Introduction

Multiple sclerosis (MS) is the most prevalent autoimmune inflammatory disease of the CNS, which leads to inflammatory demyelination and subsequent neurological symptoms1. The majority of patients initially experience episodes of intermittent neurological deficits followed by varying degrees of recovery, referred to as relapsing–remitting MS (RRMS). Most RRMS patients subsequently transition to secondary progressive MS (SPMS), characterized by irreversible and progressive worsening of neurological symptoms2. In these progressive disease stages, an incomplete understanding of the underlying pathomechanisms is mirrored by a lack of effective treatments.

Sphingosine-1-phosphate is a bioactive lipid mediator, which binds to the sphingosine-1-phosphate receptor (S1PR), a G-coupled receptor with five isoforms differentially expressed in numerous cell types3 including T cells, B cells, monocytes, dendritic cells4, and CNS resident glial cells5,6,7,8. S1PRs are attractive targets for the relapsing stage of MS, as they regulate lymphocyte trafficking9,10. Fingolimod (FTY720), a structural analogue of sphingosine-1-phosphate, has demonstrated clinical efficacy in RRMS and is approved for the treatment of relapsing disease forms11,12. After phosphorylation, its primary effect on lymphocyte homing is mediated through the binding to the S1PR subtypes 1, 3, 4, and 513. This bioactive form of FTY720 also crosses the blood–brain barrier (BBB) and, therefore, has the potential to act on S1PRs on CNS resident cells, including astrocytes and microglia13,14,15,16. Indeed, S1PR modulation demonstrated clinically relevant CNS intrinsic effects in a chronic progressive mouse model of experimental autoimmune encephalomyelitis (EAE) recapitulating several aspects of PMS, due to its actions on CNS resident and infiltrating immune cells and a decrease in pro-inflammatory cytokine expression in astrocytes, microglia and pro-inflammatory monocytes16. In these lines, S1PR modulation exerts beneficial effects in autoimmune CNS inflammation, which has recently been mirrored by the approval of novel S1PR modulators, including Siponimod, Ozanimod and Ponesimod both in relapsing and progressive MS17,18,19.

Interferon-beta (IFN-β), a member of the Type I interferon family of cytokines, is one of the pioneer medications for the treatment of RRMS12,20. Since exogenously administered IFN-β does not cross the blood–brain barrier (BBB), its effects are mediated by dendritic cell modulation and suppression of pro-inflammatory cytokine production in T cells14,21,22. Interestingly, CNS intrinsic glial cells such as astrocytes exhibit a type I Interferon signature during autoimmune CNS inflammation, which promotes anti-inflammatory effects by down-regulating NF-kB-dependent pro-inflammatory pathways in astrocytes. In addition, intranasal application of IFN-β leads to its crossing across the intact BBB and the amelioration of acute and late stages of EAE by acting on astrocytes, microglia, and pro-inflammatory monocytes23.

Here, we investigated whether combining S1PR modulation with intranasal IFN-β would improve therapeutic effects relative to FTY720 alone in progressive CNS inflammation. We assessed the effects of this combinatorial therapy in the non-obese diabetic (NOD) EAE mouse model, a well-established progressive EAE model, particularly valuable for MS research as it closely simulates the chronic and progressive disease course observed in human SPMS24. Furthermore, NOD mice exhibit poor spontaneous recovery, making this model suitable for investigating irreversible neurological damage and disability progression. This capacity to model progressive disease aspects makes it a clinically relevant system for evaluating therapeutic targets. Of note, our decision not to include intranasal IFN-β monotherapy in the present study was based on prior work demonstrating protective effects of IFN-β23, and we therefore focused on the FTY720 plus intranasal IFN-β treatment condition.

Using this approach, we show that treatment with FTY720 plus intranasal IFN-β is associated with improved clinical and immunological outcomes compared with FTY720 monotherapy, accompanied by altered CNS-resident glial responses. Transcriptomic and functional analyses implicate SOCS1-associated signaling as a modulatory component of treatment-induced glial regulation. Together, these findings provide mechanistic insight that supports further exploration of CNS-targeted combinatorial strategies in progressive MS.

Results and discussion

Progressive MS still has unmet needs for targeted and efficient therapeutic strategies. Here, we investigated the combined use of FTY720 and CNS-targeted intranasal interferon-β (nIFN-β), two agents with distinct modes of action, to assess whether CNS-directed IFN-β application can enhance treatment effects relative to FTY720 alone.

Combined therapy of FTY720 and IFN-β ameliorates chronic progressive EAE

While S1PR modulation has demonstrated its efficiency in progressive MS mainly targeting NF-κB dependent pathways16, IFN-β is known to regulate anti-inflammatory factors in glia cells23. To assess whether CNS-targeted IFN-β application can enhance treatment effects relative to S1PR modulation alone during progressive disease, EAE was induced in non-obese diabetic (NOD) mice by immunization with myelin oligodendrocyte glycoprotein (MOG35-55) in Complete Freund’s adjuvant (CFA). The NOD EAE model is characterized by an initial acute disease episode followed by chronic disease progression, therefore recapitulating several pathophysiologic and clinical aspects of progressive MS16. Treatment commenced at the onset of the secondary progressive disease stage (day 30), with animals receiving daily FTY720 alone or FTY720 in combination with intranasal IFN-β, while vehicle-treated mice served as controls (Fig. 1A).

