CKD-506: A novel HDAC6-selective inhibitor that exerts therapeutic effects in a rodent model of multiple sclerosis

Despite advances in therapeutic strategies for multiple sclerosis (MS), the therapy options remain limited with various adverse effects. Here, the therapeutic potential of CKD-506, a novel HDAC6-selective inhibitor, against MS was evaluated in mice with myelin oligodendrocyte glycoprotein35–55 (MOG35–55)-induced experimental autoimmune encephalitis (EAE) under various treatment regimens. CKD-506 exerted prophylactic and therapeutic effects by regulating peripheral immune responses and maintaining blood–brain barrier (BBB) integrity. In MOG35–55-re-stimulated splenocytes, CKD-506 decreased proliferation and downregulated the expression of IFN-γ and IL-17A. CKD-506 downregulated the levels of pro-inflammatory cytokines in the blood of EAE mice. Additionally, CKD-506 decreased the leakage of intravenously administered Evans blue into the spinal cord; CD4+ T cells and CD4−CD11b+CD45+ macrophage/microglia in the spinal cord was also decreased. Moreover, CKD-506 exhibited therapeutic efficacy against MS, even when drug administration was discontinued from day 15 post-EAE induction. Disease exacerbation was not observed when fingolimod was changed to CKD-506 from day 15 post-EAE induction. CKD-506 alleviated depression-like behavior at the pre-symptomatic stage of EAE. In conclusion, CKD-506 exerts therapeutic effects by regulating T cell- and macrophage-mediated peripheral immune responses and strengthening BBB integrity. Our results suggest that CKD-506 is a potential therapeutic agent for MS.


CKD-506 alleviated the clinical symptoms of EAE in mice subjected to the prophylactic regimen.
To investigate the therapeutic efficacy and effective dose of CKD-506 for MS, the efficacy of CKD-506 was compared with that of fingolimod in MOG  -induced EAE under the prophylactic regimen. Previous studies have demonstrated that CKD-506 is effective at doses in the range of 10-30 mg/kg bodyweight in rodent models of inflammatory bowel disease, SLE, and RA. Hence, in this study, mice were treated with CKD-506 at a dose range of 3-100 mg/kg bodyweight to examine the effective dose for the prophylactic regimen. Additionally, the dose of fingolimod was determined based on the blood exposure level in humans. The mouse equivalent dose of fingolimod is 0.1-0.3 mg/kg bodyweight 5 . CKD-506 dose-dependently decreased the clinical scores and delayed the disease onset time. The clinical scores were significantly lower, while the disease onset was significantly delayed in the groups treated with CKD-506 at doses of ≥ 30 mg/kg bodyweight than in the vehicle-treated group [ Fig. 1a, Table 1; F(8, 144) = 62.65; p < 0.001; one-way Analysis of variance (ANOVA)]. Furthermore, the mean clinical score and final clinical score in the CKD-506 (30 mg/kg bodyweight)-treated group did not differ from that in the fingolimod (1 mg/kg bodyweight)-treated group [ Fig. 1b; F(8, 144) = 45.10; p < 0.001; one-way ANOVA, and Table 1; F(8, 144) = 62.65; p < 0.001; one-way ANOVA].
CKD-506 alleviated spinal cord damage in mice with EAE. Histological analysis was performed to examine the therapeutic effects of CKD-506. The spinal cord section was stained with Luxol fast blue (LFB) to examine the demyelination pattern. The LFB-stained area in the vehicle group (36.25 ± 14.19%) was approximately 36% lesser than that in the non-immunized groups (0.07 ± 0.04%). The area with decreased LFB staining intensity was correlated with immune cell infiltration (Fig. 1e). The LFB-stained area in the CKD-506-treated and fingolimod-treated groups was approximately 5-15% lower than that in the vehicle-treated group [ Fig. 1d; F(8, 68) = 3.575; p = 0.0016; one-way ANOVA].

