Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 https://doi.org/10.1186/s12974-022-02386-1 RESEARCH Open Access Administration of Maresin‑1 ameliorates the physiopathology of experimental autoimmune encephalomyelitis Alba Sánchez‑Fernández1,2, Stephanie Zandee3, Mauricio Mastrogiovanni4, Marc Charabati3, Homero Rubbo4, Alexandre Prat3 and Rubèn López‑Vales1,2* Abstract Background: Resolution of inflammation is an active and regulated process that leads to the clearance of cell debris and immune cells from the challenged tissue, facilitating the recovery of homeostasis. This physiological response is coordinated by endogenous bioactive lipids known as specialized pro-resolving mediators (SPMs). When resolution fails, inflammation becomes uncontrolled leading chronic inflammation and tissue damage, as occurs in multiple sclerosis (MS). Methods: SPMs and the key biosynthetic enzymes involved in SPM production were analysed by metabololipidom‑ ics and qPCR in active brain lesions, serum and peripheral blood mononuclear cells (PBMC) of MS patients as well as in the spinal cord of mice with experimental autoimmune encephalomyelitis (EAE). We also tested the therapeutic actions of the SPM coined Maresin-1 (MaR1) in EAE mice and studied its impact on inflammation by doing luminex and flow cytometry analysis. Results: We show that levels of MaR1 and other SPMs were below the limit of detection or not increased in the spinal cord of EAE mice, whereas the production of pro-inflammatory eicosanoids was induced during disease progression. Similarly, we reveal that SPMs were undetected in serum and active brain lesion samples of MS patients, which was linked to impaired expression of the enzymes involved in the biosynthetic pathways of SPMs. We demonstrate that exogenous administration of MaR1 in EAE mice suppressed the protein levels of various pro-inflammatory cytokines and reduced immune cells counts in the spinal cord and blood. MaR1 also decreased the numbers of Th1 cells but increased the accumulation of regulatory T cells and drove macrophage polarization towards an anti-inflammatory phenotype. Importantly, we provide clear evidence that administration of MaR1 in mice with clinical signs of EAE enhanced neurological outcomes and protected from demyelination. Conclusions: This study reveals that there is an imbalance in the production of SPMs in MS patients and in EAE mice, and that increasing the bioavailability of SPMs, such as MaR1, minimizes inflammation and mediates therapeutic actions. Thus, these data suggest that immunoresolvent therapies, such as MaR1, could be a novel avenue for the treatment of MS. *Correspondence: Ruben.Lopez@uab.cat 1 Institut de Neurociencies and Departament de Biologia Cel lular, Fisiologia i Immunologia, Facultat de Medicina, Universitat Autonoma de Barcelona, 08193 Bellaterra, Catalonia, Spain Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 2 of 18 Keywords: Experimental autoimmune encephalomyelitis, Inflammation, Maresin-1, Multiple Sclerosis, Resolution of inflammation, Specialized pro-resolving mediators Background cerebrospinal fluid (CSF) of highly active MS patients Acute inflammation is a key physiological mechanism [6]. A more recent work uncovered that progressive and that promotes the repair of injured tissues and eliminates relapsing–remitting MS patients had increased plasma infectious organisms and toxic agents. This response is levels of pro-inflammatory eicosanoids, whereas SPMs tightly controlled and typically ends with the elimination were reduced or undetectable [20]. Moreover, this study of the immune cells and cellular debris from the tissue, also found lower transcripts of the SPM biosynthetic paving the way for the recovery of homeostasis [1–3]. enzymes and receptors in leukocytes from MS patients, Contrariwise, uncontrolled inflammation, as it occurs suggesting an imbalance in peripheral induction of the when the resolution phase fails, becomes harmful to the resolution pathways in MS patients [20]. Interestingly, tissue, and it is a hallmark of a wide variety of pathologi- prophylactic administration of two SPM known as RvD1 cal conditions, including in MS [2, 4–8]. and LXA4 has resulted in beneficial actions experimental For many years, resolution of inflammation was autoimmune encephalomyelitis (EAE) [19, 21]. However, believed to be a passive process triggered by the dilution whether impairment in SPM production occurs in the of pro-inflammatory mediators at the challenged tissue, CNS of MS patients and EAE mice, and whether admin- thus halting the recruitment of leukocytes from circula- istration of SPMs that are undetected in blood and CNS tion. However, it has now become clear that this process samples of MS patients halters inflammation and exerts is an active and coordinated event orchestrated by a class therapeutic effects in animal models of MS has not been of lipids known as SPMs [2, 7]. This bioactive lipid family elucidated. includes lipoxins (LX), resolvin D series (RvD), resolvin E Herein, we reveal that there is impaired production series, (RvE), protectins (PD/NPD) and maresins (MaR) of SPMs in active brain lesions and serum samples of [2, 7, 9]. In immune cells, SPMs are naturally synthesized MS patients, as well as in the CNS of mice undergoing from omega-3 and omega-6 polyunsaturated fatty acids EAE. We further connect these results with a defective (PUFA) through a family of lipoxygenases (i.e., 12- and expression of enzymes involved in the synthesis of SPMs 15-LOX) [3, 10]. Failure to produce adequate amounts of in peripheral blood mononuclear cells (PBMCs) and in SPMs has been associated with persistent inflammation brain lesions of MS individuals, as well as, in the spinal in many inflammatory disorders, such as asthma, ath- cord of EAE mice. We also reveal that the exogenous erosclerosis, ulcerative colitis, spinal cord injury or Alz- administration of the DHA-derived SPM coined MaR1 heimer’s disease, among others [2, 4, 5, 11, 12]. Despite in EAE mice once they showed the clinical signs of the emerging data revealing that SPMs might control inflam- disease modulates different mechanisms of inflammation mation in central nervous system (CNS) disorders, there and confers protection against neurological deficits and are very few studies addressing their importance in neu- demyelination. roinflammatory conditions, such as MS. MS is a chronic neuroinflammatory and demyelinat- Methods ing disease of the CNS affecting around 2.5 million indi- MS brain tissue collection viduals worldwide [13–15]. Although the exact cause of Autopsies on patients from Centre de Recherche du Cen- MS is not yet known, a hallmark of its physiopathology tre Hospitalier de l’Université de Montreal diagnosed is the presence of persistent numbers of immune cells in with clinical and neuropathological MS according to the the white and grey matter of the CNS [13–16]. Indeed, revised 2010 McDonald’s criteria [22] had informed con- neuropathological studies show that inflammation is pre- sent as approved by the Centre Hospitalier de l’Université sent in all MS stages and that infiltrating leukocytes drive de Montréal ethics committee (BH07.001). Autopsy the formation of demyelinated and neurodegenerative samples were cryopreserved, and lesions classified using lesions [15, 17, 18]. Luxol Fast Blue (LFB)/Haematoxylin & Eosin stain- To our knowledge, there are currently few studies ing and Oil Red O staining as previously published [23, addressing the contribution of SPMs in MS patients on 24]. For qPCR analysis, normal-appearing white matter animal models [6, 19–21]. In an early report, a connec- (NAWM) and active lesion (AL) representing 7 second- tion between disease severity and SPMs production was ary-progressive MS (SPMS) and 2 relapsing–remitting suggested in a small cohort of MS patients since the MS (RRMS) were used. The median age at death was authors found increased levels of RvD1 and NPD1 in the 50 years old (range from 26 to 65 years old) (Additional Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 3 of 18 file 1: Table S1). Regions of interest (lesion or NAWM) for in 100 µL sterile saline were administered at the time of RNA isolation were dissected manually from 5 to 6 50 µm induction and again 48 h later. All the mice were housed cryosections per block. For lipidomic analysis, normal- with food and water ad libitum at a