Experimental Neurology 257 (2014) 50–56

Contents lists available at ScienceDirect

Experimental Neurology

j ourna l homepage: www.e lsev ie r .com/ locate /yexnr

Regular Article

Neuroprotective dimethyl fumarate synergizeswith immunomodulatoryinterferon beta to provide enhanced axon protection inautoimmune neuroinflammation

Christiane Reick a,b,1, Gisa Ellrichmann a,⁎,1, Jan Thöne a, Robert H. Scannevin c, Carsten Saft a,Ralf A. Linker d,2, Ralf Gold a,b

a Department of Neurology St. Josef-Hospital, Ruhr-University Bochum, Germanyb International Graduate School of Neuroscience, Ruhr-University Bochum, Germanyc BiogenIdec, Inc. Cambridge, 02142 MA, USAd Department of Neurology, University of Erlangen, Germany

⁎ Corresponding author at: Department of NeuroloD-44791 Bochum, Germany. Fax: +49 234 509 2414

E-mail address: gisa.ellrichmann@rub.de (G. Ellrichma1 These authors share first authorship.2 Current address.

http://dx.doi.org/10.1016/j.expneurol.2014.04.0030014-4886/© 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 15 December 2013Revised 26 March 2014Accepted 4 April 2014Available online 13 April 2014

Keywords:Multiple sclerosisEAEImmunomodulationNeuroprotectionInterferon-betaDimethyl fumarateCombination therapy

Introduction: Despite recent advances in development of treatments for multiple sclerosis, there is still an unmetneed for more effective and also safe therapies. Based on the modes of action of interferon-beta (IFN-β) anddimethyl fumarate (DMF), we hypothesized that anti-inflammatory and neuroprotective effects may synergizein experimental autoimmune encephalomyelitis (EAE).Methods: EAEwas induced in C57BL/6mice by immunizationwithMOG35–55-peptide.Murine IFN-βwas injecteds.c. every other day at 10.000 IU, and DMF was provided at 15 mg/kg by oral gavage twice daily. Control micereceived PBS injections and were treated by oral gavage with the vehicle methylcellulose. Mice were scoreddaily by blinded observers and histological, FACS and cytokine studies were performed to further elucidate theunderlying mechanism of action.Results: Combination therapy significantly ameliorated EAE disease course in comparison to controls and mono-therapy with IFN-β. Histological analyses showed a significant effect on axon preservation with almost twice asmuch axons present in inflamed lesions as compared to control. Remarkably, the effect on axonal preservationwas more pronounced under combination therapy than with both monotherapies. Neither monotherapy nor

combination therapy demonstrated modulation of cytokines and frequency of antigen presenting cells.Discussion: Combination of IFN-β and DMF resulted in greater beneficial effects with improved tissue protectionas compared to the respective monotherapies. Further combination studies of these safe therapies in humandisease are warranted.

© 2014 Elsevier Inc. All rights reserved.

Introduction

Multiple sclerosis (MS) is a chronic neurological disease of thecentral nervous system (CNS) of presumed autoimmune origin that ischaracterized by inflammation-driven demyelination, and subsequentaxonal damage and progressive neurological deficits. Studies in MSand experimental autoimmune encephalomyelitis (EAE), an establishedanimal model mimicking several pathophysiological features of MS,demonstrated that demyelinating lesions are accompanied by T celland macrophage infiltration (Barnabe-Heider and Miller, 2003; Goldet al., 2006; Weiner, 2009). Subsequent to T cell and macrophage

gy, Ruhr University Bochum,.nn).

activation and infiltration, these cells release pro-inflammatory cyto-kines, interact with antigen presenting cells (APCs) and are involvedin destruction of the myelin sheet, which is associated with chronicaxonal damage (Herrero-Herranz et al., 2008).

For as long as 20 years interferon-beta (IFN-β) has been anestablished immunomodulatory agent in patients with relapsing remit-ting MS (RRMS). In phase III studies IFN-β showed beneficial effects onrelapse rate, disease progression and development of new CNS lesionsas shown by magnetic resonance imaging (MRI) (Burks, 2005; Jacobs,1996). So far, different mechanisms of action have been described forIFN-βwith a focus on downregulation of macrophage and microglia ac-tivity (Mendes and Sa, 2011).