Fig. 1
Fig. 1
Full size image

Combined therapy with FTY720 and nIFN-β ameliorates chronic progressive EAE in NOD mice. (A) Experimental setup of the experiment. (B) Clinical score of NOD mice treated daily with vehicle, FTY720 and vehicle, or FTY720 and nIFN-β from day 30 on (n = 10–12 mice per group; two-way ANOVA). Data are shown as mean ± SEM and are representative of two independent experiments. (**P < 0.01 and ****P < 0.0001).

FTY720 monotherapy prevented chronic worsening of disease, as we and others have described16,25. Notably, combined treatment with FTY720 and intranasal IFN-β resulted in a significantly greater clinical improvement compared to FTY720 monotherapy, including attenuation of previously established neurological deficits at later EAE stages (Fig. 1B). This finding is of relevance, as it indicates that CNS-directed IFN-β application can enhance the therapeutic efficacy of S1PR modulation during progressive disease stages.

Indeed, combined FTY720 and nIFN-β treatment enhanced clinical outcome in progressive EAE compared to FTY720 alone, consistent with recent findings demonstrating that FTY720 exerts both peripheral and central effects that together ameliorate progressive EAE16. Mechanistically, FTY720 primarily functions by inducing lymphatic trapping of T and B cells, thereby inhibiting their egress from lymph nodes and reducing the number of CNS-infiltrating lymphocytes13. In addition to these peripheral effects, phosphorylated FTY720 can cross the BBB26,27 to exert protective effects directly on glial cells and previously infiltrated monocytes28. Similarly, intranasal IFN-β application allows it to bypass the intact BBB, enabling it to act directly on glial and immune cells within the CNS, thus attenuating the clinical course23. Together, these data suggest that CNS-intrinsic and peripheral mechanisms linked to S1PR and IFNAR signaling contribute to the improved clinical disease course observed under combined treatment. We next sought to investigate the molecular and cellular correlates underlying this clinical improvement.

Mitigation of immune cell infiltration and pathogenic cytokine production

We analyzed the cellular composition and molecular signatures of CNS-infiltrating and resident immune cells in FTY720-treated mice with or without intranasal IFN-β during the progressive phase of EAE. Fluorescence intensity of the myelin marker FluoroMyelin (FM) was markedly increased in mice receiving combined FTY720 and intranasal IFN-β treatment (Fig. 2A), whereas the signal intensity of SMI32, an established marker of axonal damage, was reduced in both monotherapy groups and further decreased under combined treatment (Fig. 2B). Iba1⁺ microglia displayed enhanced activation in vehicle-treated mice compared to all treatment groups, indicating pronounced local myeloid activation. In line with this, increased CD45⁺ cell infiltrates were detectable at lesion sites of vehicle-treated animals, further supporting the immunomodulatory effects of S1PR-mediated cell trafficking (Fig. 2C).

Fig. 2
Fig. 2
Full size image

Reduced demyelination and axonal damage under combined FTY720 and intranasal IFN-β treatment. Immunofluorescence (IF) analysis of glial and immune cells in the spinal cord white matter of NOD EAE mice on day 63 treated daily from day 30 with vehicle, FTY720 alone, or FTY720 plus intranasal IFN-β. (A) Representative IF images of FluoroMyelin-labeled myelin (left) and quantification of mean fluorescence intensity (MFI) (right). Scale bar: 200 µm. (B) Representative IF images of SMI32-stained axons in spinal cord sections (left) and corresponding MFI quantification (right). Nuclei were counterstained with DAPI. Scale bar: 50 µm. (C) Representative IF images of Iba1⁺ microglia and CD45⁺ immune cells (left) with corresponding MFI quantification (right). To specifically assess infiltrating immune cells, CD45⁺Iba1⁻ cells were quantified. Nuclei were counterstained with DAPI. Scale bar: 50 µm. All data are presented as mean ± SEM. P values < 0.05 were considered significant and were determined using Welch’s one-way ANOVA (Brown–Forsythe) followed by Dunnett’s T3 multiple-comparisons test.

Combined FTY720 and intranasal IFN-β treatment reduced overall absolute cell numbers in the CNS (Fig. 3A). Both CNS infiltrating T cells and monocytes, including pro-inflammatory Ly6Chigh monocytes, were notably reduced in both treatment groups (Fig. 3B-D), pointing to an FTY720-driven reduction of CNS immune cell infiltration. Furthermore, production of the pro-inflammatory cytokines IFN-γ and IL-17 in T cells was lowered in both treatment groups (Fig. 3C). In addition to the effects of FTY720 on CNS-infiltrating cells, analysis of ex vivo FACS-sorted glial cells revealed reduced pro-inflammatory cytokine levels in astrocytes (Suppl. Figure 1A) and microglia (Suppl. Figure 1B) mediated by FTY720, a notion supported by previous findings of glial modulatory activities on FTY720 on astrocytes16 and microglia29. Importantly, enhanced effects were observed in FACS-sorted astrocytes under combined treatment, where tissue-degenerative pathways, such as Nos2 expression leading to the generation of neurotoxic nitric oxide, were further reduced compared to FTY720 alone (Fig. 3E). Concomitantly, microglia numbers were diminished in mice receiving combined treatment (Fig. 3F), a relevant finding given that microglia proliferation has been implicated in preventing recovery from chronic inflammatory lesions in Multiple Sclerosis patients30. Together, the mechanisms involved in nIFN-β administration include several cell types due to the broad expression of interferon α/β receptors (IFNAR)21,31,32,33, which may complement S1PR-mediated pathways targeted by FTY720.