room temperature of appearing white matter (NAWM) and active lesions (AL) 22 ± 2ºC under 12:12 h light–dark cycle. EAE onset was representing 6 secondary-progressive MS (SPMS) and 2 considered on the day animals showed the first signs of relapsing–remitting MS (RRMS) were used. The median disease (EAE score 0.5 or 1; around day 9–10 post-immu- age at death was 50 years old (range from 26 to 65 years nization). Disease peak was considered on the day EAE old) (Additional file 1: Table S2). score did not increase from the previous day (day 16–18 post-immunization). Chronic phase of the disease was MS and healthy donor PBMC and serum collection considered at day 21 post-immunization if EAE score For isolation of the peripheral blood mononuclear decreased or was maintained with respect to disease cells (PBMCs), blood samples were collected from 8 peak. patients with relapsing–remitting MS (RRMS) and 3 with secondary-progressive MS (SPMS), recruited at the MS EAE functional evaluation Clinic at the Centre de Recherche du Centre Hospitalier Mice were daily scored from day 0 to day 21 after induc- de l’Université de Montréal. The mean age was 43 years tion of EAE. The researcher was blind to experimental old (ranged from 27 to 65) (Additional file 1: Table S1). groups during the functional evaluation. A 6-point scale 8 healthy volunteers were included as controls, whose was used to evaluate the clinical signs of EAE: 0 = normal mean age was 35 years old (range from 24 to 42 years old) walking, 0.5 = partially paralysed tail, 1 = fully paralysed (Additional file 1: Table S1). PBMCs were isolated from tail, 2 = mild hind limb weakness, quick righting reflex, blood samples collected in EDTA-coated Vacutainer 3 = severe hind limb weakness, slow righting reflex, tubes (BD Biosciences, Oakville, ON, Canada), using a unable to bear weight, 3.5 = severe hind limb weakness, Ficoll density gradient as previously described [25]. partial paralysis of hind limb, 4 = complete paralysis of Moreover, serum samples were also obtained from 10 at least one hind limb, 4.5 = complete paralysis of one patients with RRMS and 10 healthy volunteers. The mean or both hind limbs and trunk weakness, 5 = complete age was 47 years old (ranging from 36 to 55) and 43 years paralysis of one or both hind limbs, forelimb weakness or old (ranging from 35 to 55) in the MS and healthy donor paralysis, and 6 = moribund. group, respectively (Additional file 1: Table S2). A written informed consent was obtained from Drug administration patients and healthy donors in accordance with the local EAE-induced mice were randomly assigned to the treat- ethics committee (CRCHUM research ethic committee ment and control experimental groups. Daily injec- approval number BH07.001). tions (i.p.) of 1 µg of MaR1 (7R,14S-dihydroxy-4Z, 8E, 10E,12Z, 16Z, 19Z-DHA; Cayman Chemical)) in 200 µL Experimental autoimmune encephalomyelitis of sterile saline (0.9% NaCl) or vehicle were initiated in All experimental procedures were approved by the Uni- each individual mouse the first day of the clinical signs versitat Autònoma de Barcelona Animal Experimenta- until the end of the study 21 days after induction. The tion Ethical Committee (CEEAH 2878) and followed the dose of MaR1 was chosen based on the effectivity of this European Communities Council Directive 2010/63/EU, SPM in spinal cord injury [4]. and the methods were carried out in accordance with the approved guidelines. Real‑time quantitative PCR assay (qPCR) Female adult C57Bl/6 (8–10 weeks old; Charles River Unimmunized and immunized female C57Bl6/J mice at Laboratories) were sedated with intramuscular injec- different stages of the disease (onset: 11/12 days after tion of a mixture of ketamine (22 mg/kg) (Imalgen 1000, EAE induction, peak: 17 days after EAE induction, and Merial) and xylazine (2.5 mg/kg) (Rompun, Bayer). chronic phase: 21 days after EAE induction) were tran- Experimental autoimmune encephalomyelitis (EAE) was scardially perfused with 60 mL of sterile saline (0.9% actively induced by subcutaneously immunization with NaCl) and lumbar spinal cords were harvested. Total 300 µg of myelin oligodendrocyte glycoprotein peptide RNA was isolated using RNeasy Lipid Tissue Mini Kit 35–55 ­(MOG35-55 MEVGWYRSPFSRVVHLYRNGK, (Qiagen) according to the manufacturer’s instruction. Thermo Fisher Scientific, MA, USA) in 200 µL Complete RNA was isolated from post-mortem MS brain samples Freund’s Adjuvant (CFA) (Difco, MI, USA) supplemented and PBMCs from MS patients and healthy donors using with 4 mg/mL of heat inactivated Mycobacterium tuber- RNeasy Lipid Tissue Micro Kit (Qiagen) and QIAamp culosis (Difco, MI, USA). Intraperitoneal (i.p.) injections RNA Blood Tissue Mini Kit (Qiagen), respectively, of 400 ng of pertussis toxin (Sigma-Aldrich, ON, USA) according to the manufacturer’s guidelines. Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 4 of 18 1 µg of RNA isolated from mouse spinal cords and paraformaldehyde (PFA) in 0.1 M phosphate buffer human PBMCs, and 0.1 µg of RNA from human brain (PB). Lumbar segments of spinal cords were harvested, samples, was then retrotranscribed using cDNA Reverse post-fixed in 4% PFA for 2 h and cryoprotected in Transcription Kit (Applied Biosystems). qPCR was per- 30% sucrose in 0.1 M at 4 °C for at least 48 h. Spinal formed using TaqMan reagents from Applied Biosystems cords were embedded in TissueTek OCT (Sakura), cut and the following TaqMan-designed primers (Thermo into transversal Section (15 µm thick) with a cryostat Fisher Scientific): mouse Alox-12/15 (Mm00507789_m1), (Leica) between L3 and L5 segments and serially picked human ALOX12 (Hs00167524_m1) and human ALOX-15 up on gelatine-coated glass slides. Samples were stored (Hs00993765_g1). Mouse Gapdh (Mm99999915_g1) and at −20 °C. human GAPDH (Hs02786624_g1) were used as house- To measure the myelinated area, sections were keeping gene. qPCR was performed using the Bio-Rad stained with LFB (Sigma-Aldrich). Briefly, after a CFX384 apparatus and software. The amount of cDNA graded dehydration, sections were placed in 1 mg/mL was calculated based on the threshold cycle (CT) value of LFB solution in 96% EtOH and 0.05% acetic acid and was standardized by the amount of Gapdh using the overnight at 37 °C and protected from light. Then, ­2−ΔΔCT method [26]. slides were washed with 96% EtOH, rehydrated in dis- tilled water and placed in a 0.5 mg/mL ­Li2CO3 solution Lipidomic analysis for 3–5 min at room temperature. Finally, sections were Lipidomic analysis was done on brain AL and NAWM washed in distilled water, dehydrated again in 100% tissue obtained from MS patients, serum samples col- EtOH and mounted in DPX (Sigma-Aldrich). To assess lected from MS and healthy volunteers and spinal cord the demyelinated area in the spinal cord, 6 random samples harvested from both naïve and EAE mice at images per mouse were captured at 10X magnifica- different stages of the disease (onset, peak and chronic tion with an Olympus BX51 and the attached Olympus phase). DP73 Camera. Demyelination was assessed by measur- All samples were thawed and weighted and 180 µL ing the demyelinated areas within the spinal cord white MeOH containing internal deuterated standard mix was matter. These analyses were done using the Image J added. Homogenization was performed with zirconium analysis software. oxide beads in a bullet blender (Next Advance) according To analyse axonal damaged, sections were incubated to the manufacturer’s instructions and further centrifu- overnight at 4ºC with an antibody against neurofilament gation at 14.000 g was done. Supernatants were diluted (1:1000; Millipore). Next, sections were washed twice with 0.1% formic acid (1:9) and oxylipin analysis was per- in phosphate-buffered saline (PBS) and incubated for formed. Human serum samples (500 µL) were thawed 1 h at room temperature with the secondary antibody and diluted in equal volume of 20% MeOH containing donkey anti-rabbit Alexa Fluor 594 (1:500; Invitrogen) internal deuterated standard mix. After centrifugation at and washed again in PBS and PB. Finally, sections were 14.000 g, supernatants were submitted to oxylipin analy- stained with Dapi (1:1000; Sigma-Aldrich) to label nuclei sis as described below. and mounted in DPX (Sigma-Aldrich) after a graded Oxylipin analysis was performed as previously dehydration. To analyse the samples, random images were