Background: In early autoimmune neuroinflammation, interferon (IFN)γ and its upregulation of the immunoproteasome (iP) is pathologic. However, during chronic multiple sclerosis (MS), IFNγ is shown to have protective properties. Although dysregulation of the iP has been implicated in neurodegeneration, its function remains to be fully elucidated. Further, although MS affects the spinal cord in a majority of MS patients, there is still much to learn regarding its pathology. Here, we demonstrate that IFNγ signaling in spinal cord astrocytes induces the iP and promotes protection of the central nervous system (CNS) during chronic autoimmunity.
Methods: In MS brain, we evaluated mRNA expression and labeled postmortem MS brainstem and spinal cord for iP subunits and indicators of oxidative stress. Primary regional human astrocytes were analyzed for iP regulation and function by quantitative PCR, Western blot, OxyBlot, and reactive oxygen species and caspase activity detection assays. Following immunization with myelin oligodendrocyte glycoprotein (MOG)35-55, the role of IFNγ signaling and the iP during chronic experimental autoimmune encephalomyelitis (EAE) were assessed using pharmacologic inhibition of the iP and genetic interruption of IFNγ signaling specifically in astrocytes. CNS tissues were analyzed by immunofluorescent analysis and cell-specific colocalization was quantified.
Results: In MS tissue, iP expression was enhanced in spinal cord compared to brainstem lesions, which correlated with a decrease in oxidative stress. In vitro, IFNγ stimulation enhanced iP expression, reduced reactive oxygen species burden, and decreased oxidatively damaged and poly-ubiquitinated protein accumulation preferentially in human spinal cord astrocytes, which was abrogated with the use of the iP inhibitor, ONX 0914. During the chronic phase of an MS animal model, EAE, ONX 0914 treatment exacerbated disease and led to increased oxidative stress and poly-ubiquitinated protein build-up. Finally, mice with astrocyte-specific loss of the IFNγ receptor exhibited worsened chronic EAE associated with reduced iP expression, enhanced lesion size and oxidative stress and poly-ubiquitinated protein accumulation in astrocytes.
Conclusions: Taken together, our data reveal a protective role for IFNγ signaling in spinal cord astrocytes during chronic neuroinflammation and identify a novel function of the iP in astrocytes during CNS autoimmunity.
A significant proportion of the inflammatory cells that infiltrate the central nervous system (CNS) of patients with multiple sclerosis (MS) and animals with experimental autoimmune encephalomyelitis (EAE), are myeloid cells, including monocyte-derived macrophages and dendritic cells. A large body of literature supports a pathogenic role of these subsets during the acute phase of CNS demyelinating lesion formation. However, we and others have shown that CNS-infiltrating myeloid cells are more heterogeneous than traditionally thought, and that their phenotype evolves during the clinical course. Arginase 1 (Arg1) is an inhibitor of nitric oxide (NO) and a canonical marker of alternatively activated myeloid cells (AAMC) with anti-inflammatory properties. Arg1+ myeloid cells accumulate in the CNS from peak EAE through remission, but are not detectable in the blood or peripheral tissues. Furthermore, myeloid cells expressing AAMC markers are dominant in the quiescent core, as opposed to the active demyelinating rim, of MS lesions.
The objective of this study is to determine the role of AAMCs in the resolution of inflammation and recovery of neurological function in a relapsing-remitting model of EAE.
To address that hypothesis, we constructed a transgenic mouse strain with the human diphtheria toxin receptor (hDTR) gene (Hbegf) linked to the Arg1 promoter (Arg1-hDTR mice). In order to restrict hDTR expression to hematopoietic cells we reconstituted lethally irradiated WT mice with bone marrow cells from Arg1-hDTR donors. EAE was induced in Arg1-hDTR→WT and WT→WT chimeric mice by the adoptive transfer of purified myelin-reactive CD4+ TH17 cells and treated with diphtheria toxin (DT). CNS inflammatory cell composition was evaluated via flow cytometry and cell subsets were isolated via FACS for transcriptional analysis.
Depletion of Arg1+ inflammatory cells in Arg1-hDTR→WT chimeric mice exacerbated EAE induced by the transfer of encephalitogenic TH17 cells, and resulted in a high mortality rate compared with PBS-treatment and WT→WT controls. Monocyte-derived dendritic cells (mDC) possessing an anti-inflammatory phenotype were most susceptible to depletion. The more severe clinical course observed following depletion was associated with an increase in the absolute number of Arg1- inflammatory macrophages present in the spinal cord at peak disease, coincident with elevated levels of proinflammatory cytokines and chemokines in spinal cord homogenates. Conversely, targeted depletion of Arg1 in immune cells had no impact on EAE incidence or severity.