Fig. 3
Fig. 3
Full size image

Immune cells and glial cells in the inflamed CNS react to the respective therapy. Quantification of immune cells within the CNS of NOD EAE mice on day 63 treated with vehicle, FTY720 and vehicle, or FTY720 and nIFN-β daily from day 30 on (n = 10–12 mice per group). (A) Absolute cell number within the CNS (B) Absolute number of infiltrating monocytes (left) and Ly6C positive monocytes (right) within the CNS. (C) Absolute number of infiltrating CD4 T cells and the amount of IFN-γ and IL-17-producing CD4 T cells in the CNS. (D) Representative FACS plot of infiltrating IFN-γ and IL-17-producing CD4 + T cells. (E) Nos2 production of sorted astrocytes displayed by fold change in mRNA expression. (F) Absolute number of microglia within the CNS. (G) Quantification of CD4 positive T cells within the lymph node and the corresponding number of IFN-γ and IL-17-producing T cells. All data are mean ± SEM and representative of two independent experiments. Outliers were identified using ROUT with Q = 1%. P-values of p < 0.05 were considered significant and determined by Brown-Forsythe and Welch ANOVA tests. Dunnett`s T3 multiple comparison test was used to compute individual variances for each comparison.

Together, FTY720 limits CNS immune cell infiltration and attenuates pro-inflammatory cytokine production in CNS resident cells, while intranasal IFN-β mainly affects CNS resident cell populations, resulting in a further attenuation of pro-inflammatory pathways in astrocytes and microglia compared to FTY720 alone.

In acute EAE and MS, the effects of FTY720 are largely mediated by immune cell trapping in peripheral lymph nodes due to egress inhibition13, thereby reducing the number of pathogenic immune cells entering the CNS. However, as we have previously demonstrated, this peripheral mode of action is less relevant in progressive stages of autoimmune CNS inflammation, where neurotoxic inflammation develops behind a sealed BBB and de novo immune cell influx is limited. In these lines, we detected no significant difference in the absolute number of T cells and IL-17-producing T cells between the groups in the draining lymph nodes. Nevertheless, IFN-γ production decreased under FTY720 treatment (Fig. 3G), suggesting a potential polarization shift in peripherally trapped immune cells. Additional intranasal IFN-β treatment did not exert detectable peripheral effects, further highlighting its primarily CNS-intrinsic mode of action. Although the NOD EAE model does not incorporate viral triggers, type I interferons can, in principle, mediate antiviral responses within the CNS. In light of increasing evidence linking Epstein–Barr virus and other viral exposures to MS pathogenesis, CNS-directed delivery of IFN-β may therefore have broader conceptual relevance beyond immunomodulation; however, antiviral mechanisms were not examined in the present study34,35.

While these data demonstrate a beneficial effect pertaining to the number of infiltrating immune cells, the progression of EAE is also critically dependent on the activation state of resident CNS cells including astrocytes, as we and others have previously shown36,37. We thus decided to focus on astrocyte-intrinsic pathways to elucidate mechanisms leading to a reduction in neuroinflammation.

FTY720 and IFN-β induce protective signaling pathways

We have previously demonstrated that both FTY720 and IFN-β can mitigate pro-inflammatory pathways in astrocytes and microglia by acting on NF-κB, SOCS2, and the aryl hydrocarbon receptor (AhR)16,23. To evaluate transcriptional changes observed under the FTY720 plus IFN-β treatment condition relative to FTY720 alone, we performed RNA sequencing on neonatal primary astrocyte cultures treated with FTY720 in the presence or absence of IFN-β. This analysis identified a transcriptional shift in astrocytes exposed to FTY720 plus IFN-β relative to FTY720 alone, with 550 differentially expressed genes (Fig. 4A). Using Ingenuity Pathway Analysis (IPA), we focused on genes differentially expressed in astrocytes treated with FTY720 plus IFN-β relative to FTY720 alone and associated with CNS inflammation and neurodegeneration (Fig. 4B). We identified several genes associated with neuroinflammation and myelination such as Cd81 and Cathepsin D (Ctsd), but also Thrombomodulin (Thbd). Previous findings have shown that the inhibition of Cathepsin D and CD81 ameliorates EAE38,39, while Thrombomodulin is crucial for myelination within the CNS40. Furthermore, we observed downregulation of Steap4, which has been reported to increase EAE severity induced by Th17 cells and contribute to neuroinflammation within the CNS41. Other downregulated genes contributing to a reduced inflammatory environment included Apoe, Cxcl5 and Lyz2, whereas upregulated genes decreasing inflammation included Crif1, Irf2 and Six1 (Fig. 4B). Together, these findings indicate a multifaceted transcriptional regulation associated with a mitigated disease phenotype. Importantly, these differentially expressed genes and pathways were linked to predicted protective downstream effects such as increased cellular homeostasis and survival, alongside decreased inflammation within the CNS and reduced cytolysis (Fig. 4C).