captured at 40X magnification with an Olympus conditioned Strata™-X SPE cartridges (Phenomenex, described [27]. Briefly, samples are transferred to pre- BX51 and the attached Olympus DP73 Camera. Torrance, CA) and washed with 10% MeOH solution in acidified water (HCl, pH 3). Then oxylipins were eluted with MeOH under vacuum, dried and reconstituted with Cytokine expression 50% MeOH. Extracts were separated on a Luna C18(2) EAE mice were euthanized at peak of the disease (17 days column (Phenomenex, Torrance, CA) and analysed by after EAE induction) with an overdose of pentobarbi- negative mode electrospray ionization using scheduled tal sodium (Dolethal) and transcardially perfused with multiple reaction monitoring (sMRM) on a spectrometer 60 mL of sterile saline (0.9% NaCl). Spinal cords were QTRAP4500 (ABSciex, Framingham, MA). Identification harvested and rapidly frozen in liquid nitrogen. Pro- and quantitation assessed by internal standard methods tein isolation from the spinal cord samples and cytokine were performed with Analyst 1.6.2 software (ABSciex). quantification were performed as described previously [28]. The protein levels of the following cytokines (IL-4, Histological analysis IL-10, IL-1α, IL-1β, IL-3, IL-6, TNF-α, IFN-γ, IL-17A, EAE mice were euthanized on day 21 post-immu- CSF-3, CCL-5, CCL-2, CXCL-2 and CXCL-10) were ana- nization with an overdose of pentobarbital sodium lysed using a custom-designed Cytokine Magnetic Bead (Dolethal) and transcardially perfused with 4% Panel (Invitrogen) on a MAGPIX system (Millipore). Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 5 of 18 using FlowJo® software V.10. Microglia cells were defined Flow cytometry Cytometer (BD Bioscience) and all data were processed Immune cell infiltration was determined from blood, lymph nodes and spinal cord of EAE mice at the peak as ­CD45low and ­ CD11b+, whereas macrophages were high of EAE (17 days after EAE induction). Briefly, mice were CD45 , ­CD11b+ and F4/80+ and granu- defined as ­ euthanized with an overdose of pentobarbital sodium locytes (mainly neutrophils) as ­CD45high, ­CD11b+ and (Dolethal). 15 µL of blood was collected from a cardiac F4/80− according to previous publications [4, 27] puncture and stored in heparinized vials at 4ºC. Then, mice were transcardially perfused with 60 mL of sterile Statistical analyses saline (0.9% NaCl) and the spinal cords and lymph nodes Data are shown as mean ± standard error of the mean (cervical and inguinal) were collected. (SEM). The Kolmogorov–Smirnov test was used to check Blood samples were incubated with red blood cell lysis whether the data was normal distributed. Two-tailed Stu- buffer (BioLegend) according to manufactures’ guide to dent’s t test was used for the comparison between two obtain a cell suspension enriched in leukocytes. Spinal different groups (qPCR, histological analysis, normalized cords and lymph nodes were cut into small pieces and cytokine levels and flow cytometry analysis). EAE clini- enzymatically dissociated in 1 mL of Hank’s Balanced cal score was analysed by using two-way repeated meas- Salt Solution (HBSS; Sigma-Aldrich) without ­Ca2+/Mg2+ ures ANOVA with post hoc Bonferroni’s test for multiple containing 0.1% collagenase and 0.1% DNase for 30 min comparisons. Cytokine levels expressed in picograms/ at 37 °C, and then, mechanically disintegrated through a grams of protein in four different groups were analysed 70-µm cell strainer to obtain a cell suspension [28]. by using a one-way ANOVA with post hoc Tukey’s test Cell suspensions were split into different 1.5 mL for multiple comparison. Differences were considered microcentrifuge tubes according to the number of anti- significant at p < 0.05. body combinations. For extracellular staining, samples were split, and unstained cells and isotype-matched Results control samples were generated to control for nonspe- There is an imbalance between pro‑inflammatory cific binding of antibodies and for autofluorescence. For and pro‑resolving lipid mediators in EAE mice surface staining the following antibodies from eBiosci- We first conducted metabolomic analysis in spinal cord ence were used at a 1:300 concentration: CD45-PerCP, samples of EAE mice to study whether the production of CD11b-PE or PE-Cy7, F4/80-PE or -APC, Ly6C-FITC, SPMs was induced in the CNS of one of the most widely Ly6G-PE, CD3-FITC-APC-PerCP; CD4-APC-Cy7, CD8- used animal models of MS. We found that levels of DHA APC, CD49b-PE, CD24-PE, CD16/32 PE, CD206 FITC, dropped in the spinal cord of EAE, especially at chronic CD11a-PE, CD49d-FITC, ICAM-1-FITC and CD62L- stages of the disease, although this reduction did not PE-Cy7. Samples were incubated with the primary anti- reach statistical significance (Fig. 1A). bodies for 1 h at 4 °C with gentle agitation, washed with Further analysis of DHA metabolites revealed some Dulbeccos’ Modified Eagle Medium (DMEM)-10% foetal DHA mediators produced by 12/15LOX, such as bovine serum (FBS) and centrifuged twice at 300 g for 14HDoHA and 17HDoHA, which are pathway markers 10 min at 4 °C to remove debris and then fixed with 1% for the formation of Maresins, RvD series and Protectin PFA. D1 (PD1), tended to increase at these EAE stages, but For intracellular staining in those samples where not at the significant level (Fig. 1A). Although PD1, RvD1 required, the following antibodies were also used at and RvD5 were detected in the spinal cord at physiologi- 1:300 concentration: FoxP3- PE-Cy7, tBet-PerCP, RORγ- cal conditions, their levels did not increase in any of the APC, GATA3-PE, IFNy-Alexa488, IL-17A-Alexa488, EAE stages analysed (Fig. 1A). Other SPMs derived from IL-4-APC, IL-10-Alexa488 or APC (all of them from eBi- DHA, such as MaR1, MaR2 and some members from oscience) iNOS-Alexa 647 (Abcam) and Arg1-Alexa 488 the RvD series, such as RvD2, Rv3 and Rv4, were unde- (Santa Cruz). Cells were fixed and permeabilized with tected in the mouse spinal cord at physiological condi- FoxP3 Transcription Factor Staining Buffer Set (eBiosci- tions and during EAE disease. The only DHA metabolite ence). Samples were immunostained with the intracellu- that was significantly augmented in EAE mice was the lar antibodies for overnight at 4ºC. In the case cytokines 16(17)-EpDPE, which is produced via cytochrome P450 antibodies, samples were incubated 4 h at 37 °C in agi- (CYP450) and has unknown functions (Fig. 1A). tation with Cell Stimulation Cocktail 1X (eBioscience) Levels of EPA remained altered in the spinal cord of and Brefeldin A 10 µL/ml (Biolegend) before the fixation EAE mice, although they tended to be higher at chronic and permeabilization for intracellular staining. Finally, stages of the disease (Fig. 1B). However, lipid mediators stained cells were washed with PBS twice and fixed with generated from EPA by 12/15LOX, such as 12-HEPE 1% PFA. Samples were analysed on a FACS Canto Flow and 15-HEPE, were increased in the spinal cord of EAE, Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 6 of 18 Fig. 1 Lipidomic analysis of SPM synthesis pathways markers in the CNS of naïve and EAE mice. A, B Graphs showing the quantification of A DHA metabolome and B EPA metabolome in the spinal of naïve and EAE mice (n = 5 per group). Dotted line highlights the naïve value. *p < 0.05; **p < 0.01; vs. vehicle. Two-way ANOVA with repeated measures, Bonferroni’s post hoc test. Data are shown as mean ± SEM Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 7 of 18 especially at the chronic phase of the disease, but only by 12/15LOX, such as 12-HETE and the intermediate 12-HEPE levels reached statistical significance (Fig. 1B). metabolite in LXA4 and LXB4 synthesis, 15-HETE, were The levels of 18-HEPE, the RvE-series lipid intermediate also increased in the spinal cord of EAE mice. However, produced via CYP450, were significantly increased in the significant differences were only observed in 15-HETE spinal cord of chronic EAE (Fig. 1B). However, none of levels at chronic stages of the disease (Fig. 2), while LXA4 the members of the RvE series were detected. and LXB4 were undetected. As for the AA metabolome, we found that the levels In contrast to SPMs, most of the pro-inflammatory of the precursor (AA) were reduced in the spinal cord lipid mediators derived from AA, such as LTB4, PGD2, of EAE mice, reaching statistical significance at chronic PGE2, PGF2 and TXB2, were significantly increased