Dimethyl fumarate (DMF; study name BG-12) has shown beneficialeffects in both EAE and inmodels of neurodegenerative diseases as wellas in two phase III studies in patients with RRMS (Ellrichmann et al.,2011; Ghoreschi et al., 2011; Linker et al., 2011). So far, different mech-anisms of action, including neuroprotection, have been described. DMF

51C. Reick et al. / Experimental Neurology 257 (2014) 50–56

was shown to mediate neuroprotection by inducing the nuclearfactor E2-related factor 2 (Nrf2) pathway and thereby addressingtoxic-oxidative stress (Ghoreschi et al., 2011; Johnson et al., 2008;Williamson et al., 2012). In APCs, DMF induced glutathione depletionand thus stimulated type II dendritic cells (DCs)with an impaired secre-tion of interleukin- (IL) 12 and IL-23. Simultaneously, DMF induced bothIL-4 producing T helper (Th) 2 cells and generated type II dendritic cells(DCs) that produce IL-10 (Ghoreschi et al., 2011).

Based on their different modes of action we postulated that IFN-βand DMF synergize in early and early chronic EAE. We show that thecombination of IFN-β andDMF resulted in a synergistic effect at the clin-ical level, as reflected by decreased disease severity of EAE. This effectwas confirmed histologically by reduced inflammation and in particularpreservation of axonal density in the spinal cord of EAE mice. The ben-eficial effects of combination therapy with IFN-β and DMF may resultfrom effects on different pathophysiological pathways in EAE.

Material and methods

Animal models

8–10week old female C57BL6/J micewere obtained from the HarlanLaboratories (Borchen, Germany, nowadays Rossdorf, Germany) andkept under standardized, pathogen free conditions at the local animalfacility, Ruhr-University, Bochum, Germany. Food and water weregiven ad libitum to all animals. Experiments were approved by thelocal authorities for animal experimentation (approval no. 8.87-50.10.32.08.032).

EAE induction and treatment

An emulsion consisting of myelin oligodendrocyte glycoprotein 35–55 (MOG35–55) in phosphate buffered saline (100 μg) and an equalvolume of complete Freund's adjuvant containing 100 μg of Mycobacte-rium tuberculosis (Difco, BD Bioscience) was prepared and subcutane-ously injected at the tail base of 8–10 week old female C57BL6/J mice.In addition to immunization with MOG35–55 emulsion, on days 0 and2, 100 ng pertussis toxin (List, Quadratec) was administered i.p. DMF-treated animals received 15 mg/kg body weight (bw) DMF dissolvedin 0.08% methylcellulose twice a day per oral gavage. Mice treatedwith murine IFN-β received an s.c. injection of 10.000 units everyother day. IFN-β was dissolved in endotoxin free recombinant mouseserumalbumin (MSA) 1%. Preventive treatment started at theday of im-munization. Animals were weighed and clinically monitored daily. Theblinded observer used a 10-scale score (Linker et al., 2002) for clinicalsigns. Mice were sacrificed on day 10 for ELISA and FACS and on day18 for histological analysis. For ELISA and FACS spleens and lymphnodes (superficial cervical lymph nodes, deep cervical lymph nodes,lumbar lymph nodes)were removed to prepare a single cell suspension.For histological studies, mice were deeply anesthetized with ketamineand were transcardially perfused with 4% paraformaldehyde.

Immunohistochemistry

Immunohistochemistry was performed on 5 μm paraffin embeddedspinal cord cross sections (lumbar, thoracal and cervical part). If neces-sary, antigen unmasking was performed via boiling of sections in citricacid buffer. After inhibition of nonspecific binding with 10% bovineserum albumin (BSA), sections were incubated overnight at 4 °C withthe appropriate primary antibody in 1% BSA. After blocking of endoge-nous peroxidase with 0.3% H2O2, the peroxidase-based ABC detectionsystem (Vectastain, Vector Laboratories, via Linaris, Wertheim,Germany) was employed with diaminobenzidine as the chromogenicsubstrate. Specificity of staining was confirmed by omitting the primaryantibody as a negative control. T cells were labeled by rat anti-CD3(1:200; Serotec, Düsseldorf, Germany), macrophages by rat anti-mouse

Mac-3 (1:200; BD Pharmingen, Heidelberg, Germany). Bielschowsky sil-ver impregnation was employed to assess axonal pathology and LuxolFast Blue staining was used to assess demyelination as described previ-ously (Linker et al., 2011). Axonal damage was additionally visualizedby double staining with CNPase and the neurofilament antibody SMI31(1:1000; Invitrogen, Karlsruhe, Germany) with anti-mouse Alexa 488-and Cy3-conjugated secondary antibodies. The sections double labeledfor neurofilaments (SMI31) and CNPase using immunofluorescencewere analyzed on a fluorescence microscope (Olympus BX51).