CKD-506 decreased the infiltration of T cells and macrophages in the spinal cord of mice with EAE. The infiltration of peripheral immune cells, such as T cells and macrophages, into the spinal cord can
induce adverse clinical symptoms, including myelin and axonal pathology in EAE. Thus, the number of immune cells in the spinal cord was examined on day 29 post-EAE induction. The H&E-stained spinal cord (L3-L5) sections of the vehicle group exhibited enhanced infiltration of immune cells [ Fig. 1c; F(8, 144) = 26.14; p < 0.001; one-way ANOVA]. The highest infiltration of immune cells was observed in the peripheral region of white matter (Fig. 1e). CKD-506 decreased the infiltration of immune cells into the spinal cord. The decrease in the infiltration of immune cells in the spinal cord of the CKD-506 (30 mg/kg bodyweight)-treated group was similar to that in the spinal cord of the fingolimod (0.1 mg/kg bodyweight)-treated group (Fig. 1c) [Fig. 1c; F(8, 144) = 26.14; p < 0.001; one-way ANOVA]. Immunohistochemical (IHC) analysis revealed that the expression of CD3 (T cell marker) and CD68 (macrophage/microglia marker), which were mainly localized in the parenchyma of white matter, in the vehicle group was higher than that in the non-immunized groups (Fig. 2d). Moreover, the expression of PAX5 (B cell marker) was upregulated in the spinal cord of the vehicle group. However, Figure 1. CKD-506 dose-dependently decreased the clinical scores, inflammatory cell infiltration, and demyelination in the experimental autoimmune encephalitis (EAE) mouse model under the prophylactic regimen. C57BL/6 mice were orally administered CKD-506 (3, 10, 30, or 100 mg/kg bodyweight) or fingolimod (0.1, 0.3, or 1 mg/kg bodyweight) daily from day 6 post-myelin oligodendrocyte glycoprotein  immunization (a-e). The clinical score (a), maximum clinical score (n = 6-24 per group) (b), inflammatory cell infiltration [n = 6 per group for the fingolimod (1 mg/kg)-treated group; n = 8 per group for the fingolimod (0.1 mg/kg)-treated group; n = 9 per group for the fingolimod (0.3 mg/kg)-treated group; n = 17 per group for the non-immunized groups and CKD-506 (3 mg/kg)-treated group; n = 24 per group for the other groups] (c), and demyelination [n = 6 per group for the fingolimod (1 mg/kg)-treated group; n = 8 per group for the non-immunized groups, CKD-506 (3 or 10 mg/kg)-treated group, and fingolimod (0.1 or 0.3 mg/kg)-treated group; n = 9 per group for the other groups] (d) were examined. (e) Representative images of H&E-stained and Luxol fast blue-stained sections. Data are presented as mean ± SD; one-way ANOVA, followed by multiple comparisons post-hoc test with Tukey's test for (b, d); Kruskal-Wallis test, followed by multiple comparisons post-hoc test with Dunnett's test for (c). ###p < 0.001, non-immunized groups versus vehicle group; **p < 0.05 and ***p < 0.001, drug-treated group versus vehicle group; aaap < 0.001, CKD-506-treated group versus fingolimod (0.1 mg/kg)-treated group; bp < 0.05, bbbp < 0.001, CKD-506-treated group versus fingolimod (0.3 mg/kg)-treated group; ccp < 0.01, cccp < 0.001, CKD-506-treated group versus fingolimod (1 mg/kg)-treated group. N, non-immunized groups; V, vehicle group; Fin, fingolimod-treated group.  CKD-506 downregulated the levels of Th1 cell-related and macrophage-related pro-inflammatory cytokines in the spinal cord and peripheral blood of mice with EAE. Next, the effect of CKD-506 on the immune-mediated inflammatory response was examined. The levels of immune cell-related cytokines in the spinal cord and peripheral blood were analyzed using the multiplex assay. The levels of Th1 cell-related and macrophage-related cytokines, such as IFN-γ, IL-12, TNF-α, and IL-1β were upregulated in the spinal cord and blood of the vehicle group on day 29 post-EAE induction. However, the levels of IL-17, secreted by activated Th17 cells, and IL-4, secreted by the Th2 cells, were not affected in the blood and spinal cord of the vehicle group. Treatment with CKD-506 downregulated the levels of Th1 cell-and macrophage-secreted cytokines in the spinal cord and blood of the vehicle group (Fig. 3). However, treatment with fingolimod decreased the levels of IFN-γ and IL-12 but not those of TNF-α, and IL-1β in the peripheral blood and spinal cord of the vehicle group. These findings suggest that CKD-506 suppressed Th1 cell-and M1 macrophage-mediated immune responses, whereas fingolimod suppressed only T cell-mediated immune responses.

