Intrastriatal injection of interleukin-1 beta triggers the formation of neuromyelitis optica-like lesions in NMO-IgG seropositive rats

Abstract

Background: Neuromyelitis optica (NMO) is a severe, disabling disease of the central nervous system (CNS) characterized by the formation of astrocyte-destructive, neutrophil-dominated inflammatory lesions in the spinal cord and optic nerves. These lesions are initiated by the binding of pathogenic aquaporin 4 (AQP4)-specific autoantibodies to astrocytes and subsequent complement-mediated lysis of these cells. Typically, these lesions form in a setting of CNS inflammation, where the blood-brain barrier is open for the entry of antibodies and complement. However, it remained unclear to which extent pro-inflammatory cytokines and chemokines contribute to the formation of NMO lesions. To specifically address this question, we injected the cytokines interleukin-1 beta, tumor necrosis factor alpha, interleukin-6, interferon gamma and the chemokine CXCL2 into the striatum of NMO-IgG seropositive rats and analyzed the tissue 24 hours later by immunohistochemistry.

Results: All injected cytokines and chemokines led to profound leakage of immunoglobulins into the injected hemisphere, but only interleukin-1 beta induced the formation of perivascular, neutrophil-infiltrated lesions with AQP4 loss and complement-mediated astrocyte destruction distant from the needle tract. Treatment of rat brain endothelial cells with interleukin-1 beta, but not with any other cytokine or chemokine applied at the same concentration and over the same period of time, caused profound upregulation of granulocyte-recruiting and supporting molecules. Injection of interleukin-1 beta caused higher numbers of blood vessels with perivascular, cellular C1q reactivity than any other cytokine tested. Finally, the screening of a large sample of CNS lesions from NMO and multiple sclerosis patients revealed large numbers of interleukin-1 beta-reactive macrophages/activated microglial cells in active NMO lesions but not in MS lesions with comparable lesion activity and location.

Conclusions: Our data strongly suggest that interleukin-1 beta released in NMO lesions and interleukin-1 beta-induced production/accumulation of complement factors (like C1q) facilitate neutrophil entry and BBB breakdown in the vicinity of NMO lesions, and might thus be an important secondary factor for lesion formation, possibly by paving the ground for rapid lesion growth and amplified immune cell recruitment to this site.

Figure 1

Blood–brain barrier breakdown induced by the intrastriatal injection of cytokines and chemokines, as indicated by the extravasation of immunoglobulins. (a-l) Cerebral hemispheres from juvenile NMO-IgG seropositive Lewis rats stained for the presence of rat immunoglobulins (brown; a-f) and adjacent sections stained for the presence of human immunoglobulins (brown; g-l) 20–24 hrs after intrastriatal injections of 0.3 μl PBS containing 30 ng of IL-1β (a,g), TNF-α (b,h), IFN-γ (c,i), CXCL2 (d,j), IL-6 (e,k) or no further additives (f,l). The differences of IgG leakage between the different experimental groups was not significant. Dashed line: needle tract. Scale bar: 1 mm.

Figure 2

Pathological findings at the needle tract. The extent of AQP4 loss (a-c) and the infiltration of W3/13⁺ granulocytes (d-f), ED1⁺ macrophages/activated microglial cells (g-i), and CD3⁺ T cells (j-l) are presented. All histological samples and quantitative data (c,f,i,l) shown derive from the striatum of 3-week old Lewis rats injected with 30 ng of IL-1β (histological data), and of IL-1β, TNF-α, IL-6, IFN-γ, CXCL2, or vehicle (PBS, veh) as indicated (quantifications). Animals were seropositive for NMO IgG or control IgG as indicated. Statistically significant differences between individual cytokine treatments are shown (*p<0.05, ANOVA-Dunett T3). Scale bars: 1 mm (a,d,g,j) and 100 μm (b,e,h,k).

