF9 (p = 0.0442) ( Figure 6A,B). Phosphorylation of Akt (p = 0.0109, Figure 7A,B), and expression of TrkB (p = 0.0023) and BDNF (p = 0.0012) were increased in the cerebellum of Fgfr1 ind−/− mice ( Figure 6A,B). There was no regulation in phosphorylation of ERK1/2 in the cerebellum ( Figure 7A,B).

Discussion
In the present EAE study, we investigated the function of FGFR1 in the cerebellum by generating an inducible and cell-specific deletion of FGFR1 in oligodendrocytes. At day 62 p.i. chronic phase of EAE, myelin and axon degeneration were less severe; furthermore, lymphocyte and macrophage/microglia cell infiltration reduced. Further analyses revealed enhanced expression of the FGF/FGFR1 signaling molecule Akt activation, upregulation of BDNF/TrkB, and reduced expression of proinflammatory cytokines.
In MS and its EAE model, inflammatory infiltrates cause destruction of oligodendrocytes and myelin sheaths [37]. In EAE, oligodendrocytes directly involved in inflammation and immune modulation in the CNS [38]. FGFs (1, 2, 3, 21, 23)/FGFRs (FGFR1) play important roles in the regulation of inflammatory responses in various diseases including multiple sclerosis [39]. In the spinal cord, oligodendroglia FGFR1 and FGFR2 mice showed less infiltrating CD3(+) T-cells, B220(+) B-cells, and Mac3(+) macrophages/activated microglia during chronic phase of EAE. In agreement with recent findings in the spinal cord, the number of infiltrating CD3(+) T-cells, B220(+) B-cells, Mac3(+) macrophages/activated microglia, or IgG producing plasma cells were reduced in the cerebellum of Fgfr1 ind−/− mice. Our data supports the notion that FGFR1 may directly influence the immune cell infiltration in the cerebellum.
At day 62 p.i., Fgfr1 ind−/− mice showed less FGF2 and FGF9 expression in the cerebellum. To date, it is unclear why deletion of FGFRs in oligodendrocytes causes a reduction in inflammation in the spinal cord and the cerebellum [15,16]. One may speculate that deletion of FGFRs results in a reduction of FGF2 and FGF9 by regulating the downstream signaling of ERK and Akt, which may reduce inflammation by altering the proinflammatory cytokines. In the cerebellum, the proinflammatory cytokines IL1β and IL6 were reduced in Fgfr1 ind−/− mice at day 62 p.i. Cytokines play a key role in recruiting activated immune cells into the CNS [40][41][42]. IL1β plays a role in neuronal degeneration via p53-mediated apoptosis of neurons [43]. IL1β-deficient mice are resistant to EAE [44], and IL1β promotes differentiation of pathogenic Th17 and lymphocyte trafficking into the CNS [45]. FGF2 was downregulated in this study. FGF2 enhances the inflammatory response of monocytes/macrophages, which increases the concentrations of cellular IL-1β in human vascular smooth muscle cells [39]. IL6-deficient animals are fully resistant to EAE [46] and astrocytic deletion of IL-6 leads less inflammatory infiltrates and decreased demyelination in the spinal cord at day 20 p.i [47]. These findings demonstrate that reduced FGF expression was associated with less inflammation and reduced proinflammatory cytokines in the cerebellum of Fgfr1 ind−/− mice.
FGF2 induces downregulation of myelin proteins in vitro [48]. In contrast, conditional deletion of either FGF2 or FGFR enhances remyelination in the cuprizone demyelination model at 12 weeks of cuprizone ingestion. Extended and elevated expression of FGF9 causes failure of remyelination [49]. In vitro, inhibition of FGFR in oligodendrocytes enhances myelin protein expression of PLP and CNPase [50]. Consistent with the cuprizone demyelination model and oligodendrocyte cell culture findings, our results demonstrate that oligodendrocyte-specific deletion of FGFR1 in MOG 35-55 -induced EAE leads to increased PLP expression in the cerebellum and less FGF2 and FGF9 at day 62 p.i. Two different stains were used to assess axonal damage in the cerebellum, namely SMI-31 (NF-H) and APP. To investigate the extent of neurofilaments in the cerebellum, tissues were stained with the SMI31 antibodies. APP is a well-characterized marker to detect axonal injury in neurodegenerative diseases such as MS [51,52]. In damaged axons, the accumulation of APP occurs in case of a disturbed fast axonal transport. Analysis of axonal structure revealed that there were increased phosphorylated neurofilaments in Fgfr1 ind−/− mice cerebellums. This indicates that FGFR1 deletion in oligodendrocytes protected myelin against inflammation and resulted in reduction of axon degeneration and myelin loss in during EAE.
FGFR1 is upregulated in OPCs in active and chronic demyelinating lesions in the cortical and periventricular brain tissue of MS patients [12]. Further, extensive demyeli-nation was present in the cerebellum of patients with progressive MS [25]. In cuprizone and lysolecithin models, FGFR1/2 double knockout mice showed less differentiated oligodendrocytes and suppressed myelin recovery [53]. In our previous EAE studies, we found oligodendrocyte-specific deletion of FGFR1 and FGFR2 does not alter oligodendroglial cell numbers in the spinal cord. In agreement with these recent studies, we did not observe an effect on the number of OPC or mature oligodendrocytes in the cerebellar WML at day 62 p.i. These findings suggest that the impaired FGFR1 in oligodendrocytes does not inhibit oligodendrocyte differentiation in the cerebellum during EAE. The differences in the findings of oligodendrocyte populations could be the lack of inflammatory signals in the cuprizone and lysolecithin model in comparison to the inflammatory environment in the EAE model.
In the cerebellum, Fgfr1 ind−/− mice show increased Akt phosphorylation, which is known to modulate inflammation and TrkB/BDNF expression and enhance myelination in the EAE model. In our previous studies, Fgfr1 ind−/− mice had an increased expression of TrkB and increased phosphorylation of downstream Akt kinase in the spinal cord [15,16]. Akt kinase is a key downstream protein that regulates inflammation [54], and continuous activation of Akt in oligodendrocytes increases myelin synthesis [55]. Due to the suppressive character of Akt2 on the expression of Th1/Th17 cells by increasing the number of T regulatory cells [56,57], a loss of Akt3 leads to severe EAE symptoms and reduces neuroregeneration [58]. Compared to our previous study on the effects of FGFR1 knockout in the spinal cord [16], we could not find an increased phosphorylation of ERK1/2 in the cerebellum. The importance of the functional ERK signaling for adequate remyelination and axon integrity were described in several investigations [59,60]. However, there might be an alternative pathway to ensure sufficient remyelination or decrease axonal loss without upregulation of ERK. Increased expression of BDNF and its corresponding receptor TrkB could possibly enhance remyelination [61].