CKD-506 inhibited the proliferation of peripheral immune cells in the MOG-stimulated splenocytes.
The progression of EAE is associated with the activation of peripheral immune responses against MOG, including the differentiation, activation, and expansion of immune cells and the secretion of various pro-inflammatory cytokines in the peripheral lymphoid organs 21,22 . To evaluate the role of suppressed peripheral immune response activation in spinal cord damage, the peripheral autoimmune response against MOG was examined using CCK8 and cytokine multiplex assays and the proportion of CD3 + T cells in the blood of EAE mice was assessed. The results of the CCK8 assay revealed that CKD-506 (     were orally administered CKD-506 (10, 30, or 100 mg/kg) and fingolimod (0.3 mg/kg) daily from day 6 postmyelin oligodendrocyte glycoprotein  immunization and pro-inflammatory cytokine levels in their blood or spinal cord were evaluated on day 29 post-induction (a, b). The levels of IFN-γ, IL-12, IL-17A, IL-1β, IL-4, and TNF-α in the plasma (a) and supernatant from spinal cord homogenate (b) were quantified using the multiplex cytokine assay. Data are presented as mean ± SD; one-way ANOVA, followed by Dunnett's post-hoc for IFN-γ, IL-12, IL-1β, TNF-α from (a) and IFN-γ, IL-17A, IL-1β, TNF-α from (b); Kruskal-Wallis test, followed by post-hoc test with Dunnett's test for IL-17A, IL-4 from (a) and IL-12, IL-4 from (b). #p < 0.05, ##p < 0.01, and ###p < 0.001, non-immunized groups versus vehicle group; *p < 0.05, **p < 0.01, and ***p < 0.001, drug-treated group versus vehicle group. N, non-immunized groups; V, vehicle group; Fin, fingolimod-treated group.   www.nature.com/scientificreports/ CKD-506 exerted therapeutic efficacy after drug discontinuation and delayed the exacerbation of EAE after fingolimod administration was discontinued. Clinically, the discontinuation of fingolimod administration results in the exacerbation of MS symptoms. Thus, this study examined the therapeutic efficacy of CKD-506 after the discontinuation of drug administration, as well as the preventive effects of CKD-506 on the exacerbation of EAE symptoms after the discontinuation of fingolimod administration. The administration of CKD-506 and fingolimod was discontinued on day 15 post-EAE induction. The discontinuation of CKD-506 did not affect EAE symptoms, whereas that of fingolimod exacerbated EAE symptoms (Fig. 6c). The final clinical score in the CKD-506-treated group was significantly lower than that in the vehicle group

CKD-506 did not reduce the white blood cell (WBC), red blood cell (RBC), and platelet (PLT) counts in normal mice unlike LBH-589, a pan-HDAC inhibitor. Previous studies have demonstrated
that HDAC inhibition has a suppressive effect on the bone marrow, and induces severe thrombocytopenia, leukopenia, or anemia 11 . Thus, WBCs, RBCs, and PLTs were evaluated in the blood of the normal mice to confirm the effects of CKD-506 (30 mg/kg, per oral) in comparison to LBH-589 (10 mg/kg, intraperitoneal). Following a daily administration of both drugs for 3 consecutive days, LBH-589 significantly reduced WBC and PLT counts, but not that of RBCs [Supplementary data; F(2, 12) = 179.4; p < 0.0001 for WBC, F(2, 12) = 153.0; p < 0.0001 for PLT; one-way ANOVA]. However, CKD-506 (30 mg/kg) showed no effects on the levels of these cells.