To calculate the inflammatory index, spinal cord cross sections werestained with haemalaun. Histological and immunohistochemical quan-tifications were performed by a blinded observer as described earlier(Linker et al., 2011).

Splenocyte and lymphocyte culture and ELISA experiments

Splenocytes were cultured in 24-well plates at a concentration of5 × 106 cells/ml of complete GluMax medium (GIBCO®, Invitrogen,Germany), containing 10% heat inactivated fetal calf serum (FCS), 1%sodium pyruvate, 1% L-glutamine, 1% MEM-NEAA and 1% penicillin/streptomycin. Cellswere stimulatedwith 10 μg/mlMOGor 1.25 μg Con-canavalin A (ConA) for 24, 48 and 72 h at 37 °C, 5% CO2. At the indicatedtime points, supernatants were collected and centrifuged to eliminatecellular debris. Cell culture supernatants were stored at −20 °C andbrought to room temperature immediately before the measurement ofdifferent cytokines, including IL-6 (BD OptEIA 555240, Pharmingen,Germany), IL-10 (R&D Systems DY417, Heidelberg, Germany), IL-17(R&D Systems DY421, Heidelberg, Germany) and IFN-γ (R&D SystemsDY485, Heidelberg, Germany) according to manufacturers' protocols.Results shown are the mean of triplicates ± SEM.

Flow cytometry (FACS) experiments

FACS analyses were performed using a FACS Canto II and CellQuestsoftware (BD). Splenic single cell suspensions and lymphocyte fromlymph nodes were incubated for 20 min at 4 °C and washed withFACS buffer, centrifuged and resuspended in FACS buffer. Monoclonalantibodies purchased fromBD (Heidelberg, Germany) andMiltenyi Bio-tech (Gladbach, Germany) were used to detect CD11b (clone M1/70)CD11c (clone HL3), CD80 (clone 16-10A1), CD86 (clone GL1) andMHC-II (clone 2G9) and CD25CD4 in lymph nodes.

Statistical analysis

All histological analyses were performed completely blindedwith respect to treatment. Data are presented as mean ± SEM. Sta-tistical analysis was performed by one-way analysis of variance(ANOVA) or Kruskal–Wallis test (all analyses done by Graph PadPrism 6, San Diego, CA, USA). As post-hoc tests Bonferroni's MultipleComparison Tests were performed. A probability level (p-value) of*p b 0.05, **p b 0.01 and ***p b 0.001 was considered to be statisti-cally significant for all tests. All error bars represent SEM.

Results

Combination of DMF and IFN-β improves the clinical disease course in EAEmice

To compare monotherapy of DMF and IFN-β with a combination ofboth compounds, we analyzed clinical symptoms and disease severityof MOG-EAE in C57BL/6 mice until the early chronic phase of the dis-ease. The incidence of EAE was 100% in the DMF-group (25/25) and inthe IFN-β-group (24/24) versus 22/24 in the combination-group(92%) and 19/23 in the sham treated control group (83%). The differencewas statistically not significant. There were no unexpected adverseevents upon daily inspection of mice with no differences in mortality

52 C. Reick et al. / Experimental Neurology 257 (2014) 50–56

or body weight between all treatment groups (data not shown). Micereceiving combination therapy displayed a significantly less severeEAE disease course compared to the vehicle only control group(Fig. 1). Combination therapy also yielded a significantly greater benefi-cial effect on clinical course vs. IFN-β alone. Whilst the combinationgroup only exhibited limp tail and unsteadiness of gait, IFN-β treatedmice suffered from paraparesis and mice in the vehicle only controlgroup displayed severe gait ataxia. These analyses were confirmed intwo further independent experiments with similar group size.

Combination treatment improved histological outcome in EAE

To correlate clinical data with histology, we next quantified inflam-matory cell infiltration in the CNS as well as demyelination and axonaldamage. Evaluation of axonal densities revealed a significant preserva-tion of axons in mice treated with combination therapy, as comparedto monotherapy and vehicle only (Figs. 2E–H and Table 1).

Comparing DMF/IFN-β and vehicle only, inflammatory cell infiltra-tion was significantly reduced after combination therapy. Combinationtherapy reduced macrophage infiltration as compared to control andIFN-β monotherapy (Figs. 2I–Q, Table 1). Luxol Fast Blue stainingshowed significantly less demyelination in mice treated with DMF/IFN-β combination therapy as compared to control animals and atrend towards preservation of myelin versus mice treated with DMFor IFN-β alone (Figs. 2R–U and Table 1).