Fig. 4
Fig. 4
Full size image

Treatment with FTY720 plus IFN-β is associated with a protective glial transcriptional program in astrocytes. Primary murine neonatal astrocytes were treated with vehicle, FTY720 + vehicle, or the combination of FTY720 and IFN-β overnight and subsequently stimulated with IL-1β and TNF-α for 4 h. (A) Heatmap of 550 differentially expressed genes in at least two out of the groups (n = 3 biological replicates). (B) Heatmap of selected genes differentially expressed in the FTY720 plus IFN-β condition relative to FTY720 alone. (C) Ingenuity Pathway Analysis identifying predicted downstream effects in astrocytes treated with FTY720 plus IFN-β compared with FTY720 alone. (D) Fold change in mRNA expression of the indicated genes, shown as log2 (combined/FTY720). (E) Measurement of Socs1 in activated astrocytes (n = 5 biological replicates). A p-value of p < 0.05 was considered significant and determined by one-way ANOVA followed by Tukey´s multiple-comparisons test. (F) Activated primary murine astrocytes were pre- treated with a SOCS1 inhibitor (pJAK2(1001–1013); 20 µM) or a control peptide (20 µM) at the beginning of the pre-treatment phase and maintained throughout cytokine stimulation with IL-1β (10 ng/mL) and TNF-α (5 ng/mL) for 4 h. qPCR was performed for indicated genes (n = 3 biological replicates). (G) Activated primary murine astrocytes were treated with FTY720/IFN-β and a control peptide or pJak2(1001–1013). qPCR was performed for indicated genes (n = 3 biological replicates). (H) Astrocyte-conditioned medium (ACM) was generated by activating primary astrocytes with IL-1β (10 ng/mL) and TNF-α (5 ng/mL) for 24 h, followed by treatment with FTY720/IFN-β and either the control peptide or pJAK2(1001–1013) (20 µM). The medium was replaced the next day and collected after an additional 24 h. The harvested ACM was added to the lower chamber of a transwell system, and 200,000 isolated monocytes were seeded into the upper insert for a 3 h migration assay. (I) Number of migrated monocytes was counted after 3 h. P-values of p < 0.05 were considered significant and determined by Student´s t-test.

Next, we examined regulatory pathways associated with the FTY720 plus IFN-β treatment condition. Interestingly, Suppressor of Cytokine Signaling 1 (Socs1) emerged as one of the prominently regulated signaling components in astrocytes treated with FTY720 plus IFN-β relative to FTY720 alone (Fig. 4D). Socs1 is a member of the STAT induced STAT-inhibitory protein family, known for dampening pro-inflammatory signaling during inflammation42,43,44,45 and playing a crucial role in controlling CNS autoimmunity through its regulatory effects on cytokine signaling46. The therapeutic efficacy mediated by the use of SOCS1 mimetics in EAE underline the relevance of this factor in autoimmune CNS disorders47,48. In astrocytes, SOCS1 has been shown to be stabilized by OTUB1 (which was also upregulated under combined treatment compared to FTY720 alone; Fig. 4D), thereby enhancing SOCS1-dependent protective effects49. Increased Socs1 expression in astrocytes under the FTY720 plus IFN-β treatment condition relative to FTY720 alone was confirmed in additional transcriptional analyses (Fig. 4E).

To evaluate the functional relevance of SOCS1-mediated modulation of astrocytes, we used a specific SOCS1-inhibiting peptide, pJAK250, to modulate SOCS-1 signaling in astrocytes. Inhibition of SOCS1 by pJAK2 led to an increase of pro-inflammatory cytokines, in parallel with a decrease in tissue-preserving and regenerative pathways, as evidenced by reduced expression levels of Ngf and Hbegf (Fig. 4F). To further elucidate the role of SOCS1 in mediating the effects of combined FTY720/IFN-β treatment, we detected a marked increase in the expression of Il1b, Nos2, and Ccl2, alongside a reduction in Ngf expression in the presence of pJAK2 (Fig. 4G). These results indicate that SOCS1-associated signaling contributes to the downstream transcriptional effects observed under the FTY720 plus IFN-β treatment condition.

To analyze the functional relevance of these findings, we next generated astrocyte conditioned medium (ACM) by stimulating astrocytes with IL-1β and, TNF-α in the presence of FTY720/IFN-β plus either a control peptide or pJAK2. After changing the stimulation medium, fresh cell culture medium was added to the activated astrocytes, and ACM was collected 24 h later. We then performed migratory analyses using a transwell system, where splenic monocytes were added to the upper chamber and ACM was placed in the lower chamber (Fig. 4H). The protective effects of FTY720/IFN-β, mediated by SOCS1, significantly impacted the migration capacity of monocytes, as ACM generated with the combined treatment and pJak2(1001–1013) led to a higher monocyte migration compared to ACM generated with the control peptide (Fig. 4I), indicating that SOCS1-associated astrocyte signaling modulates the inflammatory properties of the astrocyte secretome. These findings suggest that differences in tissue-regenerative and pro-inflammatory cytokine expression observed under the FTY720 plus IFN-β treatment conditions are associated, at least in part, with SOCS1-dependent signaling in astrocytes.