This study suggests that CNS accumulation of AAMC during later stages of autoimmune demyelinating disease suppresses the local inflammatory response and ameliorates neurological deficits by an Arg1- independent mechanism.
Cortical plaque loads in primary progressive (PP) and secondary progressive MS (SP) appear equal whereas meningeal perivascular mononuclear/lymphocytic infiltration (PMI) may be more pronounced in SP. Basal ganglia may also be a predilection site in progressive MS.
To compare cortical and basal ganglia plaque formation and PMI in large tissue areas from clinically well-defined PP and SP patients.
Large brain sections from clinically well-described patients with PP (N=12), SP (N=14) and 11 age-matched controls were stained with HE, Luxol-fast blue and in patients additionally for proteolipid protein and CD68. T- and B-cells were confirmed in PMI’s by CD3 and CD 20 in 3 PP and 3 SP patients. We measured total plaque, microglia-rimmed slowly expanding plaque and macrophage-filled filled active plaque area loads (% of cortex or basal ganglia area), PMI densities (#/Cm2 or #/Cm meningeal length) and mean PMI size. PMI’s with min. size score one had 3-4 perivascular cell rows and 20-50 cells whereas max. score six had >19 rows and >500 cells. In one PP patient, cortical slowly expanding rims were confirmed by immunofluorescence showing IgG. Expanded Disability Status Scale (EDSS) score was estimated by chart review in a blinded fashion among 20 patients (10 PP and 12SP) from whom charts were available from two time points including one within three yrs. before death. Among these, we calculated ΔEDSS/yr. before death.
Tissue areas and meningeal lengths were similar in PP and SP. Both groups displayed similarly elevated cortical plaque loads and PMI densities in the cortex and meninges. However, meningeal PMI size as well as basal ganglia (6 SP vs 8 PP) plaque load and PMI density tended to be higher in SP vs PP. Meningeal PMI density correlated with total cortical plaque load while meningeal PMI size correlated with cortical slowly expanding plaque. Cortical PMI density and size correlated with active and slowly expanding (but not total) cortical plaque. PMI’s and plaque also correlated in the basal ganglia, albeit only in SP. Overall, age at death correlated negatively with PMI’s in the meninges, cortex and in the basal ganglia. Δ EDSS/yr. before death correlated with plaque load and tended to correlate with PMI’s, albeit only in the basal ganglia.
Perivascular immune activation in the meninges, cortex and basal ganglia may be pathogenic, driving grey matter demyelination in progressive MS.
Multiple Sclerosis (MS) is characterized by central nervous system infiltration of peripheral immune cells, the largest fraction of which are macrophages. The macrophage role in MS is multifaceted, in a pro-inflammatory or ‘M1’ state instructing demyelination and axonal loss, while the anti-inflammatory ‘M2’ state holds a key role in tissue repair and regeneration. Previously, we have identified that IL-10 inhibition of miR-155 is a prominent mechanism utilized by macrophages to maintain an M2 state. Moreover, using a miR-155 floxed x LysMCre model, where miR-155 is specifically deleted from myeloid cells, there was reduced disease onset and less lesion burden in the experimental autoimmune encephalomyelitis (EAE) animal model. Thus, we hypothesize miR-155 inhibition may favorably modulate the macrophage population to an ‘M2’ or pro-repair phenotype, reducing inflammation, alleviating disease progression, mimicking an IL-10 mediated effect.
To investigate the therapeutic potential of a miR-155 anti-miRNA oligonucleotide (AMO) packaged in nanoparticle-based carriers to enhance uptake into macrophages.
4 AMOs were investigated for their ability to inhibit mir-155 in Raw 264.7 and bone marrow-derived macrophages (BMDM). The downstream effect of macrophage pro-inflammatory function in response to mir-155 inhibition was examined by measuring a range of macrophage polarisation parameters, including pro-inflammatory cytokine and nitric oxide (NO) production, expression of M2 markers Arginase-1 and CD206, two markers intricately tied with metabolism in the context of polarisation. PLGA and novel star-shaped polypeptides were also assessed for in vitro macrophage delivery.
A locked nucleic acid (LNA) modified AMO showed the most promising results for mir-155 inhibition in both Raw 264.7 and BMDM. In further studies we show changes in expression of mir-155 target genes that mimic an IL-10 mediated phenotype, while mir-155 independent increases in M2 marker Arginase-1 suggest a more M2 like phenotype. Additionally, star-shaped polypeptides demonstrate the capacity for AMO delivery to BMDM.
mir-155 inhibition can be achieved through delivery of an AMO, and mimics crucial aspects of an IL-10 mediated macrophage phenotype. Star polypeptides represent a promising avenue towards macrophage-specific uptake and future in vivo delivery in MS animal models.