in stages of the disease (Fig. 2A). However, we found a sig- the spinal cord of EAE mice at the peak and/or chronic nificant increase in the levels of 11,12-EET, which is gen- stage (Fig. 2). erated by CYP450, at the peak and chronic phase of EAE disease (Fig. 3C). Lipid mediators generated from AA Fig. 2 Lipidomic analysis of arachidonic acid pro-inflammatory metabolites in the CNS of EAE and naïve mice. Graphs showing the levels of lipid mediators of the AA metabolome in the spinal of C57Bl/6 mice in physiological conditions and at different stages of EAE (n = 5 per group). Red arrows and text indicate the pro-inflammatory metabolite conversion and its implicated enzyme. Dotted line highlights the naïve value. *p < 0.05; **p < 0.01; ***p < 0.001 vs. vehicle. Two-way ANOVA with repeated measures, Bonferroni’s post hoc test. Data are shown as mean ± SEM Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 8 of 18 Fig. 3 Lipidomic analysis of SPM synthesis pathways markers in serum of healthy controls and MS patients. A–C Graphs showing the quantification of A DHA metabolome, B EPA metabolome C AA metabolome from serum samples of healthy volunteers (n = 10) and MS patients (n = 10). *p < 0.05 vs. Control. Unpaired t-test. Data are shown as mean ± SEM Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 9 of 18 Fig. 4 Lipidomic analysis of SPM synthesis pathways markers in brain samples of MS patients. A–C Graphs showing the quantification of A DHA metabolome, B EPA metabolome and C AA metabolome in NAWM and AL (n = 8). Paired t-test. Data are shown as mean ± SEM These data suggest there is an imbalance between pro- levels of 11,12-EET, 5-HETE and 12-HETE were also inflammatory and pro-resolving lipid mediators in the increased in the serum of MS patients yet not signifi- spinal cord of EAE mice, in favour of the former, cantly (Fig. 3C). Despite we found several SPM interme- diate precursors derived from DHA, EPA and AA in the Synthesis of SPMs is impaired in MS patients serum samples, SPMs were undetected in serum samples We then performed metabolomic analysis on serum and of MS patients and healthy donors. brain samples of MS patients. These studies revealed that We then performed metabolomic studies in active the serum levels of DHA, EPA and AA did not change lesions (AL) and normal-appearing white matter between MS patients and healthy controls (Fig. 3A–C). (NAWM) of MS post-mortem brain tissue to investigate Analysing the lipid mediators derived from the DHA whether the synthesis of SPMs was defective at the spe- metabolome in serum samples, we found that 14-HDoHE cific regions of the CNS undergoing inflammation. We and 17-HDoHE, the maresin and resolvin D-series path- found that DHA, EPA and AA levels did not vary sig- way markers, were increased in MS patients yet not sig- nificantly in AL relative to NAWM, although half of MS nificant (Fig. 3A). As for the EPA metabolome, the levels individuals showed mark increase in EPA levels in AL the RvE1 metabolic intermediate, 18-HEPE, as well as (Fig. 4A–C). We also detected some intermediate precur- 12-HEPE and 15-HEPE were higher in MS patients, sors of SPM synthesis in brain MS samples, although no although only 15-HEPE reached statistically significance significant differences were found in AL as compared to (Fig. 3B). Furthermore, as for the AA metabolome, the NAWM (Fig. 4A–C). From the DHA metabolome, the Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 10 of 18 only lipid mediator detected was the 4-HDoHE, a LOX-5 serum samples, we did not detect any SPM in the MS metabolite that in animal models has been shown to brain AL and NAWM. Collectively, these data suggest mediate the antiangiogenic effect of omega-3 fatty acids that similar to EAE mice, the class switch from pro- [29]. However, the levels of 4-HDoHE in AL were higher inflammatory to pro-resolution lipid mediators does not in only 2 out 8 MS patients, but no significant differences occur properly in the periphery and in the CNS of MS were found between AL and NAWM (Fig. 4A). Lipid patients. mediators generated from the EPA metabolome were undetected in AL and NAWM. As for the AA metabo- Feeble expression of ALOX12 and ALOX15 in MS patients lome, we found that 5-HETE, 12-HETE and 15-HETE and EAE mice levels in AL were higher in 4 out of 8 patients, but no Since our data suggest there is impaired activation of the significant changes were observed (Fig. 4C). Similar to resolution programmes of inflammation in MS patients Fig. 5 Expression of lipoxygenase 12 and 15 in MS patients and lipoxygenase 12/15 in EAE mice. A–D Graphs showing the mRNA levels of ALOX12 A, C and ALOX15 B, D in PBMCs of MS individuals (n = 11) and healthy donors (n = 8) B, C and in MS brain AL and NAWM C, D (n = 9). E Time course expression of Alox12/15 mRNA levels in the spinal cord of EAE mice over disease progression (n = 3–4 per time point). Unpaired t-test in A–D and one-way ANOVA, Bonferroni’s post hoc test in (E). Data are shown as mean ± SEM Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 11 of 18 and EAE mice, we then assessed whether this was asso- ciated with a failure in the expression of the key biosyn- thetic enzymes involved in SPM production. Thus, we measured the expression of lipoxygenase-12 (ALOX12) and lipoxygenase-15 (ALOX15) in PBMCs of active MS patients and healthy donors, as well as in active MS lesions in post-mortem brain tissue. qPCR analy- sis revealed that mRNA levels of ALOX12 and ALOX15 were undetected or found at very low levels PBMCs of either healthy donors or MS patients (Fig. 5A, B). Then, when assessing the expression of these enzymes in post- mortem MS brain samples, we did not find significant changes in the expression within the lesioned areas com- pared to the NAWM, despite some variability across patients (Fig. 5C, D). Indeed, only 2 out of 9 patients showed increased expression of these enzymes in the AL Fig. 6 MaR1 reduces pro-inflammatory cytokines in the spinal cord of EAE mice. Plot showing the fold change in cytokine protein levels compared to NAWM, while the rest did not show any of different cytokines in the spinal cord of EAE mice treated with change or a slight downregulation (1 patient in ALOX12 MaR1 relative to those treated with vehicle (dotted line). Grey bars and 2 in ALOX15) (Fig. 5C, D). highlight those cytokines that are significantly regulated by MaR1. We then studied whether this phenomenon also *p < 0.05; **p < 0.01; ***p < 0.001 vs. vehicle. Unpaired t-test (n = 4 per occurred in EAE, one of the most widely used animal group). Data are shown as mean ± SEM models of MS. For this purpose, we analysed the expres- sion of Alox12/15, the murine orthologue of the previous mentioned enzymes in humans and the initiating enzyme assay revealed that MaR1 reduced significantly the lev- in the SPM biosynthetic pathway. qPCR analysis revealed els of 5 out 9 pro-inflammatory cytokines (IL-1α, IL-1β, that, similar to MS patients, mRNA levels of Alox12/15 IL-6, TNFα and INFγ) and 2 out the 4 chemokines meas- were barely found in the CNS of EAE mice, although the ured (CCL-2, CCL-5) (Fig. 6; Additional file 1: Table S3). transcripts of this enzyme tended to be greater at the Since cytokines and chemokines coordinate the chronic stage of the disease (Fig. 5E). migration of immune cells into tissues, we next studied Overall, these results suggest that the key enzymes whether MaR1 modulates the number of immune cells involved in the production of SPMs are not induced in in the spinal cord of EAE mice at the peak of the dis- MS individuals nor EAE mice. ease. Flow cytometry experiments revealed that MaR1 strongly reduced the numbers of T cells, including both Effects of MaR1 on inflammation in EAE CD4 + T cells (CD3 + , CD4 +) and CD8 + cells (CD3 + , Since we demonstrated that production of SPMs is not CD8 +), and B cells (CD45 + , CD11b–, CD3–, CD24 +). induced in MS patients and in EAE mice, we proceeded MaR1 also reduced the accumulation of macrophages to investigate whether exogenous administration of SPMs (CD45high, CD11b + , F4/80 +) and activated micro- attenuates inflammation in EAE. Among the different glial cells (CD45low, CD11b +) but not neutrophils SPMs, we tested the effects of MaR1 since (i) this SPM (CD45high, CD11b + , F4/80–, Ly6G +) at this time point was undetected in the serum and brain active lesions (Fig. 7A; Additional file 