CKD-506 alleviated depression at the pre-symptomatic stage of EAE. Clinical studies have
reported that the administration of new drugs to patients with MS can induce depression. More than 50% of the patients with MS exhibit depression during their lifetime 24 . Thus, a tail suspension test (TST) was performed and the expression levels of acetylated α-tubulin and α-tubulin in the brain were examined on day 8 post-EAE induction to evaluate the CNS-penetrating ability of CKD-506 to alleviate depression-like behavior in mice with EAE during the pre-symptomatic stage. In TST, mice with EAE exhibited a decreased period of immobility, which indicates depression-like behavior in this model 25 . Moreover, the CKD-506-treated group exhibited a higher immobility latency than the vehicle group. The immobility latency value in the CKD-506-treated group was similar to that in the non-immunized groups [ Fig. 7a; F(3, 47) = 3.944; p = 0.0137; one-way ANOVA]. The level of acetylated α-tubulin in the CKD-506 (10 mg/kg bodyweight)-treated group was significantly higher than that in the vehicle-treated group [ Fig. 7b, C; F(6,14) = 9.071; p < 0.001; one-way ANOVA]. This suggests that CKD-506 (10 mg/kg bodyweight) could penetrate the brain. These findings demonstrate that CKD-506 could alleviate depression-like behavior by exerting therapeutic effects in the brain.

Discussion
Over the last decade, the treatment outcomes of patients with MS have markedly improved. However, disease relapse and progression are observed in several patients. In some cases, there are limited therapeutic strategies available, which highlights the need for the continuous monitoring of the risk of developing side effects. Thus, there is a need to develop novel therapeutic strategies against MS. In this study, the potential therapeutic effects of CKD-506, a novel HDAC6-specific inhibitor, against MS were examined under various treatment regimens using the MOG 35-55 -induced EAE mouse model. Previous studies have demonstrated that pan-HDAC inhibitors could alleviate the symptoms of several autoimmune diseases by enhancing the anti-inflammatory response 8,26 . However, pan-HDAC inhibitors have a narrow therapeutic spectrum, which limits their application as therapeutics for MS 11,26 . Compared with pan-HDAC inhibitors, HDAC6 inhibitors have a better safety profile since they predominantly acetylate cytoplasmic non-histone proteins 27,28 . Moreover, phase I clinical trials for CKD-506, a highly selective HDAC6 inhibitor, showed no significant safety issue (EudarCT number:2016-002816-42), and in this study as well, CKD-506, at the therapeutic dose, did not induce bone marrow suppression, unlike LBH-589. Table 2. Values of the mean clinical score, maximum clinical score, and final clinical score in the CKD-506-and fingolimod-treated experimental autoimmune encephalitis (EAE) groups under the therapeutic regimen. Data are presented as mean ± SD (*p < 0.05, **p < 0.01 and ***p < 0.001, drug-treated group vs. vehicle group). Results were analyzed using Kruskal-Wallis test, followed by multiple comparisons post-hoc test with Dunnett's test for maximum clinical score and using one-way ANOVA followed by Dunnett's post-hoc test for mean clinical score and final clinical score. *The mean clinical score was calculated as follows: The mean clinical score = the accumulated score / evaluated days. *Fin, Fingolimod-treated group. www.nature.com/scientificreports/ CKD-506 also markedly alleviated the clinical symptoms and inhibited inflammation and demyelination in the spinal cord of the MOG 35-55 -induced EAE mouse model under both prophylactic and therapeutic regimens. The therapeutic effects of CKD-506 were comparable with those of fingolimod, which is a second-line therapeutic for MS. Currently, patients resistant to first-line therapeutics are administered fingolimod, which decreases the relapse rate by 60% within 1 year after treatment 5 .

(A) Values for clinical parameters (mean ± SD) in therapeutic mode
CKD-506 directly suppressed the proliferation of pathogenic immune cells in the MOG 35-55 -stimulated splenocytes. It also decreased the infiltration of T cells and macrophages as well as the levels of pro-inflammatory cytokines in the blood or the spinal cord of EAE mice. These findings suggest that CKD-506 regulates the proliferation of the pathogenic subpopulations of immune cells and suppresses their activity by reducing the levels of pro-inflammatory cytokines such as IL-12, IL-17A, IFN-γ, IL-1β, and TNF-α in EAE. The activation and expansion of the pathogenic immune cells from the peripheral lymphoid organs are critical for the development of MS in the initial relapse process 29,30 . Previous studies have also shown that the number of pathogenic immune cells, such as certain subpopulations of T cells and macrophages, is elevated in the blood and CNS lesions of patients with MS during the relapse phase 31,32 .