Furthermore, using confocal laser scanning microscopy after stain-ing for neurofilament and CNPase, we could show that treatment withDMF/IFN-β led to preservation of intact axons containing phosphorylat-ed, SMI31-positive axons in EAE lesions (Figs. 3A–D).

Combination of DMF/IFN-β treatment partly stimulated cytokineproduction ex vivo

We then investigated the concomitant immunological processesunder the different therapeutic regimes. To this end, we assessed IL-6,IL-10, IL-17 and IFN-γ secretion upon ex vivo stimulation of splenocytesat day 10 (Figs. 4A–D). Subsequent to ConA stimulation DMF/IFN-βcombination therapy increased IFN-γ levels significantly as comparedto vehicle (Fig. 4A). IL-6, IL-10 and IL-17 were also affected after treat-ment with both, DMF and IFN-β, in contrast to vehicle only in theacute phase underlining the same trend (Figs. 4B–D) shown in IFN-γlevels.

FACS analysis of immune cell phenotypes

To determine whether IFN-β, DMF or the combination of both hadany effect on overall T cell activation and differentiation, APCs from

Fig. 1. Combination of IFN-β and DMF ameliorated EAE disease course. Acute clinicalcourse of EAE in C57BL/6 female mice after monotherapy with IFN-β (n= 24) or DMF(n = 25) or combination therapy DMF/IFN-β (n = 24) versus control (n = 23).Combination treatment yielded a less severe course of disease compared to IFN-βmonotherapy and control mice. Data represent themean of three independent exper-iments ± SEM.

each group were analyzed by flow cytometry from spleen and lymphnode in the acute phase (CD11b+, CD11b/c+, CD11c+). MHC II expres-sion on CD11b+ cells or CD11b/c+ cells was not different between thecontrol group, monotherapies and DMF/IFN-β treated EAE mice (datanot shown).

T cell proliferation studies

To investigate the effect of DMF, IFN-β or the combination of both onMOG35–55-specific T cell responses, isolated splenocytes of immunizedanimals were cultured in the presence of MOG35–55 or ConA. After incu-bation with [3H] thymidine for 16 h proliferation wasmeasured by [3H]thymidine incorporation. In all three treatment groups, no significantinhibition of T cell proliferation was observed (data not shown).

Discussion

This study demonstrates that a combination of DMF and IFN-β canameliorate EAE disease clinical course more effectively compared tosham treatment and IFN-β monotherapy. In contrast, there was no su-periority of combination therapy compared to DMF monotherapy. Inline to clinical data, histological analyses revealed less inflammation, de-myelination and in particular preservation of axonal damage.

We hypothesized that combination of IFN-β and DMFwith putative-ly complementary mechanisms acts synergistic in EAE: this combina-tion may tackle both anti-inflammatory and central neuroprotectiveeffects. Basically cross-connectivity between neurodegeneration and in-flammation under autoimmune inflammatory conditions has alreadybeen described (Trapp and Nave, 2008): With its anti-inflammatory,anti-oxidative and pro-metabolic potential in a variety of cell types,DMF may inhibit immune response and hereby exert a beneficial effecton inflammatory activity (Scannevin et al., 2012; Schilling et al., 2006)IFN-β reduces inflammation by positively influencing the balance ofpro- and anti-inflammatory mechanisms in the periphery, and alsoentry of monocytic cells into the CNS via MMP activity (Floris et al.,2002; Schmidt et al., 2001). Additionally, IFN-β may increase the pro-duction of nerve growth factor with the potential capacity to improveneuronal survival and repair (Kieseier, 2011). Thus a combination ofcompounds has the potential tomaximize beneficial effects and thus in-duce synergies in EAE.

Focusing on the clinical course of EAE the effectiveness of combina-tion therapywith DMF/IFN-βmay be explained by histopathological ef-fects acting on both inflammation, especially on macrophages and ondegeneration, especially axonal damage.

We observed higher axonal densities and milder demyelination inall treated groups but the most marked beneficial effect was seen inthe DMF/IFN-β treated group. Activated macrophages and microgliacan be found nearby degenerating axons, myelin sheaths and oligoden-drocytes (Ferguson et al., 1997; Kornek et al., 2000; Trapp et al., 1998).Prior studies in EAE reported that axonal damage in mice was reducedafter DMF therapy (Linker et al., 2011). Axonal damage has the stron-gest impact on the development of irreversible neurological symptomsinMS (Ferguson et al., 1997; Kornek et al., 2000; Trapp andNave, 2008).Common pathways between neurodegeneration and inflammationunder autoimmune inflammatory conditions have already been dem-onstrated (Trapp and Nave, 2008) and it is known that inflammationis a driving effect for degeneration. In the current study an appropriatetrend for this effect was confirmed. Here, axon protection or more cor-rectly axonal sparing may be the consequence of the simultaneousanti-inflammatory/immunomodulatory properties of the combinationof DMF/IFN-β. We demonstrated that DMF and IFN-β treatment alonepreserved axonal loss compared to sham treatment but a combinationof both substances yieldedmore pronounced effects, highlighting puta-tive synergies mainly on axonal targets.