1: Fig. S1). of MS patients; (ii) unlikely other SPMs, such as RvD1, Due to the role of cytokines and SPMs in modulating T RvD5 and NP1, MaR1 was undetected in the spinal cord cell and macrophage phenotypes [4, 30], we investigated of EAE mice; (iii) previous studies revealed that this SPM whether MaR1 affects these immune cells responses. We is also undetected in plasma [20] and CFS [6] samples of found that MaR1 reduced the number of CD4 T cells MS patients; (iv) MaR1, in contrast to other SPMs that expressing IFNγ in the spinal cord of EAE mice, suggest- were also undetected in MS and EAE patients, such as ing that the number of pathogenic Th1 cells was reduced RvD2–RvD4, has already demonstrated efficacy in mini- (Fig. 7B; Additional file 1: Fig. S2B). However, MaR1 did mizing inflammation in the CNS [20]; and (v) the contri- not alter the percentage of T cells expressing tBet, a tran- bution of this SPM in demyelinating diseases is unknown. scription factor associated with Th1 responses (Fig. 7C). For this purpose, we administered 1 µg to EAE mice, MaR1 did not attenuate the infiltration of Th2 nor once a day from disease onset (0.5–1 EAE score) and we Th17 cells since the expression signature cytokines and assessed the protein levels of 14 cytokines in the spinal transcription factors, GATA3 and RORγ, respectively, cord at the disease peak (4–4.5 EAE score). Luminex remained stable after treatment (Fig. 7B, C). However, Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 12 of 18 MaR1 increased the proportion of Tr1 cells (CD3 + , Next, we investigated whether the lessened leuko- CD4 + , CD49b +) (Fig. 7C; Additional file 1: Fig. S2B), cyte infiltration in the CNS after MaR1 treatment is which have a crucial role in maintaining tolerance to linked with the attenuation of immune cells expansion self-antigens, and increased in ~ 50% the percentage of in the lymph nodes or in the blood stream. We found this lymphocyte subset producing IL-10, although not to that MaR1 slightly reduced the number of the different the significant level (Fig. 7D; Additional file 1: Fig. S2C). leukocyte populations in the lymph nodes, although the MaR1 treatment, however, did not alter the percentage of result was not statistically significant (Fig. 8A). How- Treg cells (CD3 + , CD4 + , FoxP3 +) (Fig. 7C; Additional ever, MaR1 reduced by ~ twofold the number of dendritic file 1: Fig. S2C), another subset of regulatory lympho- cells (CD45high, CD11b + , F4/80 + , CD11c), which cytes, but it tended to increase their production of IL-10 have a decisive role in T cell priming. However, in con- (Fig. 6D; Additional file 1: Fig. S2C). trast to that observed in the spinal cord, MaR1 did not As for macrophages, we found that MaR1 reduced modulate the phenotype of CD4 T cells and monocytes. the accumulation of this myeloid cells in the spinal (Fig. 8C–D). cord of EAE mice, but it increased the percentage of Despite the minimal actions of MaR1 on lymphocyte CD206 + macrophages (CD45high, CD11b + , F4/80 + , and monocyte in the lymph nodes, we observed fewer CD206 +) (Fig. 7A, E; Additional file 1: Fig. S3A), indi- CD4 T cells, B cells and monocytes in the blood of EAE cating that this SPM increased the proportion of anti- mice (Fig. 9A). MaR1 also led to changes in the pheno- inflammatory macrophages. In this line, we also found type of circulating CD4 T cells, increasing the percent- that MaR1 bolstered the ratio ­Ly6Clow/Ly6Chigh in mac- age of Tr1 cells and also reducing that of pathogenic Th1 rophages by ~ threefold, further supporting that this cells (Fig. 9B). Nonetheless, MaR1 did not modify the SPM favoured the anti-inflammatory responses of mac- phenotype of circulating monocytes, which had a pre- rophages. (Fig. 5F; Additional file 1: Fig. S4B). Similar to dominant pro-inflammatory state in both treated and no macrophages, microglia counts were markedly reduced treated mice (CD45high, CD11b + , F4/80 + , ­Ly6Chigh) by MaR1 (Fig. 7A). However, this SPM led to increased (Fig. 9C). These results suggested that some immune cell proportion of anti-inflammatory CD206 + microglia responses observed in the spinal cord of MaR1-treated (CD45low, CD11b + , CD206 +) (Fig. 7G; Additional mice might be, in part, due to the actions of this SPM in file 1: Fig. S3C). the periphery. We then assessed whether the reduction in leukocyte infiltration in the spinal cord by MaR1 was associated Systemic administration of MaR1 attenuates EAE with a reduction in the expression of adhesion molecules To investigate whether the effects of MaR1 on inflam- in different population of immune cells. We found that matory processes carry a clinical outcome, we assessed MaR1 did not alter the expression of any of the stud- the effects of exogenous delivery of MaR1 on EAE symp- ied adhesion molecules (CD11a, ICAM-1, VLA-4 and toms. We found that mice treated with MaR1 were mark- CD62L) in macrophages, microglia, granulocytes or edly protected against neurological impairments, despite lymphocytes (Additional file 1: Fig. S4A–C), suggesting the treatment being initiated at disease onset (data that the reduced infiltration of leukocytes into the CNS pooled from two independent experiments) (Fig. 10A, of EAE mice treated with MaR1 was likely related to the B). Indeed, MaR1-treated mice reached an average EAE reduced expression of pro-inflammatory cytokines in the score of ~ 2.5 by the end of the follow-up, while vehicle- spinal cord rather than to an effect of MaR1 on leuko- treated mice reached ~ 4.2 (Fig. 10A). Moreover, mice cyte adhesion on blood vessels. Nonetheless, we did find treated with MaR1 had significantly lower cumulative an increase in the percentage of macrophages express- and maximum EAE scores than vehicle-treated mice ing VLA-4 (CD45high, CD11b + , F4/80 + , VLA-4 +) in (Fig. 10B). MaR1-treated mice (Additional file 1: Fig. S4A), possibly We then evaluated whether the enhancement in neu- associated with the enhanced infiltration of anti-inflam- rological functions observed in EAE mice treated with matory macrophages that we described before. MaR1 was related to preservation of myelin within the CNS. (Fig. 10C, D). However, histological sections (See figure on next page.) Fig. 7 MaR1 attenuates inflammation in the spinal cord of mice undergoing EAE. A–G Graph showing the quantification of different immune cell populations in the spinal cord (A), the number of CD4 + T cells expressing IL-4, IL-10, IFNγ or IL-17A (B), the percentage of CD4 + T cells expressing the CD49b, FoxP3, tBet, GATA3 or RORγ (C), the percentage of regulatory lymphocytes expressing IL-10 (D), the percentage of macrophages expressing polarization markers (E), the ratio of ­Ly6Clow/Ly6Chigh macrophages F and the percentage of microglial cells expressing polarization markers in the spinal cord after MaR1 or vehicle treatment. *p < 0.05; **p < 0.01; ***p < 0.001 vs. vehicle. Unpaired t-test (n = 4 per group). Data are shown as mean ± SEM Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 13 of 18 Fig. 7 (See legend on previous page.) Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 14 of 18 Fig. 8 Effects of MaR1 on immune cells in the lymph nodes at the peak of EAE. A Graph showing the quantification of different leukocytes populations and dendritic cells B in the lymph nodes of MaR1- or vehicle-treated mice. C Plot showing the percentage of CD4 + T cells expressing the transcription factors CD49b, FoxP3, tBet, GATA3 or RORγ and the cell numbers for the different macrophage subsets according to Ly6C expression D in the lymph nodes of MaR1- or vehicle-treated mice. Unpaired t-test (n = 4 per group). Data are shown as mean ± SEM Fig. 9 Effects of MaR1 on immune cells in the blood of EAE mice at disease peak. A Graph showing the number of different leukocytes populations in the blood of MaR1- or vehicle-treated mice. B Plot showing the percentages of CD4 + T cells expressing the transcription factors CD49b, FoxP3, tBet, GATA3 or RORγ and D different macrophage phenotypes according to Ly6C expression in blood of EAE mice treated with MaR1 or vehicle. Unpaired t-test (n = 4 per group). *p < 0.05; **p < 0.01 vs. vehicle. Unpaired t-test (n = 4 per group). Data are shown as mean ± SEM immunostained against NF revealed that there were observed in EAE mice treated with MaR1 was due to the very few axons within the demyelinating lesions of mice ability of this SPM to