Notably, astrocytes isolated from EAE mice were analyzed at a late disease stage, when overall inflammatory activity had already subsided under both treatment conditions, which may explain the limited transcriptional differences observed in vivo. Our in vitro experiments were therefore designed to model early glial responses under controlled conditions and to elucidate direct molecular effects of combined treatment. While these findings cannot be directly extrapolated to the in vivo setting, they reveal a SOCS1-associated transcriptional program observed in astrocytes exposed to FTY720 plus IFN-β compared with FTY720 alone, although the absence of an IFN-β monotherapy condition precludes attribution of this effect specifically to the combination. Furthermore, primary astrocytes derived from neonatal animals may differ transcriptomically and functionally from astrocytes isolated from a chronically inflamed adult CNS; thus, future studies incorporating earlier in vivo time points and glial-specific profiling approaches will be important to directly bridge in vitro and in vivo findings. Together, these data support a model in which treatment with FTY720 plus intranasal IFN-β is associated with protective effects not only through peripheral immune modulation but also through altered astrocyte-associated signaling pathways involving SOCS1. However, because IFN-β monotherapy was not included in the in vitro experiments, these data do not allow us to determine whether the observed astrocytic effects are specific to the combination or may in part be driven by IFN-β alone. A limitation of the present in vitro analyses is that an IFN-β monotherapy condition was not included. Accordingly, while the FTY720 plus IFN-β condition differed from FTY720 alone, these experiments do not permit formal attribution of the observed transcriptional effects to a combinatorial or synergistic interaction, and some effects may be driven by IFN-β alone.

FTY720 and IFN-β induce anti-inflammatory glial cell polarization

Pathogenic activities of glial cells, specifically astrocytes and microglia, are critical drivers of neurodegeneration and myelin loss in progressive MS51,52. We have previously demonstrated that microglia produce both pro- and anti-inflammatory regenerative factors, which can reciprocally act on astrocytes to induce opposing pathways53,54. Given that phosphorylated FTY720 and nIFN-β are able to cross the intact BBB, we further dissected the relevance of combined FTY720 and IFN-β treatment by first analyzing their effects on glial cells. In treated microglia, FTY720 alone and in combination with IFN-β did not significantly alter Nos2 and Tnfa expression, while Il1b expression was reduced in the presence of FTY720 in both treatment conditions as compared to vehicle (Fig. 5A). In contrast to astrocytes, expression of the neurotrophic factors Ngf, Lif, and Hbegf was reduced specifically in the combined treatment condition, whereas FTY720 alone did not significantly alter their expression compared to vehicle (Fig. 5B). In parallel, Bdnf expression was increased in treated microglia (Fig. 5C). The reduction of selected neurotrophic factors in microglia contrasts with the overall disease amelioration observed in NOD EAE, consistent with previous findings indicating that astrocytes represent the major source of tissue-protective mediators within the CNS36,37,55. To further inspect microglial behavior following FTY720 and dual-treatment stimulation, we also analyzed the activation marker Cd86; however, its expression did not show significant differences between treated microglia and controls (Fig. 5D). Together, these data suggest that the observed clinical improvement in progressive EAE is unlikely to be primarily driven by microglial intrinsic changes, further supporting astrocytes as key responder cells under combined treatment conditions. While the present study focuses on astrocyte and microglia responses, it is important to note that fingolimod has also been reported to exert direct effects on oligodendrocyte lineage cells, including promotion of oligodendrocyte differentiation and remyelination in experimental models. These observations suggest that S1PR modulation may contribute to CNS repair not only through immunomodulation and glial regulation, but also via direct effects on oligodendrocyte biology56,57. Indeed, astrocytes demonstrated reduced pro-inflammatory cytokine production upon FTY720 treatment (Fig. 5E). Importantly, additional intranasal IFN-β treatment was associated with increased expression of multiple regenerative mediators, including Hbegf, Lif, and Ngf, under combined treatment conditions (Fig. 5F). Notably, the regulation of these factors showed partial species-specific differences between murine and human glial cells ((Fig. 5A–F, I–J). Together, these findings indicate a coordinated astrocytic response to combined FTY720/IFN-β treatment, in which FTY720 attenuates pro-inflammatory cytokine expression, while IFN-β promotes the induction of tissue-protective programs.

Fig. 5
Fig. 5
Full size image

Glial cells respond to FTY720 and IFN-β treatment with cell type–specific modulation of inflammatory and regenerative programs. Primary cultures of murine glial cells were treated with vehicle, FTY720, or the combination of FTY720 and IFN-β overnight and then stimulated with IL-1β and TNF-ɑ for 4 h. qPCR of the indicated genes was performed in (A-D) microglia and (n = 3 biological replicates) and (E–F) astrocytes (n = 6 biological replicates). (G) Astrocyte-conditioned medium (ACM) was generated as described in Fig. 4H. In brief, primary astrocytes were activated with IL-1β (10 ng/mL) and TNF-α (5 ng/mL) for 24 h, followed by treatment with FTY720/IFN-β and either the control peptide or pJAK2(1001–1013) (20 µM) for 6 h in FCS-free medium. The medium was replaced the next day and collected after an additional 24 h. The harvested ACM was added to activated astrocytes under the respective conditions for subsequent qPCR analysis. qPCR was performed of indicated genes (n = 6 biological replicates). (H) Migration assay was performed as described in Fig. 4H, and the number of migrated monocytes was measured after 3 h (n = 5 biological replicates). (I) Activated human microglia (HMC3 cells; n = 6) and (J) human astrocytes (Human astrocytes Catalog #1800, ScienCell; n = 3) were treated with vehicle, FTY720 and vehicle, or the combination of FTY720 and IFN-β overnight and then stimulated with IL-1β and TNF-ɑ for 4 h. qPCR of the indicated genes was performed. All qPCR results are displayed by fold change in mRNA expression. P-values of p < 0.05 were considered significant and determined by one-way ANOVA followed by Tukey´s multiple-comparisons test.

