A role for the plasminogen activator system in inflammation and neurodegeneration in the central nervous system during experimental allergic encephalomyelitis

Abstract

Early signs of inflammatory demyelination include entry of fibrin(ogen) into the central nervous system (CNS), which is normally excluded by the blood-brain barrier, and up-regulation of components of the plasminogen activator system. Using mice deficient in tissue-type plasminogen activator (tPA-/-) and urokinase plasminogen activator receptor (uPAR-/-), we investigated the involvement of the PA system on the clinical and pathological features of experimental allergic encephalomyelitis, an animal model of multiple sclerosis. tPA-/- mice suffered an early and a more severe acute disease characterized by incomplete recovery when compared to wild-type controls, with significantly higher CNS levels of plasminogen activator inhibitor-1. This correlated with fibrin accumulation, which co-localized with nonphosphorylated neurofilament on thickened axons in experimental allergic encephalomyelitis tissue. In contrast, uPAR-/- mice had a delayed, less acute disease reflected in delayed infiltration of inflammatory cells. These animals developed chronic disease as a result of steadily increased inflammation, increased levels of urokinase-type plasminogen activator (uPA), and greater degree of demyelination. Thus, the plasminogen activator system can modulate both inflammatory and degenerative events in the CNS through the respective effects of tPA and uPAR on fibrinolysis and cell adhesion/migration, manipulation of which may have therapeutic implications for multiple sclerosis.

Figure 1

Clinical course of EAE disease in tPA^−/− and uPAR^−/− mice. WT, tPA^−/− (A), and uPAR^−/− (B) mice were immunized with 300 μg of MOG 35-55 in complete Freud’s adjuvant on days 0 and 7 and injected with pertussis toxin on days 0 and 2 to induce EAE. All mice were scored for clinical signs of disease on a scale of 1 to 5. Results are plotted as the mean clinical score for all animals in each group; n = 19, WT (three of which failed to develop EAE); n = 14, tPA^−/−; and n = 14, uPAR^−/−. A: tPA^−/− mice incurred a more prolonged and severe disease characterized by an incomplete recovery. B: uPAR^−/− mice showed a delay in the onset and peak of disease.

Figure 2

Infiltration of macrophages/microglia in EAE. Spinal cords were removed at 17 to 20 dpi and at day 35 dpi onwards of EAE and cut longitudinally. Frozen sections were stained with an antibody against F4/80 (A–I), CD45 (J–L), and CD4 (not shown) to assess the infiltration/migration of inflammatory mononuclear cells. In control animals there are no mononuclear cells in the spinal cord tissue (not shown). There is a delay in microglial migration and infiltration of macrophages and lymphocytes in uPAR^−/− mice (C) when compared to WT and tPA^−/− animals at 20 dpi (A and B). F: However, perivascular cuffs containing mononuclear cells are evident in uPAR^−/− mice at the peak of disease, 35 dpi. A high degree of persisting inflammation is evident in uPAR^−/− mice at 60 dpi (I) although not in WT animals (G). J–L: CD45 showed very similar patterns of staining to F4/80. Original magnifications: ×100 (A–L); ×400 (insets).

Figure 3

Perivascular cuff scores for WT, tPA^−/−, and uPAR^−/− EAE mice. Spinal cords from EAE mice at 17 to 20, 35, and 60 dpi were sectioned longitudinally and stained with an antibody against CD45. Total cuffs were counted in a section area of 4 cm², and each cuff was given a score according to the degree of infiltration; 1, perivascular inflammation fewer than three or fewer cells deep; 2, more than three cells deep; 3, parenchymal infiltrate. A total of three slides from different mice were counted per time point and the data are shown as the mean ± SEM; *P < 0.05 and ***P < 0.001, significance illustrated versus control unless illustrated by a bar. Cuff counts (not illustrated) showed a similar pattern as the cuff scores.

Figure 4

Demyelination and axonal pathology in EAE. i: Longitudinal frozen spinal cord sections from mice at 60 dpi were stained for MBP and SMI32 to assess the degree of demyelination and changes in axonal pathology in the CNS during EAE. MBP and SMI32 staining in control untreated mice (A–C, G–I) and in EAE at 60 dpi (D–F, J–L). ii: Density of MBP staining was assessed using a Quantimet image analysis system. Density of MBP staining in uPAR^−/− mice at 60 dpi was significantly less than controls and WT animals at the same stage of disease. iii: The number of SMI32-positive axons was increased in all genotypes at 60 dpi, significantly more so in tPA^−/− and uPAR^−/− mice. Staining for SMI35 revealed no apparent difference between different genotypes and time points (not shown). *P < 0.05, **P < 0.01, ***P < 0.001. Original magnifications, ×100.

Figure 5

Fibrin(ogen) localization in EAE. i: Spinal cords were removed from control animals and mice 35 dpi onwards of EAE and cut longitudinally. Frozen sections were stained with an antibody against fibrin(ogen), and double-fluorescence staining was performed with antibodies against fibrin(ogen) and SMI32, a marker of nonphosphorylated neurofilament. Fibrin(ogen) can be seen surrounding perivascular cuffs in sections from EAE animals (D–F), but in much greater amounts in tPA^−/− mice. Co-localization of fibrin(ogen) on SMI32-positive axons is seen in sections from tPA^−/− EAE mice (H) but not in sections from WT or uPAR^−/− animals (G or I). Original magnifications: ×100 (A–F); ×400 (G–I).

Figure 6

Western blotting of fibrin(ogen) and plasminogen. Spinal cords from control and EAE mice were homogenized for protein extraction. Levels of fibrin(ogen) and plasminogen were detected by Western blotting and were quantitatively measured by densitometry scanning (A) and results are shown as arbitrary densitometry units ± SEM. Blots were reprobed with anti-actin to ensure equal loading of proteins. Levels of fibrin(ogen) (B) and plasminogen (C) were significantly increased during the acute phase of EAE in tPA^−/− mice when compared to control mice and WT mice at the same stage of disease. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 7

Fibrinolysis in the mouse CNS. Spinal cords from control and EAE mice at specific time points were homogenized for protein extraction. The fibrinolytic capacity was investigated using a clot lysis assay that measures the degradation of an in vitro-formed clot using spectrophotometry. Results are presented as the mean clot degradation over time for WT (A), tPA^−/− (B), and uPAR^−/− (C) mice. Samples from tPA^−/− mice were incubated for a further 20 hours to detect fibrinolysis initiated by uPA. tPA^−/− mice at 60 dpi had a significantly faster clot degradation than either control or 20 dpi mice. *P < 0.05.

Figure 8

Levels of PAI-1 and uPA during EAE. Levels of PAI-1 and uPA were determined in spinal cord homogenate samples from control and EAE mice by a modified sandwich ELISA. Results are shown as ng antigen/mg protein ± SEM. A–C: PAI-1 is significantly increased in WT, tPA^−/−, and uPAR^−/− mice during the acute/peak phase of EAE. D–F: In addition levels of uPA were significantly increased during the chronic phase of EAE in tPA^−/− and uPAR^−/− mice (E and F), which was not mirrored in WT animals (D). *P < 0.05 versus control.