reduce the size of the CNS lesions treated with MaR1 or vehicle (Additional file 1: Fig. S5). rather than to remyelination. These data suggest that the reduced neurological deficits Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 15 of 18 2 (COX-1 and 2) and 5-LOX from AA, such as prosta- glandins, thromboxanes and leukotrienes, contribute to inflammation by promoting chemotaxis and activation of immune cells and increasing vascular permeability [33, 34]. Nevertheless, not all the actions of eicosanoid are pro-inflammatory since a number of in vivo studies revealed that 15d-Prostaglandin J2, for instance, attenu- ates inflammation and neurological decline in EAE and spinal cord injury [35, 36]. In contrast, SPMs are key players in the resolution of inflammation. Except for lipoxins, SPMs are derived from omega-3 PUFA, in particular, from DHA and EPA. DHA can give rise to several resolution agonists, includ- ing those of the resolvin D series, protectin D and MaRs. EPA can also yield some SPMs, namely, resolvin E series [2, 7, 9, 11, 37, 38]. Fig. 10 Effects of MaR1 on neurological deficits and myelin loss Omega-3 PUFAs are enriched in oils derived from fish in EAE mice. A, B Graphs showing the clinical score of EAE mice and algae. Various epidemiological studies have indicated treated with MaR1 or vehicle over disease progression (A), as well as the maximum and cumulative EAE score (B). C Graph showing the that omega-3 PUFA dietary supplementations are associ- quantification of myelin loss in the lumbar spinal cord of MaR1- or ated with improved clinical outcome of various patholo- vehicle-treated mice at 21 days post-induction. D Representative gies, such as metabolic syndrome or cardiovascular histological spinal cord tissue sections stained with LFB from EAE diseases [39, 40]. In MS, however, the beneficial actions mice treated with vehicle and MaR1. *p < 0.05; **p < 0.01; ***p < 0.001 of omega-3 PUFA supplementation are inconsistent, as vs. vehicle. Two-way ANOVA with repeated measures, Bonferroni’s post hoc test in A (n = 12 per group). Unpaired t-test in B (n = 12 per one trial failed to show clinical improvement despite the group) and C (n = 9 in vehicle and n = 11 in MaR1). Data were pooled effectively increased omega-3 PUFA levels in patients as from two different experiments. Data are shown as mean ± SEM a result of the supplementation [41, 42]. Thus, it is very plausible that this lack of efficacy stems from the failure of MS patients to produce adequate amounts of SPMs, Overall, these data reveal the therapeutic potential of rather than from the bioavailability of omega-3 PUFA in MaR1 in attenuating neuroinflammation and protecting their serum. from tissue damage and clinical signs of EAE. Previous studies have reported that lipid mediator pathways are dysregulated in various nervous system- Discussion related pathologies [4, 6, 19, 20, 43]. For instance, inap- In the present study, we report that the production of propriate SPM production has been observed in other the main lipid mediators that activate the resolution pro- CNS conditions, such as in Alzheimer’s [43] and in mice grammes of inflammation is deficient in serum and active after spinal cord contusion injury [4]. A recent report has brain lesions of MS patients, as well as, in the CNS of uncovered that most SPMs are absent in plasma sam- EAE mice. These findings are associated with impaired ples of MS patients at different stages of the disease, and induction of key enzymes involved in the biosynthetic only RvD1, RvD5 and NP1 could be detected [20]. These pathways of SPMs. We also reveal that the administra- observations are in agreement with our findings shown tion of the resolution agonist, coined MaR1, in EAE here, although we were unable to detect any SPM in mice attenuates the pro-inflammatory environment of serum or inflammatory regions of the brain. These small the CNS, modulates T cell and macrophage responses differences between both studies could be due to type of towards a more regulatory or anti-inflammatory pheno- sample analysed, and in case of the active brain lesions type, and reduces the blood counts of different leukocyte samples, to the limited amount of tissue available. How- populations. Importantly, we demonstrate that the effects ever, RvD1, RvD5 and NPD1 were the three SPMs that we of MaR1 on inflammation are associated with marked detected in the spinal cord of mice at physiological condi- protection against neurological decline and myelin loss tions, but their levels did not increase at different stages despite treatment initiation at disease onset. of EAE disease. Interestingly, we found that the impaired PUFAs play a critical role in the regulation of inflam- production of SPMs in serum plasma of MS patients was mation by controlling several processes involved in its associated with deficient induction in the expression of onset and termination [2, 31, 32]. In general, eicosa- the main SPM biosynthetic enzymes in PBMCs, which is noids generated via the actions of cyclooxygenase 1 and in agreement with the work of Kooij and colleagues [20]. Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 16 of 18 Besides, we also provided evidence indicating that this pathogenic Th1 and Th17 cells and their generation inefficient induction in 12- and 15-LOX expression also from naïve CD4 cells [30]. We did not observe any sig- occurs in regions of the brain undergoing inflammation nificant change in the cell number of different Th cell in MS patients, as well as in the spinal cord of EAE mice. populations in the lymph nodes, suggesting that MaR1 Therefore, these results suggest that there is a failure in does not regulate the peripheral activation of Th cells in peripheral and central production of SPMs in MS, which EAE, at least when its administration occurs after dis- may be responsible, in part, of the dysregulated deleteri- ease onset. MaR1, however, reduced Th1 cell number in ous inflammatory response characteristic of this disease. the blood and increased that of non-classical Tregs. In Interestingly, Kooij and colleagues observed inverse this study, we also observed that macrophages present correlation between the plasma levels of these RvD1, in the spinal cord of EAE adopted a more anti-inflam- RvD5 and NP1 and clinical severity in MS patients [20], matory phenotype, similar to our previous findings in and that LXA4, LXB4, RvD1 and PD1 reduced MS- spinal cord injury [4]. Because this was only observed derived monocyte activation and cytokine production, in the CNS and not also in the lymph nodes or blood, it suggesting that administration of SPM could attenuate suggests that the modulation of the inflammatory envi- inflammation in MS patients. In this line, prophylac- ronment of the CNS by MaR1 is most likely to be the tic administration of RvD1 in EAE alleviated the clinical responsible for the effects on macrophages. However, signs of disease, suggesting that the exogenous admin- the impact of MaR1 on immune cell numbers is unlikely istration of SPMs could revert the pathological signs of dependent on the modulation of adhesion molecules in neuroinflammation [19]. peripheral leukocytes or microglial cells since this SPM MaR1 is an understudied member of the SPM super- fail to reduce their expression. Indeed, MaR1 increased family [9, 44]. There are just few studies demonstrating the expression of VLA-4 in macrophages. This cell the beneficial actions of this SPM in the CNS. MaR1 adhesion has been recently shown to be important for not only enhanced inflammatory resolution after spi- ­ y6Clow macrophages [52] and may the infiltration of L nal cord injury and reduced myelin loss and neuro- account for the higher infiltration of anti-inflammatory logical deficits [4], but also promoted neuroprotection macrophages observed in the CNS of EAE mice treated in spinal muscular atrophy [45] and cerebral ischemia with MaR1. [46]. Thus, we complete this scenario since herein we provide data demonstrating that the administration of MaR1 in EAE mice also conferred protection against Conclusions functional deficits and myelin loss, despite initiating Overall, we provide clear evidence that there is inef- the treatment after disease onset. We also found that ficient production of SPMs in the spinal cord of MaR1 modulates different inflammatory mechanisms. EAE mice as well as in the CNS and periphery