The levels of IL-12, IFN-γ, and IL-17A, mainly expressed in the Th1/17 cells, and of IL-1β and TNF-α, mainly expressed in the microglia/macrophages are usually higher in the blood, CSF, and lesion sites of the patients with MS, compared with healthy individuals and those cytokines have close relations with the progression of MS [33][34][35][36] . IL-12 is an enhancer of the polarization of T cells to Th1 cells, a pathogenic Th subset that contributes to the progression of MS directly by damaging the myelin sheath of the oligodendrocytes and indirectly by enhancing the activation of potentially pathogenic macrophages through the secretion of TNF-α or IFN-γ 37 . MOG-induced symptoms and CNS lesions are alleviated in Il12 knockout mice 38 . IFN-γ is a hallmark of Th1 cells inducing inflammation and autoimmune responses, such as those seen in MS. The effects of IFN-γ on MS pathology are controversial and disease stage dependent. However, the expression levels of IFN-γ in patients with MS are usually higher than those in healthy individuals 33 , and blocking IFN-γ significantly improved the symptoms in secondary progressive MS 39 . Thus, Lowering IFN-γ could be one treatment strategy for MS. IL17A, a key cytokine of Th17, has also demonstrated efficacy against MS in a pilot study with its antibody, sekukinumab 34 , while the blockade of IL-17A ameliorated the disease symptoms in various models of EAE mice 40,41 . Although www.nature.com/scientificreports/ the clinical effects of IL-1β and TNF-α are not clear since their respective antibodies show no positive response in clinical studies, various studies have suggested their important roles in the pathogenic responses of EAE mice and those seen in MS. Clinical symptoms of EAE were significantly attenuated in IL-1 receptor-deficient and IL-1β-deficient mice 35 . Decreased levels of TNF-α improved EAE symptoms by protecting the BBB permeability and inhibiting the immune cell migration in various conditions of TNF-α deficit 36 . BBB dysfunction also contributes to the progression of EAE and MS by enabling the migration of activated pathogenic cells and toxic molecules into the brain, which induces neuronal inflammation, demyelination, and neural cell death 23 . In this study, CKD-506 protected the BBB by upregulating the expression of occludin, a major component of BBB. Occludin, a transmembrane protein, enhances the tight junction integrity by forming a scaffold with ZO-1. Previous studies have also demonstrated that the downregulation of occludin enhances BBB permeability 42 . CKD-506 downregulates the levels of ICAM-1, VCAM-1, and IP-10 in LPS-induced human peripheral blood mononuclear cells and an RA rat model 20 . Additionally, CKD-506 regulates the morphology of immune cells by rearranging tubulin and actin and consequently, inhibits their migration 20,43 . In this study, a decrease in the number of macrophages in the spinal cord also demonstrated the inhibitory effects of CKD-506 on immune cell migration. Clinical studies on natalizumab have demonstrated that the inhibition of pathogenic immune cell chemotaxis and migration into the brain decreases the risk of MS relapse 44 . These findings indicate that CKD-506 could alleviate MS by enhancing the BBB integrity and regulating the chemo-attraction and migration of immune cells.
The US Food and Drug Administration has warned that fingolimod discontinuation may exacerbate disease symptoms, which was also reported in a clinical study 45 . We found that fingolimod discontinuation rapidly exacerbated EAE symptoms and increased the number of leukocytes and levels of Th1-related cytokines such as IFN-γ in the blood after 3 days. This suggests that the rapid deterioration of EAE symptoms may be attributed to the increased escape of pathogenic T cells into the blood from the lymphoid organ immediately after fingolimod discontinuation. However, the therapeutic effects of CKD-506 were observed even after the drug was discontinued. TNF-α and IFN-γ were maintained at lower levels 3 days after drug discontinuation. However, it remains unclear as to why the efficacy of CKD-506 was maintained.
Additionally, CKD-506 delayed the exacerbation of symptoms and prevented the rapid escape of leukocytes after the discontinuation of fingolimod. The mechanisms underlying the CKD-506-mediated delay in symptom exacerbation and inhibition of rapid leukocyte escape have not been elucidated. However, the findings suggest that CKD-506 could delay the progression of EAE even when the number of active pathogenic cells markedly increased in the systemic circulation, which is a characteristic of the relapse phase of MS.