Figure 3

Location and frequency of lesions with AQP4 loss and neutrophil infiltrates outside the needle tract, after intrastriatal injection of IL-1β and intraperitoneal application of NMO-IgG. The location (red dots in the schemes of [6]; a,b) and histology (c,d) of perivascular lesions with AQP4 loss and neutrophil infiltration in adult animals seropositive for the NMO-IgGs, J0 (a,c) and I GF (b,d), that received intrastriatal injection of IL-1β (gray area represents the injection site). Similar lesions were not observed after transfer of IgGs from a NMO-IgG negative NMO patient (e), or three NMO-IgG negative MS patients (f-h). The number of those lesions was significantly higher in IL-1β/NMO-IgG injected animals (n=29) than in the controls (n=47), according to Mann Whitney U test with Bonferroni Holm correction (i). The following NMO-IgG or control IgG preparations were used: NMO-IgG J0 (1, 2, 7-14), NMO- IgG I GF (3, 4), NMO-IgG7 (5) and NMO-IgG8 (6). As controls, IgG preparations of two AQP4 antibody negative NMO patients (J3 and J4; 15), three NMO-IgG negative MS patients (J5, J6, J7; 16 and 17), and subcuvia (18, 19) were used. Experiments 1, 3, 7, 9, 11, 12, 13, 15 and 18 were performed using juvenile animals, and experiments 2, 4, 5, 6, 8, 10, 14, 16, 17 and 19 using adult rats. Cytokines were injected as indicated. The AQP4-specific antibodies found in the NMO-IgG preparations are responsible for the formation of lesions, as revealed by absorption studies using two different NMO-IgG preparations (j-l). Lesions were present (j, l; white circles) when NMO-IgG had been exposed to emGFP transfected HEK 293 cells, but were absent (k, l; black circles) when NMO-IgG had been exposed to AQP4-emGFP transfected HEK 293 cells. Bar=25 µm.

Figure 4

The effects of cytokines and chemokines on rat brain endothelial cells in vitro. Rat brain endothelial cells (EC) were isolated from the cerebra of 3-week old (a,c) and adult (b,d) Lewis rats, and are shown here in phase-contrast microscopy (a,b) and after staining with an antibody against zonula occludens 1 (ZO-1, red; c, d; scale bar = 25 μm). (e) Endothelial cells from 3-week old (left panel) and adult (right panel) Lewis rats were stimulated for 22 hours with vehicle, IL-1β, IL-6, IFN-γ, TNF-α, or CXCL2, and then subjected to PCR analysis of the following genes: Cxcl1, Cxcl2, Csf3, Ccl2, Ccl5, Icam1, Vcam1, and GAPDH. These data are representative of 2 different, independently performed experiments. (f,g,h) The lysates of rat brain endothelial cells treated for 12 hours with IL-1β (f) or vehicle (g) were analyzed with antibody arrays and revealed changes in protein expression of CXCL1 (red rectangle), CXCL3 (gray rectangle) and CCL2 (green rectangle). The fold-change in signal intensity of these different proteins was then calculated using the Image J software of the National Institute of Health. (i,j) Rat brain endothelial cells cultured for 24 hours in the presence of IL-1β (i) reveal a marked increase in ICAM1 protein expression (green), which was not seen after 24 hours of culture in the presence of vehicle (j). All cells were counterstained with antibodies against ZO-1 (red). Bar = 25 μm. Data are representative of 2 different experiments.

Figure 5

Perivascular C1q reactivity in the cytokine-injected hemisphere of animals harboring serum NMO-IgG. Histological analysis (a-f) of blood vessels with cellular C1q reactivity covering > 50% of the vessel abluminal surface, which were found after the intrastriatal injection of IL-1β (a,b) and TNF-α (c,d), but not after injection of vehicle (PBS, e,f), IFN-γ, IL-6 or CXCL2 (not shown). Consecutive sections were stained with antibodies against C1q (red, a,c,e) and human IgG (brown, b,d,f), and counterstained with hematoxylin to reveal nuclei (blue). C1q⁺ glial cells are indicated by arrows. Scale bars: 25 μm.

Figure 6

IL-1β expression in NMO lesions. IL-1β is expressed in active lesions of NMO patients (a). Perivascular lesions with pronounced IL-1β expression (b) show the characteristic pattern of complement C9neo deposition (c). IL-1β reactivity was absent in active lesions of acute MS cases (d,g), which were characterized by a high extent of demyelination (e, Kluever staining) and marked microglia activation (f, shown here by Iba1 immunostaining). No signal in NMO lesions was detected upon application of secondary antibody alone (h). Confocal microscopy shows macrophages in active NMO lesions stained with IL-1β⁺ (green, i) and Iba1⁺ (red, j). An overlay of these two different stainings is shown in (k). Scale bars: 100 μm (a,d,e,f,h), 50 μm (b,c), 25 μm (g), and 10 μm (i-k).