Ethics Statement
All scientific procedures on animals were approved by the regional council of Hesse, Giessen, Germany (GI 20/23-Nr. 31/2008) in accordance with the German animal welfare act and the European legislation for the protection of animals used for scientific purposes (2010/63/EU). AVMA guidelines for the euthanasia of animals were followed. Animal studies were performed according to the guidelines of FELASA. All efforts were made to minimize pain and suffering

Induction and Clinical Assessment of MOG35-55-Induced EAE
Induction and clinical evaluation of EAE was done as previously described [15,16]. Briefly, 8-to 12-week-old female Fgfr1 ind−/− and control mice were injected subcutaneously in the flank with 300 µg of myelin oligodendrocyte glycoprotein peptide (MOG  ; Charité Hospital, Berlin, Germany) that was emulsified in complete Freund's adjuvant (Sigma, Steinheim, Germany) containing 10 mg of Mycobacterium tuberculosis (Difco, Detroit, MI, USA). Pertussis toxin was administered i.p. on days 0 and 2 postimmunization (p.i.). Symptoms were evaluated using a scale that ranged from 0 to 5 as described and presented earlier [16]. Mice were sacrificed in the chronic phase of EAE on day 62 postimmunization.

Histology and Immunohistochemistry
Fgfr1 ind−/− mice and controls were anesthetized and transcardially perfused with 4% paraformaldehyde on day 62 p.i. The cerebellum was dissected, embedded in paraffin blocks, and stained for inflammatory infiltrates (hematoxylin and eosin) and myelin loss (Luxol fast blue/periodic acid-Schiff) as described recently [16]. For immunohistochemistry, the cerebellum was deparaffinized, rehydrated, and antigen retrieval was performed by boiling the sections in citrate buffer (10 mM, pH 6). Macrophages/activated microglia

Protein Extraction and Western Blot Analysis
The cerebellum from Fgfr1 ind−/− mice and controls were separately homogenized in an ice-cold lysis buffer with TissueRuptor (Qiagen Instruments, Hombrechtikon, Switzerland). The amount of protein was quantified (Pierce ® BCA Protein Assay Kit, Thermo Scientific, IL, USA) and normalized. Protein was loaded and separated by 10% SDS-PAGE and transferred (Trans Blot, Semi-dry Transfer cell, BioRad) to a nitrocellulose membrane (GE Healthcare, Amersham TM Hybond ECL, Buckinghamshire, UK). The membranes were incubated overnight at 4 • C with the target-specific primary antibodies ( Table 1). After that, membranes were incubated with respective secondary antibodies (Table 1) and developed with SuperSignal West Pico Chemiluminescent Substrate (Thermo, Pierce Biotechnology, Rockford, IL, USA) using an ECL ChemoCam Imager (INTAS Science Imaging Instruments GmbH, Göttingen, Germany). GAPDH (Santa Cruz Biotechnology, CA, USA) ( Table 1) was used as a loading control; proteins were analyzed using the ImageJ 1.53b software (National Institute of Health, Bethesda, MD, USA).

Statistics
All analyses were performed blinded to the genotype. For histology and immunohistochemistry, positively labelled cells in lesions from the cerebellum were counted. Numbers of animals per group are noted in the figure legends. Student's t-tests were performed to indicate differences between the genotypes. Statistical significance was set at p ≤ 0.05. Values are expressed as mean ± SEM. * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001. Statistical analysis and graphs were prepared using the GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA).

Conclusions
In the cerebellum, cell-specific deletion of FGFR1 in oligodendrocytes exhibited less axonal damage, less myelin loss, and decreased inflammation in EAE. These effects in the cerebellum were associated by changes in FGF/FGFR signaling, increased BDNF/TrkB and PLP expression. Impaired FGFR1 signaling was associated with reduced expression of proinflammatory cytokines such as IL1β and IL6. In addition, reduced expression of FGF2 and FGF9 as well as increased phosphorylation of Akt were found in FGFR1 knockout mice cerebellums. Furthermore, FGFR1 deletion has no effect on the number of oligodendrocytes. Taken together, inhibiting FGF receptors in the CNS may protect the cerebellum against inflammation and degeneration.  Institutional Review Board Statement: The study was conducted according to the guidelines of Justus Liebig University, Giessen, and approved by regional council of Hesse, Giessen, Germany (GI 20/23-Nr. 31/2008, approved on 19 August 2008) in accordance with the German animal welfare act and the European legislation for the protection of animals.

Data Availability Statement:
The data presented in this study are available upon request to corresponding author.