CKD-506 increased the levels of acetylated α-tubulin in the brain and alleviated the depression-like behavior at the pre-symptomatic stage of the EAE model. Currently, 50% of the patients with MS suffer from depression, which has no effective treatment. Thus, the efforts to treat MS-associated depression are on-going. Furthermore, previous studies on Tnfar and Il-1β knockout mice demonstrated that the downregulation of these cytokines alleviates depression-like behaviors 46,47 . The upregulation of these cytokines in the hypothalamus has been shown to affect emotional behavior through the upregulation of corticosterone, which is secreted from the hypothalamic-pituitary-adrenal/thyroid axis, in mice with EAE 25 . Various neurological diseases are associated with defects in axonal transport, which induce neurological defects and lead to emotional and cognitive dysfunction 17 . Moreover, studies on mice with ALS or CMT have reported that HDAC6 inhibitors reverse axonal transport defects by upregulating the acetylation of α-tubulin 48,49 . These results suggest that CKD-506 could alleviate depression by downregulating TNF-α and IL-β expression and enhancing axonal transport through the upregulation of acetylated α-tubulin in the CNS.
This study has some limitations. We did not to confirm the subpopulations of the T cells and macrophages in the blood or spinal cord as the cytokine analysis carried out was not sufficient to confirm the subpopulation of these cells; different subpopulations of T cells and macrophages have vastly different effects on the disease. Furthermore, although the efficacy of CKD-506 was maintained after its discontinuation, with the observation of a delay in symptom exacerbation, the related mechanisms have not been elucidated. Lastly, only one behavioral test was performed for the evaluation of the depression-like features in EAE mice. Animal behavioral assessment is heavily influenced by various factors. Thus, more than two types of assessments would ensure the accuracy of the results.
Summarily, this study demonstrated that CKD-506 markedly alleviates EAE in the rodent model through the regulation of the activation and proliferation of pathogenic subpopulations of T cells and M1 macrophages in the peripheral lymphoid tissue and the maintenance of BBB integrity. Additionally, the therapeutic efficacy of CKD-506 was maintained even after its discontinuation. Furthermore, CKD-506 could delay disease progression and decrease the number of activated pathogenic cells in systemic circulation. Finally, CKD-506 alleviated depression-like behavior in mice with EAE. These findings suggest that CKD-506 is a potential therapeutic strategy against MS. www.nature.com/scientificreports/ facility in CKD Research Institute labs (Yongin, Korea). All methods with animals were performed in accordance with the relevant guidelines and regulations of the Institutional Animal Care and Use Committee of the Laboratory Animal Center at Chong-gun dang, Korea (approval number: S-20-009) and the study was carried out in compliance with the ARRIVE guidelines.

Methods
EAE induction and symptom evaluation. EAE was induced as described previously 50  Histological analysis. Mice were anesthetized with isoflurane (Hana Pharm Inc., Korea) and transcardially perfused with saline (3 mL). The spinal cord was excised and fixed with 10% neutral-buffered formalin at room temperature. The samples were cut into pieces, processed, embedded in paraffin, and sectioned to 4-μm thick sections using a sliding microtome. The sections were stained with H&E (BBC Biochemical, Mount Vernon, WA, USA) or LFB (VitroVivo Biotech, # VB-3006) and observed under a light microscope. The severity of immune cell infiltration with the H&E-stained sections was assessed and scored in the meninges, parenchyma, and vessels of the spinal cord by two histologists who were blind to the treatments and grouping. Then, the scores in the three regions were added. The scoring to determine the histological score was performed as follows: for meninges and parenchyma, no infiltrating cells = 0; few infiltrating cells = 1, numerous infiltrating cells = 2, and widespread infiltration = 3; for vessels, no cuffed vessel = 0; one or two cuffed vessels per section = 1, three to five cuffed vessels per section = 2, and more than five cuffed vessels per section = 3. Additionally, the demyelinated and total areas of white matter were measured using ImageJ software. The percentage of demyelination was calculated as follows: demyelinated area (%) = [(demyelinated area in white matter) / (total white matter area) × 100].
