Loss of Myelin Basic Protein Function Triggers Myelin Breakdown in Models of Demyelinating Diseases

Abstract

Breakdown of myelin sheaths is a pathological hallmark of several autoimmune diseases of the nervous system. We employed autoantibody-mediated animal models of demyelinating diseases, including a rat model of neuromyelitis optica (NMO), to target myelin and found that myelin lamellae are broken down into vesicular structures at the innermost region of the myelin sheath. We demonstrated that myelin basic proteins (MBP), which form a polymer in between the myelin membrane layers, are targeted in these models. Elevation of intracellular Ca(2+) levels resulted in MBP network disassembly and myelin vesiculation. We propose that the aberrant phase transition of MBP molecules from their cohesive to soluble and non-adhesive state is a mechanism triggering myelin breakdown in NMO and possibly in other demyelinating diseases.

Graphical abstract

Figure 1

Myelin Pathology in Focal Experimental NMO Lesions Starts with Vesiculation of Myelin at the Inner Tongue (A) Representative images of cross-sections of corpus callosum of Lewis rats injected with human AQP4 antibody (Ab) and complement or PBS (Ctrl) at indicated time points (scale bar, 500 nm). (B) Myelin fragmentation profiles with representative images. (C) Quantification of the myelin breakdown patterns at the different time points post-injection (1 hr, 18 hr, 5 days, and 7 days post-injection [p.i.]). Bars represent mean with SEM (n = 3 animals, >300 axons per animal; ^∗∗p < 0.01, ^∗∗∗p < 0.001, one-way ANOVA). See also Figures S1 and S2.

Figure 2

Myelin Fragmentation Patterns in Anti-MOG Antibody-Injected Lewis Rats and Biozzi EAE (A) Representative images of cross-sections of MOG antibody (Ab) and complement (right) or PBS (Ctrl) injections (left) into the corpus callosum of adult Lewis rats at 12 hr p.i. Scale bar in all images, 500 nm. (B) Pattern of myelin breakdown detected after focal injection of MOG antibody and complement. (C) Quantification of the different myelin fragmentation profiles depicted in (B) as percentage of total myelin sheaths at 1 hr, 8 hr or 12 hr p.i. of MOG antibody/complement or PBS (Ctrl). (D) Representative images of cross-sections of Biozzi EAE spinal cord lesions (first relapse). The left panel displays control area, whereas the right panel displays border of demyelinated EAE lesion. (E) Myelin breakdown profiles detected in EAE lesions. (F) Quantification of percentage of all different fragmentation patterns in EAE lesion and control area. Bars display mean with SEM (n = 3 animals, >300 axons per animal; ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0,001, one-way ANOVA). (G) Schematic drawing of proposed model of myelin fragmentation. A normal myelin sheath with its compacted layers (in green) wrapped around an axon (with microtubules in blue and neurofilaments in pink) is depicted. The cytoplasm-rich domains of myelin at the outer and inner tongue are shown in orange. In the focal NMO model, first the inner tongue becomes enlarged before it collapses into small vesicles at the inside of the myelin sheath. The vesiculation progresses outward until the whole myelin sheath is transformed into vesicular profiles. The lower panel shows additional patterns observed in the anti-MOG antibody model consisting of focal splitting and vesiculations with focal bulging. See also Figures S1 and S2.

Figure 3

Lack of MBP Leads to Myelin Vesiculation (A) Adult Lewis rats were injected with human AQP4 antibody (Ab) and complement or PBS (Ctrl) in the corpus callosum. The subcellular localization of MBP (top) and PLP (bottom) was determined by cryoimmuno-electron microscopy at 18 hr p.i. (scale bar, 500 nm; gold size, 15 nm for MBP and 10 nm for PLP). The vesicles at the inner tongue are enlarged in the white boxes. (B) Quantification of the number of gold particles in compact myelin as compared to gold particles in vesiculated myelin for MBP and PLP labeling. Bars shown mean with SEM (n = 3 animals, >70 axons per animal; ^∗∗∗p < 0.001, Student’s t test). (C and D) Representative images of high-pressure frozen optic nerves of wild-type and shiverer mice at P10, P14, and P21 (C) with magnification of a vesiculated sheath in (D). (E) Quantification of vesiculated membrane profiles. Bars shown mean with SEM (n = 3 animals, >80 axons per animal; ^∗∗∗p < 0.001, Student’s t test). See also Figure S1.

Figure 4

Calcium Influx Triggers Epitope Unmasking in MBP in Early NMO Lesions (A) Representative images of coronal sections of adult Lewis rat brains injected with human AQP4 antibody (Ab) and complement or PBS (Ctrl) in the contralateral hemisphere for 1 hr or 18 hr p.i. stained for MBP in green and QD9 in red. (B) Magnification of the red box in (A) and no primary antibody control to indicate that there is no cross-reactivity with the antibody injected. Scale bar, 1 mm (A and B). (C) Quantification of the MBP to QD9 signal intensity ratio of the AQP4 antibody lesion normalized to the control (PBS injection) (n = 3 animals, 3–5 regions of same size per animal, ^∗∗∗p < 0.001, Student’s t-test). (D) Adult Lewis rat brains were injected with 1 μl human AQP4 antibody or PBS and complement. After 12 hr p.i., the rats were sacrificed, and acute brain slices were maintained for 2 hr in aCSF or aCSF supplemented with 25 mM EGTA. Sections were stained with DAPI in blue (upper panel), QD9 in yellow (middle panel), and MBP in grey (lower panel). Scale bar, 1 mm. The lesion area marked by the red line was determined by GFAP staining. (E) Representative images of a GFAP staining (in green) in control and AQP4 Ab injected rats. The loss of staining or astrocyte fragmentation is marked by the red line. Scale bar, 200 μm. (F) Quantification of GFAP signal intensity around the injection site and lesion size. (G) Quantification of lesion area based on GFAP loss in (E). (H) Quantification of MBP to QD9 integrated density ratio as normalized to the 12 hr p.i. time point for control and EGTA treated acute brain sections. Graphs show the mean with SEM (n = 3 animals, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, one-way ANOVA). See also Figures S1, S3, and S4.