To further investigate the functional consequences of astrocyte modulation, we added ACM generated from FTY720/IFN-β treated astrocytes to activated astrocytes in an in vitro culture system and measured the cytokine production. Within this system, we observed a reduced production of Il1b and Tnfa (Fig. 5G). In addition, ACM derived from FTY720/IFN-β–treated astrocytes reduced monocyte migration in transwell assays (Fig. 5H), indicating a less pro-inflammatory astrocyte-derived microenvironment.

Finally, we sought to validate our findings in human astrocytes and microglia. Indeed, treatment of both human microglia (Fig. 5I) and astrocytes (Fig. 5J) with FTY720 and IFN-β induced regenerative mediators including NGF, LIF and HBEGF, supporting conserved glial responses to combined treatment across species.

Taken together, these data suggest that treatment with FTY720 plus intranasal IFN-β is associated with modulation of CNS-resident glial cells with reduced pro-inflammatory cytokine expression and induction of regenerative mediators in astrocytes. Because IFN-β monotherapy was not included in the in vitro assays, these experiments do not allow formal dissection of the relative contribution of each agent to the observed glial responses. These coordinated effects on astrocytes provide a mechanistic framework supporting further exploration of CNS-targeted combinatorial strategies in progressive MS.

Materials and methods

Mice

Female NOD/ShiLtJ mice between 6–8 weeks were obtained from Charles River and kept in the mouse facility of the ZPF (animal care facility of the Klinikum rechts der Isar), where C57Bl6/J neonates were bred. Animal experiments are reported according to the ARRIVE guidelines. All procedures were performed according to the guidelines of the institutional animal care and use committee at the Technical University of Munich and were approved by the Government of Upper Bavaria (Regierung von Oberbayern; approval nos. 55.2.2–2532-2–1306 and 55.2.2–2532-2–1927-18). For terminal procedures, adult mice were deeply anesthetized with isoflurane (induction 3–4%, maintenance 1.5–2% in oxygen) until loss of pedal reflex and euthanized by exsanguination via transcardial perfusion with ice-cold PBS. Death was confirmed by cessation of circulation and absence of reflexes prior to tissue collection. Neonatal C57BL/6 J mice (P1–P5) used for primary glial cultures were anesthetized by hypothermia until unresponsive and euthanized by decapitation.

EAE induction and treatment

Mice were immunized subcutaneously with 200 µg MOG35-55 peptide dispensed in Complete Freund´s Adjuvant (CFA) followed by an intraperitoneal injection of 200 ng pertussis toxin (List Biological Laboratories, Inc.) on day 0 and day 2. CFA consisted out of 85% paraffin oil (Fluka), 15% mannide monooleate (Gerbu) and 5 mg/mL M. tuberculosis (BD). On day 30 after disease induction, corresponding to the onset of the secondary progressive disease phase, mice were treated daily with intraperitoneal injections of FTY720 (0.3 mg/kg body weight; Sigma) plus vehicle, or with the combination of FTY720 and intranasal application of IFN-β (5,000 IU; R&D Systems). Vehicle-treated mice served as controls. The selected doses of FTY720 (0.3 mg/kg, i.p.) and intranasal IFN-β (5,000 IU) were based on prior studies demonstrating robust efficacy and tolerability in chronic and progressive EAE models. In particular, these doses have been shown to achieve effective S1P receptor modulation and CNS bioavailability of IFN-β while minimizing excessive peripheral immunosuppression23,58. The dosing strategy was chosen to allow mechanistic comparison with established literature rather than to perform formal dose–response analyses.

EAE clinical scoring

Clinical symptoms were assessed and evaluated as follows: 0, no signs of disease; 1, limp tail; 2, weakness in hind legs; 3, paralyzed hind legs; 4, paralyzed hind legs and forelegs; 5, moribund.

Isolation of cells from adult mouse CNS

The CNS was digested in 3 mL 0.9 mg/mL Papain (Sigma) for 20 min and added 1 ml of 10,000 U/mL DNAse I (Sigma) and 1 mg/mL Collagenase (Sigma) for an additional 10 min. The suspension was filtered through a (100 µm cell strainer) and centrifuged over a Percoll gradient to remove excess of myelin. The cells were stained using the following fluorochrome-conjugated antibodies in a 1:100 dilution: CD3 (17A2), CD4 (RM4-5), CD11b (M1/70), CD45 (30-F11), Ly6G (1A8). Ly6C (HK1.4), CD105 (MJ7/18), CD140a (APA5), CD45R (RA3-6B2), TER-119, Oligodendrocyte Marker O4, IL-17 (TC11-18H10.1), IL-10 (JES5-16E3), IFN-γ (XMG1.2). Antibodies were purchased at R&D Systems, BD Biosciences, Biolegend or ThermoFisher Scientific. Astrocytes were sorted by CD11blow and CD45low after excluding Ly6G, CD105, CD140a, CD45R, TER-119 and Oligodendrocyte Marker O4 positive cells. Microglia were sorted by gating CD45high and CD11bintermediate and inflammatory monocytes were sorted by gating CD45high, CD11bhigh, and Ly6Chigh. For intracellular cytokine staining of T cells isolated cells were gated for CD3 and CD4 positive cells and additionally stained for IFN-γ, IL-17 and IL-10 upon fixation according to the manufacturers protocol. Absolute cell numbers were determined using counting beads (CountBright, ThermoFisher Scientific) according to the manufacturer’s instructions.