of MS First, MaR1 reduced the protein levels of different pro- patients. Our data indicate that exogenous administra- inflammatory cytokines in the spinal cord of EAE mice, tion of MaR1 is effective at enhancing multiple mecha- including IL-1α, IL-1β, TNFα, IFNγ and IL-6, among nisms needed for preventing inflammation in EAE mice others, but did not alter the amounts of anti-inflam- and confers protection against neurological decline and matory/immunomodulatory cytokines, such as IL-4 or myelin loss. These data therefore indicate that SPMs in IL-10. These pro-inflammatory cytokines play a critical general, and MaR1 in particular, represent a novel ther- role in the development of EAE [47–49], and are asso- apeutic avenue for the treatment of MS with promising ciated with neurodegeneration and progression in MS potential. [50, 51]. Second, MaR1 reduced the accumulation of major leukocyte populations in the spinal cord of EAE Abbreviations mice, albeit not neutrophils. This may have occurred, at AA: Arachidonic acid; CFA: Complete Freund´s adjuvant; CNS: Central nerv‑ least partially, due to the decreased levels of cytokines ous system; DHA: Docosahexaenoic acid; EAE: Experimental autoimmune in the spinal cord, and also because of MaR1’s ability to encephalomyelitis; EPA: Eicosapentaenoic acid; LOX: Lipoxygenase; LX: Lipox‑ ins; MaR1: Maresin-1; MS: Multiple sclerosis; NAWM: Normal-appearing white hamper the accumulation of immune cells also into the matter; PBMCs: Peripheral blood mononuclear cells; PBS: Phosphate-buffered blood. Third, MaR1 modulated the responses of mac- saline; PD/NPD: Protectins/Neuroprotectins; PPMS: Primary-progressive multi‑ rophages and CD4 T cells towards a more anti-inflam- ple sclerosis; PUFAs: Polyunsaturated fatty acids; qPCR: Real-time Quantitative PCR Assay; RRMS: Relapsing–remitting multiple sclerosis; RvD: Resolvin D; matory profile. In particular, MaR1 reduced pathogenic RvE: Resolvin E series; sMRM: Scheduled multiple reaction monitoring; SPMS: Th1 cells in the spinal cord and increased the number Secondary-progressive multiple sclerosis; SPMs: Specialized pro-resolving lipid of non-classical regulatory T cells. In agreement with mediators. our findings, a previous study revealed that resolution agonists, including MaR1, attenuate the activation of Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 17 of 18 Supplementary Information Author details 1 Institut de Neurociencies and Departament de Biologia Cel lular, Fisiologia The online version contains supplementary material available at https://​doi.​ i Immunologia, Facultat de Medicina, Universitat Autonoma de Barcelona, org/​10.​1186/​s12974-​022-​02386-1. 08193 Bellaterra, Catalonia, Spain. 2 Centro de Investigación Biomédica en Red de Enfermedades Neurodegenerativas (CIBERNED), Barcelona, Spain. 3 Additional file 1. Fig. S1. MaR1 reduces the accumulation of immune Department of Neuroscience, Faculty of Medicine, Université de Montréal cells in the spinal cord of mice at the peak of EAE. Fig. S2. MaR1 modu‑ and Neuroimmunology Unit, Centre de Recherche du CHUM (CRCHUM), lates T cells responses in the spinal cord of EAE mice. Fig. S3. MaR1 Montréal, Québec, Canada. 4 Departamento de Bioquímica, Facultad de modulates the macrophages and microglia phenotypes towards a more Medicina and Centro de Investigaciones Biomédicas (CEINBIO), Universidad de anti-inflammatory phenotype. Fig. S4. Effects of the treatment with MaR1 La República, Montevideo, Uruguay. in the expression of adhesion molecules on immune cells in the spinal cord at the peak of EAE. Fig. S5. Effects of MaR1 in axonal damage in the Received: 17 September 2021 Accepted: 13 January 2022 lesioned areas of the spinal cord at the peak of EAE. Table S1. Descriptive information about human samples of MS patients and healthy controls used for QPCR analysis. Table S2. Descriptive information about human samples of MS patients and healthy controls used for lipidomic analysis. Table S3. Cytokine expression in the spinal cord of naïve, vehicle- and References MaR1-treated mice at the peak of EAE. 1. Russell P. Pathologic basis of disease. Pathology. 2007;7:260. 2. Serhan CN. Pro-resolving lipid mediators are leads for resolution physiol‑ ogy. Nature. 2014;510:92–101. Acknowledgements 3. Serhan CN, Chiang N, Dalli J, Levy BD. Lipid mediators in the resolution of The authors thank Mónica Espejo, Jessica Jaramillo, Neus Hernández and inflammation. Cold Spring Harb Perspect Biol. 2015;7(2):a016311. Manuela Costa for their technical assistance. 4. Francos-Quijorna I, Santos-Nogueira E, Gronert K, Sullivan AB, Kopp MA, Brommer B, et al. Maresin 1 promotes inflammatory resolution, neuro‑ Authors’ contribution protection, and functional neurological recovery after spinal cord injury. J RL-V designed the study, analysed and interpreted the data, and wrote the Neurosci. 2017;37:11731–43. manuscript. AS-F collected murine samples, performed the experiments, 5. López-Vales R, David S. Bioactive lipids in inflammation after central nerv‑ analysed and interpreted the data, and wrote the manuscript. SZ collected ous system injury. In: Trostchansky A, Rubbo H, editors. Bioactive lipids in human samples, interpreted data and reviewed the manuscript. MC assisted health and disease advances in experimental medicine and biology, vol. in the collection of human samples and reviewed the manuscript. MM and 1127. Springer: Cham; 2019. p. 181–94. HR performed the metabololipidomic analyses in human and murine samples 6. Prüss H, Rosche B, Sullivan AB, Brommer B, Wengert O, Gronert K, et al. and reviewed the manuscript. AP provided human samples. All authors read Proresolution lipid mediators in multiple sclerosis—differential, disease and approved the final manuscript. severity-dependent synthesis—a clinical pilot trial, Bozza PT. PLoS ONE. 2013;8:e55859. Funding 7. Schwab JM, Chiang N, Arita M, Serhan CN. Resolvin E1 and protectin D1 This work was supported by grants from Spanish Ministry of Economy activate inflammation-resolution programmes. Nature. 2007;447:869–74. and Competitiveness (SAF2016-79774-R), Spanish Ministry of Science and 8. Ruiz F, Vigne S, Pot C. Resolution of inflammation during multiple sclero‑ Innovation (PID2020-120267RB-I00/AEI/1013039/501100011033), Wings for sis. Semin Immunopathol. 2019;41:711–26. Life International Foundation, Red de Terapia Celular (TERCEL), funds from 9. Serhan CN, Yang R, Martinod K, Kasuga K, Pillai PS, Porter TF, et al. Mares‑ “la Caixa” Foundation under agreement LCF/TR/CI17/10020018 and LCF/ ins: novel macrophage mediators with potent antiinflammatory and PR/HA17/52170001 to R.L-V, and by funds from CSIC Grupos Universidad de proresolving actions. J Exp Med. 2009;206:15–23. la República, Uruguay (Grant 536) to HR. A.P. holds the T1 (senior) Canada 10 Chiang N, Serhan CN. Structural elucidation and physiologic functions Research Chair in MS. of specialized pro-resolving mediators and their receptors. Mol Aspects Med. 2017;58:114–29. Availability of data and materials 11. Serhan CN. Treating inflammation and infection in the 21st century: new All the data generated or analysed during this study are included in this pub‑ hints from decoding resolution mediators and mechanisms. FASEB J. lished article ant its supplementary information files. 2017;31:1273–88. 12. Wang L, Wang F-S, Gershwin ME. Human autoimmune diseases: a com‑ prehensive update. J Intern Med. 2015;278:369–95. Declarations 13. Bhise V, Dhib-Jalbut S. Further understanding of the immunopathology of multiple sclerosis: impact on future treatments. Expert Rev Clin Immu‑ Ethics approval and consent to participate nol. 2016;12:1069–89. Tissue samples were collected from MS patients with full ethical approval 14. Dendrou CA, Fugger L, Friese MA. Immunopathology of multiple sclero‑ (BH07.001) and informed consent was approved by the local ethics com‑ sis. Nat Rev Immunol. 2015;15:545–58. mittee in Montreal (Canada). All experimental procedures were approved 15 Popescu BFG, Lucchinetti CF. Immunopathology. Handb Clin Neurol. by the Universitat Autònoma de Barcelona Animal Experimentation Ethical 2016. https://​doi.​org/​10.​1016/​B978-0-​444-​63432-0.​00006-2. Committee (CEEAH 2878) and followed the European Communities Council 16. Prat A, Antel J. Pathogenesis of multiple sclerosis. Curr Opin Neurol. Directive 2010/63/EU, and the methods were carried out in accordance with 2005;18:225–30. the approved guidelines. 