Autoimmune-induced preferential depletion of myelin-associated glycoprotein (MAG) is genetically regulated in relapsing EAE (B6 x SJL) F1 mice

Abstract

Background: Experimental autoimmune encephalomyelitis (EAE) is commonly used to investigate mechanisms of autoimmune-mediated damage to oligodendrocytes, myelin, and axons in multiple sclerosis (MS). Four distinct autoimmune mechanisms with subsequently distinct patterns of demyelination have been recognized in acute MS lesions. EAE correlates for those distinct patterns of MS lesions are unknown. An excessive loss of myelin-associated glycoprotein (MAG), as a result of distal oligodendrogliopathy, is found exclusively in the subtype III lesion. We sought to answer if types of demyelination in acute lesions during onset and relapse of EAE can replicate the specific patterns observed in MS acute lesions.

Methods: In parental H-2b (C57BL/6, B6) and hybrid H-2b/s [(B6 x SJL) F1] EAE mice, we examined spinal cord levels of MOG, MAG, and myelin basic protein (MBP), and compared to levels of axonal neurofilament (NF160) to assess axonal function, and levels of PARPp85 as an indicator of irreversible apoptosis.

Results: During disease onset, levels of MOG significantly dropped in both strains, although more profoundly in H-2b/s mice. Levels of MOG recovered in relapsing mice of both strains. Regulation of MAG was dissimilar to MOG. Modest loss of MAG was found at disease onset in both strains of mice. Unexpectedly, in relapsing H-2b/s mice, a major depletion of MAG and NF160, accompanied with sharp elevation of PARPp85 levels, was measured. PARPp85 immunoreactivity was observed in cytoplasm and nuclei of some MBP containing cells.

Conclusion: Taken together, our results show genetically controlled distinct patterns of MOG and MAG depletion, in MOG35-55 induced EAE in H-2b and H-2b/s mice. The data also suggest distinctive immune regulation of acute lesions that develop in relapsing compared to disease onset. A profound depletion of MAG, concomitant with marked depletion of axonal NF160, and sharp elevation of PARPp85 levels, occurred exclusively in relapsing H-2b/s mice. Our findings suggest concurrence of sharp decrease of MAG levels, axonal dysfunction and irreversible apoptosis with severe relapsing disease in H-2b/s mice. We propose that MOG-induced EAE in H-2b/s mice may prove as a useful model in studying mechanisms, which govern autoimmune-induced preferential loss of MAG, and its impact on oligodendroglial pathology.

Figure 1

Profound depletion of MAG levels occurs in relapsing H-2^b/s mice. Dissimilar regulation of relative levels of MAG was observed between H-2^b and H-2^b/s mice throughout EAE. Levels of MAG in spinal cord of EAE mice were measured by western blot. Approximately 20 μg of total protein was loaded per lane. Total of 3–4 mice from H-2^b/s and H-2^b strains with acute and relapsing stages of the disease were analyzed. Only H-2^b/s mice (3 mice) with chronic-sustained disease were examined. Representative blot of 3–4 similar experiments shows: acute – H-2^b/s (13 dpi, grade 4.5), H-2^b (16 dpi, grade 4); relapse – H-2^b/s (21 dpi, grade 3.5, 1^st relapse), H-2^b (27 dpi, grade 2, 1^st relapse); and chronic-sustained – H-2^b/s (72 dpi, grade 2.5). Relative levels of MAG were calculated as ration to corresponding levels of GAPDH (see Fig. 4). Compared to levels in control and in H-2^b mice, H-2^b/s mice markedly reduced their relative levels of MAG during acute and chronic. Most profound depletion of MAG was found in relapsing H-2^b/s mice. Corresponding levels of GAPDH are shown in Fig. 6. Membrane was stripped and re-probed with MOG, MBP, NF160 and PARPp85 antibodies (Fig. 2, 4, 6 and 7A).