Isolation of primary astrocytes and microglia

Brains of neonatal C57Bl6/J mice (P1-5) were dissected and digested with 0.25% trypsin–EDTA for 10 min and pipetted through a cell strainer (100 µm) to obtain a single cell suspension. The cells were then plated in a poly-L-lysine precoated 175-cm2 flask. The medium was replaced after 4 days and incubated for approx. 7 days until cells were confluent (37 °C and 5% CO2). Primary astrocytes and microglia were sub-cultured for further experiments by trypsinisation the cells (0.25% trypsin–EDTA) for 2–3 min. A CD11b MACS was performed to generate astrocyte and microglia cell cultures. The CD11b negative (astrocytes) and CD11b positive cells (microglia) were then plated in the respective poly-L-lysine precoated wells.

Cell treatment

All cells were cultured in DMEM supplemented with 10% FCS, 2% PenStrep, and 4 mM L-glutamine, unless stated otherwise in the figure legend. Primary glial cells were pre-treated with FTY720 (1 µM) ± murine IFN-β (500 IU/mL) for 12 h, followed by stimulation with murine IL-1β (10 ng/mL) and TNF-α (5 ng/mL) for 4 h, unless indicated otherwise in the respective figure legend. Where specified, other compounds such as control peptide (20 µM) or pJAK2(1001–1013) (20 µM) were added during the pre-treatment phase and maintained throughout cytokine stimulation. For selected experiments, shorter 4 h single-step stimulations were performed as indicated in the figure legends. Human astrocytes and microglia were treated with equivalent concentrations of human cytokines (hIFN-β 500 IU/mL, hIL-1β 10 ng/mL, hTNF-α 5 ng/mL).

Peptide synthesis

The peptides pJAK2(1001–1013) (sequence: KLPQDKE-pY-YKVKEP) and the control peptide (sequence: KLPQDKEAAKVKEP) were synthesized at Biogenes GmbH. All peptides have a Palmitoyl-lysine end added to the N-terminus allowing cell penetration.

Flow cytometry analysis

Cells were stained using the above-mentioned antibodies in a 1:100 dilution purchased at R&D Systems, BD Biosciences, Biolegend or ThermoFisher Scientific. Cells were then analyzed using Cytoflex S (Beckman Coulter). For intracellular cytokine staining, cells were stimulated with phorbol 12-myristate 13-acetate (PMA, 50 ng/mL) and ionomycin (500 ng/mL) in the presence of Monensin (GolgiStop; BD Biosciences) for 4 h at 37 °C. Surface staining was then performed using the respective fluorochrome-conjugated antibodies. Subsequently, cells were fixed and permeabilized for 1 h at room temperature in the dark using fixation/permeabilization solution (BD Biosciences), washed, and incubated overnight (12 h) with intracellular antibodies. After an additional washing step, cells were resuspended in 200 µL PBS and analyzed by flow cytometry.

Immunofluorescence and image analysis

For immunohistochemical analyses of the central nervous system (CNS), mice were transcardially perfused with cold phosphate-buffered saline (PBS). The spinal cord was dissected and processed for immunofluorescence staining. Tissues were post-fixed in 4% paraformaldehyde (PFA) in PBS at 4 °C for 24 h, followed by cryoprotection in 30% sucrose in PBS at 4 °C overnight. The tissue was then embedded in Tissue-Tek OCT compound, snap-frozen in liquid nitrogen–cooled 2-methylbutane and stored at − 80 °C. Transverse spinal cord Sects. (10 μm thickness) were cut using a cryostat (Leica), mounted on glass slides, and stored at − 20 °C until further use.

For immunofluorescence staining of Iba1 and CD45, sections were post-fixed in acetone for 10 min at − 20 °C, rinsed once in PBS, and incubated for 1 h in blocking buffer containing 5% bovine serum albumin (BSA), 10% donkey serum, and 0.3% Triton X-100 in PBS. Slides were then incubated overnight at 4 °C with goat anti-Iba1 (1:300; Abcam; #ab289874) and rabbit anti-CD45 (1:300; Cell Signaling; #70257S) diluted in 1% BSA, 1% donkey serum, and 0.3% Triton X-100 in PBS. After three washes (5 min each), sections were incubated for 1 h at room temperature with secondary antibodies: donkey anti-rabbit IgG Alexa Fluor 488 (1:500; Thermo Fisher Scientific; #A21206), donkey anti-goat IgG Alexa Fluor 647 (1:500; Life Technologies, #A11057). Sections were washed three times for 5 min each, counterstained with DAPI (1:100,000 in 1% BSA, 1% donkey serum, and 0.3% Triton X-100 in PBS; Sigma, #D8417) for 10 min at room temperature and washed three additional times in PBS. Finally, slides were coverslipped with ProLong Gold Antifade Mountant (Thermo Fisher Scientific, #P36930) and stored at 4 °C until analysis.

For SMI32 staining, sections were post-fixed in acetone for 10 min at − 20 °C, washed in PBS for 5 min, and blocked for 30 min in blocking buffer (5% BSA, 10% donkey serum, 0.3% Triton X-100 in PBS). Slides were then incubated overnight at 4 °C with mouse anti-SMI32 antibody (1:1000; BioLegend, #801,701) diluted in 1% BSA, 1% donkey serum, and 0.3% Triton X-100 in PBS. The following day, sections were washed three times (5 min each) and incubated for 1 h with secondary antibodies: donkey anti-mouse IgG Alexa Fluor 647 (1:500; Thermo Fisher Scientific, #A32787). After washing, nuclei were counterstained with DAPI (1:100,000; 10 min at room temperature), washed in PBS, and mounted with ProLong Gold Antifade Mountant for storage at 4 °C.

