Myeloid cell plasticity in the evolution of central nervous system autoimmunity.

OBJECTIVE: Myeloid cells, including macrophages and dendritic cells, are a prominent component of central nervous system (CNS) infiltrates during multiple sclerosis (MS) and the animal model experimental autoimmune encephalomyelitis (EAE). Although myeloid cells are generally thought to be proinflammatory, alternatively polarized subsets can serve noninflammatory and/or reparative functions. Here we investigate the heterogeneity and biological properties of myeloid cells during central nervous system autoimmunity. METHODS: Myeloid cell phenotypes in chronic active MS lesions were analyzed by immunohistochemistry. In addition, immune cells were isolated from the CNS during exacerbations and remissions of EAE and characterized by flow cytometric, genetic, and functional assays. RESULTS: Myeloid cells expressing inducible nitric oxide synthase (iNOS), indicative of a proinflammatory phenotype, were detected in the actively demyelinating rim of chronic active MS lesions, whereas macrophages expressing mannose receptor (CD206), a marker of alternatively polarized human myeloid cells, were enriched in the quiescent lesion core. During EAE, CNS-infiltrating myeloid cells, as well as microglia, shifted from expression of proinflammatory markers to expression of noninflammatory markers immediately prior to clinical remissions. Murine CNS myeloid cells expressing the alternative lineage marker arginase-1 (Arg1) were partially derived from iNOS+ precursors and were deficient in activating encephalitogenic T cells compared with their Arg1- counterparts. INTERPRETATION: These observations demonstrate the heterogeneity of CNS myeloid cells, their evolution during the course of autoimmune demyelinating disease, and their plasticity on the single cell level. Future therapeutic strategies for disease modification in individuals with MS may be focused on accelerating the transition of CNS myeloid cells from a proinflammatory to a noninflammatory phenotype. Ann Neurol 2018;83:131-141.

An IFNγ/CXCL2 regulatory pathway determines lesion localization during EAE.

BACKGROUND: Myelin oligodendrocyte glycoprotein (MOG)-reactive T-helper (Th)1 cells induce conventional experimental autoimmune encephalomyelitis (cEAE), characterized by ascending paralysis and monocyte-predominant spinal cord infiltrates, in C57BL/6 wildtype (WT) hosts. The same T cells induce an atypical form of EAE (aEAE), characterized by ataxia and neutrophil-predominant brainstem infiltrates, in syngeneic IFNγ receptor (IFNγR)-deficient hosts. Production of ELR+ CXC chemokines within the CNS is required for the development of aEAE, but not cEAE. The cellular source(s) and localization of ELR+ CXC chemokines in the CNS and the IFNγ-dependent pathways that regulate their production remain to be elucidated. METHODS: The spatial distribution of inflammatory lesions and CNS expression of the ELR+ CXC chemokines, CXCL1 and CXCL2, were determined via immunohistochemistry and/or in situ hybridization. Levels of CXCL1 and CXCL2, and their cognate receptor CXCR2, were measured in/on leukocyte subsets by flow cytometric and quantitative PCR (qPCR) analysis. Bone marrow neutrophils and macrophages were cultured with inflammatory stimuli in vitro prior to measurement of CXCL2 and CXCR2 by qPCR or flow cytometry. RESULTS: CNS-infiltrating neutrophils and monocytes, and resident microglia, are a prominent source of CXCL2 in the brainstem of IFNγRKO adoptive transfer recipients during aEAE. In WT transfer recipients, IFNγ directly suppresses CXCL2 transcription in microglia and myeloid cells, and CXCR2 transcription in CNS-infiltrating neutrophils. Consequently, infiltration of the brainstem parenchyma from the adjacent meninges is blocked during cEAE. CXCL2 directly stimulates its own expression in cultured neutrophils, which is enhanced by IL-1 and suppressed by IFNγ. CONCLUSIONS: We provide evidence for an IFNγ-regulated CXCR2/CXCL2 autocrine/paracrine feedback loop in innate immune cells that determines the location of CNS infiltrates during Th1-mediated EAE. When IFNγ signaling is impaired, myeloid cell production of CXCL2 increases, which promotes brainstem inflammation and results in clinical ataxia. IFNγ, produced within the CNS of WT recipients, suppresses myeloid cell CXCR2 and CXCL2 production, thereby skewing the location of neuroinflammatory infiltrates to the spinal cord and the clinical phenotype to an ascending paralysis. These data reveal a novel mechanism by which IFNγ and CXCL2 interact to direct regional recruitment of leukocytes in the CNS, resulting in distinct clinical presentations.

CNS-resident classical DCs play a critical role in CNS autoimmune disease.

Experimental autoimmune encephalomyelitis (EAE) is an inflammatory demyelinating disease of the central nervous system (CNS), induced by the adoptive transfer of myelin-reactive CD4+ T cells into naive syngeneic mice. It is widely used as a rodent model of multiple sclerosis (MS). The development of EAE lesions is initiated when transferred CD4+ T cells access the CNS and are reactivated by local antigen-presenting cells (APCs) bearing endogenous myelin peptide/MHC class II complexes. The identity of the CNS-resident, lesion-initiating APCs is widely debated. Here we demonstrate that classical dendritic cells (cDCs) normally reside in the meninges, brain, and spinal cord in the steady state. These cells are unique among candidate CNS APCs in their ability to stimulate naive, as well as effector, myelin-specific T cells to proliferate and produce proinflammatory cytokines directly ex vivo. cDCs expanded in the meninges and CNS parenchyma in association with disease progression. Selective depletion of cDCs led to a decrease in the number of myelin-primed donor T cells in the CNS and reduced the incidence of clinical EAE by half. Based on our findings, we propose that cDCs, and the factors that regulate them, be further investigated as potential therapeutic targets in MS.

A randomized, subject and rater-blinded, placebo-controlled trial of dimethyl fumarate for obstructive sleep apnea.

Study Objectives: To investigate the therapeutic effect of dimethyl fumarate (DMF, an immunomodulatory agent) on obstructive sleep apnea (OSA), and potential influence of any such effect by selected proinflammatory molecules. Methods: Patients with OSA who deferred positive airway pressure therapy were randomized (2:1) to receive DMF or placebo for 4 months. Participants underwent polysomnography before randomization and at 4 months. Blood was collected monthly. The primary outcome was the mean group change in respiratory disturbance index (δ-RDI). Secondary analyses focused on the association between treatment effect of DMF (on RDI) and expression of plasma cytokines and chemokines, or nuclear factor κ-B (NFκB) signaling molecules in peripheral blood mononuclear cells. Results: N = 65 participants were randomized. N = 50 participants (DMF = 35, placebo = 15) had complete data for final analyses. The mean difference in δ-RDI between groups was 13.3 respiratory events/hour of sleep: -3.1+/-12.9 vs. 10.2+/-13.1 in DMF and placebo groups, respectively (mixed-effects model treatment effect: ß = -0.14, SE = 0.062, p = 0.033). Plasma levels of TNF-α showed only nonsignificant decreases, and IL-10 and IL-13 only nonsignificant increases, in DMF-treated participants compared with placebo. No significant interaction or main effect on RDI for selected cytokines and chemokines was found. Participants with a therapeutic response to DMF did experience significant reductions in intracellular NFκB signaling molecules at 4 months. Overall, DMF was well-tolerated. Conclusions: The immunomodulatory drug DMF partially ameliorates OSA severity. Suppression of systemic inflammation through reduction of NFκB signaling may mediate this effect. Clinical Trials: ClinicalTrials.gov, NCT02438137, https://clinicaltrials.gov/ct2/show/NCT02438137?term=NCT02438137&rank=1.