Additional effects of combination treatment were seen on myeloidcells, especially on macrophage infiltration. IFN-β is still one of the

Fig. 2. Histological analyses. Representative spinal cord cross-sections of mice on day 18 post immunization (p.i.). Note infiltrates (arrows) after Haemalaun staining (A to D) in controlmice (A), IFN-β treatedmice (B) and DMF-treatedmice (C). Combination therapy diminishes number and occurrence of infiltrates (D). (E to H) Bielschowsky silver staining from control(E), IFN-β-(F), DMF- (G) and DMF/IFN-β (H) treated C57BL/6 mice. In contrast to the control group (E) there is less axonal loss after treatment (F to H). CD3 staining (I–M) revealed in-flammatory T cell infiltrates. Further histological analyses showed a stronger infiltration (arrows) after Mac-3 staining (N–Q) and increased demyelination (framed area) after Luxol FastBlue staining (R to U) in control- (N and R) and IFN-β- (O and S) or DMF-group (P and T). Scale bar = 100 μm for A to Q and 500 μm for R to U.

Table 1Histological analyses: quantification and p-values.

A Control DMF IFN-β DMF/IFN-β

Inflammatory index 5.3 ± 1.3 3.7 ± 0.5 2.4 ± 0.6 1.2 ± 0.5Demyelination of white matter in % ± SEM 8.1 ± 1.5 4.2 ± 1.1 3.8 ± 0.8 1.6 ± 0.6Axonal density ± SEM 2.3 ± 0.2 3.3 ± 0.2 3.7 ± 0.3 4.6 ± 0.2CD3 positive cells/mm2 ± SEM 305 ± 24 264 ± 33 286 ± 16 228 ± 24Mac3 positive cells/mm2 ± SEM 596 ± 33 532 ± 30 533 ± 34 385 ± 36

B Inflammatory Index Demyelination Axonal density CD3 positive cells/mm2 Mac3 positive cells/mm2

Control IFN-β n.s. n.s p** b 0.01 n.s n.sDMF n.s. n.s p* b 0.05 n.s n.sDMF/IFN-β p*** b 0.001 p** b 0.01 p*** b 0.001 n.s. p** b 0.01

IFN-β DMF n.s. n.s n.s. n.s n.sDMF/IFN-β n.s. n.s n.s. n.s p* b 0.05

DMF DMF/IFN-β n.s. n.s p** b 0.01 n.s n.s

Blinded quantification (A) of the inflammatory index after HE-staining, CD3 positive cells, Mac-3 positive macrophages/microglia, demyelination after Luxol Fast Blue staining and axonaldensities after silver impregnation were performed on day 18 post immunization marking the acute phase of the disease (n = 12 per group). Data are presented as cells/mm2 for CD3,Mac-3, percent demyelination for Luxol Fast Blue and relative axonal densities for silver impregnated axons. P-values upon comparison of the different groups are calculated byKruskal–Wallis test with Bonferroni post testing as depicted in B. n.s. = not significant.

53C. Reick et al. / Experimental Neurology 257 (2014) 50–56

Fig. 3. Combination of DMF and IFN-β protects axon density. DMF/IFN-β protects axons in EAE lesions. Representative fluorescence microscopy images of lesions from spinal cord cross-sections are shown (scale bar= 50 μm, CNPase positivemyelin is depicted in red, SMI31 positive, naked axons in green). Dashedwhite lines denote lesion borders in A–D. As compared tocontrols (A), therapywith DMF (C) and IFN-β (D) alone leads to some and DMF/IFN-β therapy (B) leads to significant preservation of SMI31-positive, naked axons in lesions (green). (Forinterpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

54 C. Reick et al. / Experimental Neurology 257 (2014) 50–56

mainstay therapies used for MS-treatment. In line with our results, sev-eral other publications have shown beneficial but limited effects of IFN-β in EAE models (Floris et al., 2002; Ruuls et al., 1996; Schmidt et al.,2001; Yu et al., 1996). While the mechanisms of IFN-β are pleiotropic,immunomodulatory properties of IFN-β are well established (Nelsonet al., 1996; Tuohy et al., 2000), with a special focus on myeloid cells,most notably monocytes (Prinz et al., 2008).