17. Lassmann H, Brück W, Lucchinetti CF. The immunopathology of multiple sclerosis: An overview. Brain Pathol. 2007;17:210–8. Consent for publication 18. Prineas JW, Kwon EE, Cho ES, Sharer LR, Barnett MH, Oleszak EL, et al. Not applicable. Immunopathology of secondary-progressive multiple sclerosis. Ann Neurol. 2001;50:646–57. Competing interests 19. Poisson LM, Suhail H, Singh J, Datta I, Denic A, Labuzek K, et al. Untar‑ A.S.-F and R.L.-V. are inventors on a patent application filled by Universitat geted plasma metabolomics identifies endogenous metabolite with Autònoma de Barcelona “Specialized pro-resolving lipid mediators for use in drug-like properties in chronic animal model of multiple sclerosis. J Biol the treatment of neurodegenerative diseases and/or autoimmune diseases” Chem. 2015;290:30697–712. (PCT EP2018/051076). The remaining authors declare no competing financial 20. Kooij G, DeradaTroletti C, Leuti A, Norris PC, Riley I, Albanese M, et al. interests. Specialized pro-resolving lipid mediators are differentially altered in peripheral blood of patients with multiple sclerosis and attenuate Sánchez‑Fernández et al. Journal of Neuroinflammation (2022) 19:27 Page 18 of 18 monocyte and blood-brain barrier dysfunction. Haematologica. 42. Torkildsen O, Wergeland S, Bakke S, Beiske AG, Bjerve KS, Hovdal H, et al. 2019;105(8):2056–70. ω-3 fatty acid treatment in multiple sclerosis (OFAMS Study). Arch Neurol. 21. DeradaTroletti C, Enzmann G, Chiurchiù V, Kamermans A, Tietz SM, Norris 2012;69:1044–51. PC, et al. Pro-resolving lipid mediator lipoxin A4 attenuates neuro-inflam‑ 43. Wang X, Zhu M, Hjorth E, Cortés-Toro V, Eyjolfsdottir H, Graff C, et al. mation by modulating T cell responses and modifies the spinal cord Resolution of inflammation is altered in Alzheimer’s disease. Alzheimer’s lipidome. Cell Rep. 2021;35:109201. Dement. 2015;11(40–50):e1-2. 22. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, et al. 44. Serhan CN, Dalli J, Colas RA, Winkler JW, Chiang N. Protectins and Diagnostic criteria for multiple sclerosis: 2010 Revisions to the McDonald maresins: New pro-resolving families of mediators in acute inflam‑ criteria. Ann Neurol. 2011;69:292–302. mation and resolution bioactive metabolome. Biochim Biophys Acta. 23. Dhaeze T, Tremblay L, Lachance C, Peelen E, Zandee S, Grasmuck C, et al. 2015;1851:397–413. CD70 defines a subset of proinflammatory and CNS-pathogenic TH1/ 45. Ohuchi K, Ono Y, Joho M, Tsuruma K, Ogami S, Yamane S, et al. A Doco‑ TH17 lymphocytes and is overexpressed in multiple sclerosis. Cell Mol sahexaenoic acid-derived pro-resolving agent, maresin 1, protects motor Immunol. 2019;16:652–65. neuron cells death. Neurochem Res. 2018;43:1413–23. 24. Kuhlmann T, Ludwin S, Prat A, Antel J, Brück W, Lassmann H. An updated 46. Xian W, Li T, Li L, Hu L, Cao J. Maresin 1 attenuates the inflammatory histological classification system for multiple sclerosis lesions. Acta Neu‑ response and mitochondrial damage in mice with cerebral ischemia/ ropathol. 2017;133:13–24. reperfusion in a SIRT1-dependent manner. Brain Res. 2019;1711:83–90. 25. Arbour N, Holz A, Sipe JC, Naniche D, Romine JS, Zyroff J, et al. A new 47. Lévesque SA, Paré A, Mailhot B, Bellver-Landete V, Kébir H, Lécuyer M-A, approach for evaluating antigen-specific T cell responses to myelin et al. Myeloid cell transmigration across the CNS vasculature triggers antigens during the course of multiple sclerosis. J Neuroimmunol. IL-1β–driven neuroinflammation during autoimmune encephalomyelitis 2003;137:197–209. in mice. J Exp Med. 2016;213:929–49. 26. Livak KJ, Schmittgen TD. Analysis of relative gene expression data 48. Schrijver HM, Van As J, Crusius JBA, Dijkstra CD, Uitdehaag BMJ. Interleu‑ using real-time quantitative PCR and the 2-ΔΔCT method. Methods. kin (IL)-1 gene polymorphisms: Relevance of disease severity associated 2001;25:402–8. alleles with IL-1β and IL-1ra production in multiple sclerosis. Mediators 27. Mastrogiovanni M, Ifrán E, Trostchansky A, Rubbo H. Lipidomic analysis of Inflamm. 2003;12:89–94. oxygenated polyunsaturated fatty acid–derived inflammatory mediators 49. Serada S, Fujimoto M, Mihara M, Koike N, Ohsugi Y, Nomura S, et al. IL-6 in neurodegenerative diseases. Biochem Soc Trans. 2011;39(5):1240–6. blockade inhibits the induction of myelin antigen-specific Th17 cells and 28 Amo-Aparicio J, Martínez-Muriana A, Sánchez-Fernández A, López-Vales Th1 cells in experimental autoimmune encephalomyelitis. Proc Natl Acad R. Neuroinflammation Quantification for Spinal Cord Injury. Curr Protoc Sci. 2008;105:9041–6. Immunol. 2018;123:1–15. 50. Kallaur AP, Oliveira SR, Simao ANC, De Almeida ERD, Morimoto HK, Lopes 29. Sapieha P, Stahl A, Chen J, Seaward MR, Willett KL, Krah NM, et al. J, et al. Cytokine profile in relapsing-remitting Multiple sclerosis patients 5-Lipoxygenase metabolite 4-HDHA is a mediator of the antiangiogenic and the association between progression and activity of the disease. Mol effect of ω-3 polyunsaturated fatty acids. Sci Transl Med. 2011. https://​doi.​ Med Rep. 2013;7:1010–20. org/​10.​1126/​scitr​anslm​ed.​30015​71. 51. Rossi S, Motta C, Studer V, Macchiarulo G, Volpe E, Barbieri F, et al. 30. Chiurchiù V, Leuti A, Dalli J, Jacobsson A, Battistini L, Maccarrone M, et al. Interleukin-1β causes excitotoxic neurodegeneration and multiple Proresolving lipid mediators resolvin D1, resolvin D2, and maresin 1 are sclerosis disease progression by activating the apoptotic protein p53. Mol critical in modulating T cell responses. Sci Transl Med. 2016. https://​doi.​ Neurodegener. 2014;9:56. org/​10.​1126/​scitr​anslm​ed.​aaf74​83. 52. Sun D, Zhang M, Sun P, Liu G, Strickland AB, Chen Y, et al. VCAM1/VLA4 31. David S, Greenhalgh AD, López-Vales R. Role of phospholipase A 2s and interaction mediates Ly6Clow monocyte recruitment to the brain in a lipid mediators in secondary damage after spinal cord injury. Cell Tissue TNFR signaling dependent manner during fungal infection. PLOS Pathog. Res. 2012;349:249–67. 2020;16:e1008361. 32. Robinson DR. Lipid mediators of inflammation. Rheum Dis Clin North Am. 1987;13:385–405. 33. Belury MA, Harris WS. Omega-6 fatty acids, inflammation and cardiometa‑ Publisher’s Note bolic health: Overview of supplementary issue. Prostaglandins, Leukot Springer Nature remains neutral with regard to jurisdictional claims in pub‑ Essent Fat Acids. 2018;139:1–2. lished maps and institutional affiliations. 34. Dennis EA, Norris PC. Eicosanoid storm in infection and inflammation. Nat Rev Immunol. 2015;15:511–23. 35. Diab A, Deng C, Smith JD, Hussain RZ, Phanavanh B, Lovett-Racke AE, et al. Peroxisome proliferator-activated receptor- agonist 15-Deoxy- 12,1412,14-prostaglandin J2 ameliorates experimental autoimmune encephalomyelitis. J Immunol. 2002;168:2508–15. 36. Kerr BJ, Girolami EI, Ghasemlou N, Jeong SY, David S. The protective effects of 15-deoxy-Δ-12,14-prostaglandin J2 in spinal cord injury. Glia. 2008;56:436–48. 37 Buckley CD, Gilroy DW, Serhan CN. Proresolving lipid mediators and mechanisms in the resolution of acute inflammation. Immunity. 2014;40:315–27. Ready to submit your research ? Choose BMC and benefit from: 38. Serhan CN, Chiang N. Endogenous pro-resolving and anti-inflamma‑ tory lipid mediators: a new pharmacologic genus. Br J Pharmacol. • fast, convenient online submission 2009;153:S200–15. 39. de Camargo TL, de Oliveira EP, Moreto F, Portero-McLellan KC, Burini RC. • thorough peer review by experienced researchers in your field Omega-3 fatty acids supplementation decreases metabolic syndrome • rapid publication on acceptance prevalence after lifestyle modification program. J Funct Foods Elsevier. • support for research data, including large and complex data types 2015;19:922–8. 40 Marik PE, Varon J. Omega-3 dietary supplements and the risk of cardio‑ • gold Open Access which fosters wider collaboration and increased citations vascular events: a systematic review. Clin Cardiol. 2009;32:365–72. • maximum visibility for your research: over 100M website views per year 41. Riccio P, Rossano R, Larocca M, Trotta V, Mennella I, Vitaglione P, et al. Anti- inflammatory nutritional intervention in patients with relapsing-remitting At BMC, research is always in progress. and primary-progressive multiple sclerosis: a pilot study. Exp Biol Med. 2016;241:620–35. Learn more biomedcentral.com/submissions