IHC analysis. The paraffin-embedded sections (4 μm) of the spinal cords were mounted on glass slides. IHC analysis of the sections was performed using an automated slide preparation system (Benchmark XT; Ventana Medical Systems Inc., Tucson, AZ, USA). Deparaffinization, epitope retrieval, and immunostaining were performed using cell conditioning solutions and the BMK ultraVIEW diaminobenzidine detection system (Ventana Medical Systems Inc., Tucson, AZ), following the manufacturer's instructions. The tissue sections were stained with anti-CD3 (#ab16669, Abcam), anti-CD68 (#ab125212, Abcam), and anti-PAX5 (#ab109443, Abcam) primary antibodies. The positive signals were amplified using ultraVIEW copper. The sections were counterstained with hematoxylin and bluing reagent. The number of CD3-, CD68-, and PAX5-immunoreactive cells in the white matter and pia mater of the spinal cord was examined using a light microscope and normalized to the total area of white matter and pia mater.
Cytokine analysis. Mice were anesthetized with isoflurane and the blood sample was collected from the inferior vena cava on day 29 post-EAE induction. Systemic perfusion was performed by infusing saline (3 mL) into the left ventricle of the heart through the opened vena cava. Lumbar spinal cord (L3-L5) was excised and homogenized in ice-cold homogenizing buffer (PBS containing 0.1% Triton X-100 and 1% protease inhibitor mixture). The blood and homogenized spinal cord samples were then centrifuged at 12,000 rpm and 4 °C for 15 min. The levels of cytokines in the supernatant were measured using the multiplex cytokine kits (Millipore, #MHSTCMAG-70 K), following the manufacturer's instructions. The protein concentration was measured using www.nature.com/scientificreports/ bicinchoninic acid protein assay reagents (Thermo Scientific Inc., # VC297454). The cytokine levels in the spinal cord homogenate were normalized with the protein concentration.
Proliferation assays and cytokine analysis. Single-cell suspensions were prepared from the spleen of EAE mice collected on day 12 post-EAE induction. The cells (2 × 10 5 cells/well) were cultured in 96-well plates in the presence of the MOG peptide (10, 20, and 40 μg/mL; Hooke laboratories). The samples were processed in duplicate plates and the experiments were repeated four times. After 48 h, the cells in one plate were incubated with CCK8 solution and the absorbance at 500 nm was measured after 3 h. The culture in the other plate was centrifuged at 12,000 rpm and 4 °C to obtain the supernatant for cytokine analysis. The cytokine level in the supernatant was evaluated as mentioned above.
Analysis of BBB permeability. BBB permeability was evaluated using EB dye as previously described 51 .
On day 21 post-EAE induction, the mice were intravenously injected with EB dye (2% in PBS; 4 mL/kg bodyweight; Sigma), which was allowed to circulate for 1 h. The mice were then perfused with saline (25 mL) and the spinal cord was excised and weighed. The tissues were homogenized in formamide (0.3 mL; Sigma Aldrich) and incubated overnight at 55 °C for EB extraction. The absorbance value of the EB in the extracts was measured at 500 nm. Additionally, the absorbance value was adjusted to the volume of EB using the standard curve prepared using different concentrations of EB. The final value was represented by dividing the volume of EB in the tissue by the tissue weight.
Complete blood count (CBC). CBC was measured using a CBC analyzer (Sysmex, #XN-1000 V) to determine the changes in the number of leukocytes in the vehicle group subjected to the drug discontinuation regimen and the levels of CKD-506 and fingolimod in the blood of the vehicle groups subjected to the drug discontinuation and drug change regimens.
TST. Depression behavior in mice was evaluated using TST. The mice were suspended by attaching the tip of the tail with adhesive tape (approximately 1 cm) 10-cm above the floor. To assess the depression-like behavior, immobility time (in seconds) of mice was recorded and evaluated for 5 min. To reduce the experimental bias, two analysts performed the test blindly.
Statistical analysis. The details of the statistical analysis carried out, including statistical tests, exact p-values, and sample size are provided in the Results section and figure legends. Briefly, the means between multiple experimental groups were analyzed using Kruskal-Wallis statistic, one-way or two-way ANOVA, followed by post-hoc Dunnett's, Tukey's, or Bonferroni's test. Student's t-test was performed to compare the means of two groups. The differences were considered significant at p < 0.05. Data are expressed as mean ± SD. All statistical analyses were performed using GraphPad Prism (ver 9.0).

Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. www.nature.com/scientificreports/