Similar to IFN-β, DMF also acts on myeloid cells (Linker et al., 2010;Schilling et al., 2006). In linewith results of Schilling et al., a reduction inmacrophage infiltration in the spinal cord was observed after DMFtreatment without relevant inhibition of T cell infiltration. In additionto their antigen presenting functionmacrophages/microglia cells are ac-tive players in myelin destruction in EAE (Bauer et al., 1994; Ghoreschiet al., 2011).

The role of macrophages/microglia as active players in myelin de-struction in EAE aswell as inMS (Bauer et al., 1996) seems to be tackledby combination of DMF/IFN-β during EAE. We found a reduction ofmacrophages/microglia in situ under combination treatment, whereasIFN-β therapy exhibit less reduced numbers of macrophages/microglia.Hereby, the less severe clinical course in EAE might additionally be ex-plained besides the more impressive effect on axons. Macrophagesand microglia are known to be important effector cells for tissue injuryin inflammatory diseases of the CNS.

We believe that the synergistic effect of combination therapy withDMF/IFN-β is based on its twomodes of action resulting in axon protec-tion: 1. a direct effect via the Nrf2 pathway, that after activation due toDMF/IFN-β increases levels of detoxifying enzymes and in consequencereduces neurotoxicity, and 2. an indirect effect viamodulation ofmacro-phages making a contribution to axonal damage.

The causal chain is induced by axonal damage (Linker et al., 2005;Nikic et al., 2011) that in turn is essential since axonal injury and neuro-nal loss are correlated with the progression of disease (Ferguson et al.,1997; Trapp et al., 1998).

Thus our data argue for a clinical trial with IFN-β plus DMF in RRMSpatients. Both substances are characterized to be safe and well-

tolerated. They have been used for many years so that potential side ef-fects arewell-known andmanageable. Additive immunosuppressive ef-fects are not expected. Although the exact mechanisms of preservationof axons after combination therapy are not dissected in total, we pro-pose a beneficial net effect of combination therapy also for MS patients.Therefore, clinical studies in RRMS seem to be very reasonable andpromising.

Conclusion

In conclusion thepresenteddata underline thedifferent putative im-munomodulatory mechanisms of IFN-β after systemic use with strongemphasis on anti-inflammatory effects resulting in axonal sparing.Combination treatment of DMF/IFN-β in an autoimmune experimentalmodel improves the clinical course of EAE. Thismay result fromputativeneuroprotective activities of DMF (Barnabe-Heider and Miller, 2003;Yamada et al., 2001) which add to the immunomodulatory effects ofboth drugs. Establishment of both, effective and safe treatment regimesin MS therapy is a major goal and different combinations of immuno-modulatory agents have been tested so far (Brod et al., 2000; Calabresiet al., 2002; Gold, 2008; Jolivalt et al., 2003; Soos et al., 2002; Vollmeret al., 2004; Weilbach et al., 2004). To better understand the underlyingmechanisms, further studies are warranted.

Authors' contributions

CR carried out the experiments, analyzed histological and immuno-logical results and drafted themanuscript. RAL, CS andRGprovided gen-eral support and participated in the design of the study. JT supportedFACS analysis. RAL and GE conceived the study, and participated in itsdesign and coordination and surveyed and supported the draft of themanuscript. RHS is a full time employee of Biogen Idec, Inc. All authorsread and approved the final manuscript.

Fig. 4.Cytokine levels. Levels of interferon-γ (IFN-γ) are higher in ConA stimulated splenocytes of DMF/IFN-β treatedmice in the acute phase (A). DMF/IFN-β treatment also increases IL-6(B) and anti-inflammatory IL-10 (C) after ConA stimulation as compared to vehicle. Different treatment regimes do not influence expression of the mentioned cytokine after MOG35–55

stimulation. Data shown are representative of three independent experiments showing the same result.

55C. Reick et al. / Experimental Neurology 257 (2014) 50–56

56 C. Reick et al. / Experimental Neurology 257 (2014) 50–56

Competing interests

RL and RG received travel support, speaker's honoraria and researchgrants from BiogenIdec. RG also has board activities for BiogenIdec. RHSis an employee of BiogenIdec; the company involved in marketing ofBG-12 (dimethyl fumarate). The other authors declare no competinginterests.