Figure 2

Distinct patterns of MOG reduction between H-2^b/s and H-2^b mice throughout disease. Distinct patterns of MOG-specific demyelination were observed between H-2^b and H-2^b/s mice throughout EAE. Relative levels of MOG were measured similarly as levels of MAG (Fig.1). As opposed to regulation of MAG, most profound depletion of MOG was observed in H-2^b/s mice during acute EAE. Lesser then in H-2^b/s, loss of MOG was noted in acute H-2^b mice during acute disease. Loss of MOG, lesser then in acute but significant compared to control values, was equally found in H-2^b and H-2^b/s mice with relapsing disease. In H-2^b/s mice with chronic-sustained disease levels of MOG were comparable to control values. Stripped and reprobed membrane was shown. Corresponding levels of GAPDH are shown in Fig. 6.

Figure 3

H&E staining reveals an extensive mononuclear cell infiltration in spinal cord of (B6 × SJL) F1 (H-2^b/s) relapsing mice (top panel). In these mice inflammatory cells are evident throughout the spinal cord parenchyma. Lesser degree of inflammation was observed in relapsing B6 (H-2^b) mice (lower panel). Representative image shows lumbar spinal cord from relapsing H-2^b/s mouse, with clinical severity score of 3, and relapsing H-2^b mouse with severity score of 2. (H&E × 10).

Figure 4

Discrete changes in MBP levels in CNS of H-2^b/s and H-2^b mice throughout EAE. Discrete changes in MBP levels were observed in spinal cord of H-2^b and H-2^b/s mice throughout EAE. Relative levels of MBP were measured similarly as levels of MAG (Fig. 1). Compared to control values, modest reduction of MBP levels was measured only in H-2^b/s mice during relapsing and chronic-sustained EAE. This antibody revealed presence of four distinct closely related 21–26 KD isoforms of MBP in control and EAE spinal cord, similarly as reported by the manufacturer of this antibody. These isoforms may be the result of posttranslational modification of MBP. Stripped and re-probed membrane was shown. Corresponding levels of GAPDH are shown in Fig. 6.

Figure 5

Correlation between levels of MAG, MOG and MBP inH-2^b/s and H-2^b mice has revealed dissimilar regulation of MAG and MOG in relapsing H-2^b/s mice.

Figure 6

Distinctive changes of NF160 levels in CNS throughout disease between H-2^b/s and H-2^b mice. Relative levels of medium chain of neurofilament (NF160) were significantly depleted in mice with relapsing and chronic disease. Relative levels of NF160 were measured similarly as levels of MAG (Fig.1). Most profound drop in NF160 levels was observed in relapsing H-2^b/s mice.

Figure 7

Differences in levels of PARPp85 between H-2^b/s and H-2^b EAE mice. Relative levels of PARPp85 were measured in spinal cord (A) and in Percol gradient separated fraction of spinal cord, which was devoid of infiltrating inflammatory cells (B). Relative levels of PARPp85 in spinal cord were determined similarly as levels of MAG (Fig. 1). The highest level of PARPp85 was found in relapsing H-2^b/s mice, and this value was arbitrarily taken as 100. Compared to this highest level, significantly lower relative levels of PARPp85 were observed during acute and chronic disease. Stripped and re-probed membrane was shown for levels of PARPp85 in spinal cord. Corresponding levels of GAPDH are shown in Fig. 6.

Figure 8

Two-color immunostaining reveals co-localization of PARPp85 and some MBP immunoreactivity (merge and merge detail). Some PARPp85 is found in cell nuclei, whose morphology is suggestive of apoptosis (asterix). Other nuclei with PARPp85 immunoreactivity appear round and small (arrowhead). The remaining PARPp85 immunostaining appears extranuclear and relates to cells expressing MBP immunoreactivity (arrows). Note relative absence of PARPp85 in submeningeal mononuclear cells (merge – upper right). (Two-color immunofluorescence × 20 and 40).

Figure 9

Correlation between MAG, NF160 and PARPp85 levels in H-2^b/s and H-2^b mice has revealed similar regulation of MAG and NF160 in relapsing H-2^b/s mice. Dissimilar to profound drop of MAG and NF160, sharp elevation of PARPp85 was observed in these mice.