For myelin visualization, FluoroMyelin Green Fluorescent Myelin Staining Kit (Thermo Fisher Scientific, #F34651) was used according to the manufacturer’s instructions.

Images were acquired using a confocal microscope under identical settings within each experiment. Mean fluorescence intensities were quantified in predefined regions of interest (ROIs) using Fiji (ImageJ, version 2.9.0; https://imagej.net/software/fiji/) in a blinded manner.

qPCR

qPCR was performed by isolating RNA using the RNAeasy Mini Kit (Qiagen), transcribing RNA to cDNA and normalizing qPCR results to Gapdh expression. Relative expression levels were then calculated using the ΔΔCt method and are presented as fold change compared to the vehicle-treated control group, which was set to 1.The following TaqMan probes for qPCR were all purchased at Applied Biosystems: Gapdh, Mm99999915_g1; Nos2, Mm00440502_m1; Csf2, Mm01290062_m1; Ccl2, Mm00441242_m1; Tnf-α, Mm00443258_m1; Il-1β, Mm00487229_m1; Vegfb, Mm00442102_m1; Socs1, Mm00782550_s1; Bdnf, Mm04230607_s1; Ngf, Mm0043039_m1; Lif, Mm00434762_g1; Hbegf, Mm00439306_m1; NGF, Hs00171458_m1; LIF, Hs01055668_m1; HBEGF, Hs00181813_m1.

Astrocyte-conditioned medium (ACM)

Isolated primary astrocytes were treated with the substances mentioned in the respective figure legend for 24 h. Cells were then additionally stimulated with 10 ng/mL IL-1β and 5 ng/mL TNF-α for an additional 24 h. The medium was removed, washed with 1 × PBS and new medium was added to the cells. This astrocyte-conditioned medium (ACM) was either utilized for stimulation experiments on activated astrocytes or used for the migration assay.

Migration assay

ACM was pipetted into 24 well plates. A transwell (Costar) was gently placed in each well and 200,000 monocytes isolated from the spleen of adult C57Bl6/J mice using the CD11b Microbead Isolation kit (Miltenyi, no. 130–049-601) were added to the transwell. Number of migrated monocytes through the filter were counted after 3 h.

Cell lines

For further in vitro experiments we used the human microglia cell line HMC3 (American Type Culture Collection [ATCC], Manassas, VA, USA; CRL-3304) and primary human astrocytes (ScienCell Research Laboratories, Carlsbad, CA, USA; Human Astrocytes, Catalog #1800). The use of the established HMC3 cell line and commercially obtained human astrocytes did not require additional ethical approval.

RNAseq analysis

Library preparation for bulk-sequencing of poly(A)-RNA was done as described previously59. Briefly, barcoded cDNA of each sample was generated with a Maxima RT polymerase (Thermo Fisher) using oligo-dT primer containing barcodes, unique molecular identifiers (UMIs) and an adaptor. Ends of the cDNAs were extended by a template switch oligo (TSO) and full-length cDNA was amplified with primers binding to the TSO-site and the adaptor. NEB UltraII FS kit was used to fragment cDNA. After end repair and A-tailing a TruSeq adapter was ligated and 3’-end-fragments were finally amplified using primers with Illumina P5 and P7 overhangs. In comparison to59, the P5 and P7 sites were exchanged to allow sequencing of the cDNA in read1 and barcodes and UMIs in read2 to achieve a better cluster recognition. The library was sequenced on a NextSeq 500 (Illumina) with 63 cycles for the cDNA in read1 and 16 cycles for the barcodes and UMIs in read2.

Gencode gene annotations M25 and the mouse reference genome GRCm38 were derived from the Gencode homepage (EMBL-EBI). Drop-Seq tools (version 1.12; https://github.com/broadinstitute/Drop-seq)60 was used for mapping raw sequencing data to the reference genome. The resulting UMI filtered count matrix was imported into R (version 4.0.5; https://www.r-project.org/) and lowly expressed genes were subsequently filtered out. Prior differential expression analysis with Limma61, the mean–variance trend was estimated with the weighted voom method followed by quantile normalization. For this purpose, a univariate model with treatment as independent variable was used. Differential testing was performed by calculating moderated t-statistics and p-values. A gene was determined as differentially regulated if the adjusted p-value was below 0.05. Heatmaps show normalized expression data after z-score transformation. Raw sequencing data is available from the European Nucleotide Archive under the accession number PRJEB48316.

(ENA Browser; https://www.ebi.ac.uk/ena/browser/view/PRJEB48316).

IPA

Data were analyzed through the use of Ingenuity Pathway Analysis. Molecules from the dataset that had an FDR < 0.05 and were associated with biological functions and/or diseases in the Ingenuity Knowledge Base were considered for the analysis. A right-tailed Fisher’s Exact Test was used to calculate a p-value determining the probability that each biological function and/or disease assigned to that data set is due to chance alone62.

Statistical analysis

Statistical analysis was performed using GraphPad Prism (version 8.2.1; https://www.graphpad.com/). The statistical tests used are mentioned in the corresponding figures and a P-value of < 0.05 was considered as significant. All error bars represent SEM unless stated otherwise in the figure legends.