Acknowledgments

The skillful technical support of Bernadette Jesionek and SilviaSeubert is highly appreciated. The original enthusiasm of Prof. Altmeyerand Prof. Przuntek, Bochum, in introducing FAE to therapy of MS isgratefully acknowledged.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.expneurol.2014.04.003.

References

Barnabe-Heider, F., Miller, F.D., 2003. Endogenously produced neurotrophins regulatesurvival and differentiation of cortical progenitors via distinct signaling pathways.J. Neurosci. 23 (12), 5149–5160.

Bauer, J., Sminia, T., Wouterlood, F.G., Dijkstra, C.D., 1994. Phagocytic activity of macro-phages and microglial cells during the course of acute and chronic relapsing experi-mental autoimmune encephalomyelitis. J. Neurosci. Res. 38 (4), 365–375.

Bauer, J., Ruuls, S.R., Huitinga, I., Dijkstra, C.D., 1996. The role of macrophage subpopula-tions in autoimmune disease of the central nervous system. Histochem. J. 28 (2),83–97.

Brod, S.A., Lindsey, J.W., Wolinsky, J.S., 2000. Combination therapywith glatiramer acetate(copolymer-1) and a type I interferon (IFN-alpha) does not improve experimentalautoimmune encephalomyelitis. Ann. Neurol. 47 (1), 127–131.

Burks, J., 2005. Interferon-beta1b for multiple sclerosis. Expert. Rev. Neurother. 5 (2),153–164.

Calabresi, P.A., et al., 2002. An open-label trial of combination therapy with interferonbeta-1a and oral methotrexate in MS. Neurology 58 (2), 314–317.

Ellrichmann, G., et al., 2011. Efficacy of fumaric acid esters in the R6/2 and YAC128modelsof Huntington's disease. PLoS ONE 6 (1), e16172.

Ferguson, B., Matyszak, M.K., Esiri, M.M., Perry, V.H., 1997. Axonal damage in acute mul-tiple sclerosis lesions. Brain 120 (Pt 3), 393–399.

Floris, S., et al., 2002. Interferon-beta directly influences monocyte infiltration into thecentral nervous system. J. Neuroimmunol. 127 (1–2), 69–79.

Ghoreschi, K., et al., 2011. Fumarates improve psoriasis and multiple sclerosis by inducingtype II dendritic cells. J. Exp. Med. 208 (11), 2291–2303.

Gold, R., 2008. Combination therapies in multiple sclerosis. J. Neurol. 255 (Suppl. 1),51–60.

Gold, R., Linington, C., Lassmann, H., 2006. Understanding pathogenesis and therapy ofmultiple sclerosis via animal models: 70 years of merits and culprits in experimentalautoimmune encephalomyelitis research. Brain 129 (Pt 8), 1953–1971.

Herrero-Herranz, E., Pardo, L.A., Gold, R., Linker, R.A., 2008. Pattern of axonal injury inmu-rine myelin oligodendrocyte glycoprotein induced experimental autoimmune en-cephalomyelitis: implications for multiple sclerosis. Neurobiol. Dis. 30 (2), 162–173.

Jacobs, L., 1996. Magnetic resonance imaging in clinical therapeutic trials of multiple scle-rosis. West. J. Med. 164 (6), 531–532.

Johnson, J.A., et al., 2008. The Nrf2-ARE pathway: an indicator and modulator of oxidativestress in neurodegeneration. Ann. N. Y. Acad. Sci. 1147, 61–69.

Jolivalt, C.G., Howard, R.B., Chen, L.S., Mizisin, A.P., Lai, C.S., 2003. A novel nitric oxide scav-enger in combinationwith cyclosporine A ameliorates experimental autoimmune en-cephalomyelitis progression in mice. J. Neuroimmunol. 138 (1–2), 56–64.

Kieseier, B.C., 2011. Themechanism of action of interferon-beta in relapsingmultiple scle-rosis. CNS Drugs 25 (6), 491–502.

Kornek, B., et al., 2000. Multiple sclerosis and chronic autoimmune encephalomyelitis: acomparative quantitative study of axonal injury in active, inactive, and remyelinatedlesions. Am. J. Pathol. 157 (1), 267–276.

Linker, R.A., et al., 2002. CNTF is a major protective factor in demyelinating CNS disease: aneurotrophic cytokine as modulator in neuroinflammation. Nat. Med. 8 (6), 620–624.

Linker, R.A., et al., 2005. EAE in beta-2 microglobulin-deficient mice: axonal damage is notdependent on MHC-I restricted immune responses. Neurobiol. Dis. 19 (1–2),218–228.

Linker, R.A., et al., 2010. Functional role of brain-derived neurotrophic factor in neuropro-tective autoimmunity: therapeutic implications in a model of multiple sclerosis. Brain133 (Pt 8), 2248–2263.

Linker, R.A., et al., 2011. Fumaric acid esters exert neuroprotective effects in neuroinflam-mation via activation of the Nrf2 antioxidant pathway. Brain 134 (Pt 3), 678–692.

Mendes, A., Sa, M.J., 2011. Classical immunomodulatory therapy in multiple sclerosis:how it acts, how it works. Arq. Neuropsiquiatr. 69 (3), 536–543.

Nelson, P.A., et al., 1996. Effect of oral beta interferon on subsequent immune responsive-ness. Ann. N. Y. Acad. Sci. 778, 145–155.

Nikic, I., et al., 2011. A reversible form of axon damage in experimental autoimmune en-cephalomyelitis and multiple sclerosis. Nat. Med. 17 (4), 495–499.

Prinz, M., et al., 2008. Distinct and nonredundant in vivo functions of IFNAR on myeloidcells limit autoimmunity in the central nervous system. Immunity 28 (5), 675–686.

Ruuls, S.R., et al., 1996. The length of treatment determines whether IFN-beta prevents oraggravates experimental autoimmune encephalomyelitis in Lewis rats. J. Immunol.157 (12), 5721–5731.

Scannevin, R.H., et al., 2012. Fumarates promote cytoprotection of central nervous systemcells against oxidative stress via the nuclear factor (erythroid-derived 2)-like 2 path-way. J. Pharmacol. Exp. Ther. 341 (1), 274–284.

Schilling, S., Goelz, S., Linker, R., Luehder, F., Gold, R., 2006. Fumaric acid esters are effec-tive in chronic experimental autoimmune encephalomyelitis and suppress macro-phage infiltration. Clin. Exp. Immunol. 145 (1), 101–107.

Schmidt, J., Sturzebecher, S., Toyka, K.V., Gold, R., 2001. Interferon-beta treatment of ex-perimental autoimmune encephalomyelitis leads to rapid nonapoptotic terminationof T cell infiltration. J. Neurosci. Res. 65 (1), 59–67.

Soos, J.M., et al., 2002. Cutting edge: oral type I IFN-tau promotes a Th2 bias and enhancessuppression of autoimmune encephalomyelitis by oral glatiramer acetate. J.Immunol. 169 (5), 2231–2235.

Trapp, B.D., Nave, K.A., 2008. Multiple sclerosis: an immune or neurodegenerative disor-der? Annu. Rev. Neurosci. 31, 247–269.

Trapp, B.D., et al., 1998. Axonal transection in the lesions of multiple sclerosis. N. Engl. J.Med. 338 (5), 278–285.

Tuohy, V.K., et al., 2000. Modulation of the IL-10/IL-12 cytokine circuit by interferon-betainhibits the development of epitope spreading and disease progression inmurine au-toimmune encephalomyelitis. J. Neuroimmunol. 111 (1–2), 55–63.

Vollmer, T.L., et al., 2004. An open-label safety and drug interaction study of natalizumab(Antegren) in combination with interferon-beta (Avonex) in patients with multiplesclerosis. Mult. Scler. 10 (5), 511–520.

Weilbach, F.X., Chan, A., Toyka, K.V., Gold, R., 2004. The cardioprotector dexrazoxane aug-ments therapeutic efficacy of mitoxantrone in experimental autoimmune encephalo-myelitis. Clin. Exp. Immunol. 135 (1), 49–55.

Weiner, H.L., 2009. The challenge of multiple sclerosis: how do we cure a chronic hetero-geneous disease? Ann. Neurol. 65 (3), 239–248.

Williamson, T.P., Johnson, D.A., Johnson, J.A., 2012. Activation of the Nrf2-ARE pathway bysiRNA knockdown of Keap1 reduces oxidative stress and provides partial protectionfrom MPTP-mediated neurotoxicity. Neurotoxicology 33 (3), 272–279.

Yamada,M., et al., 2001. Differences in survival-promoting effects and intracellular signal-ing properties of BDNF and IGF-1 in cultured cerebral cortical neurons. J. Neurochem.78 (5), 940–951.

Yu, M., Nishiyama, A., Trapp, B.D., Tuohy, V.K., 1996. Interferon-beta inhibits progression ofrelapsing-remitting experimental autoimmune encephalomyelitis. J. Neuroimmunol.64 (1), 91–100.