Abnormal Spinal Cord Myelination due to Oligodendrocyte Dysfunction in a Model of Huntington’s Disease

Animals

BACHD mice expressing the full-length human mutant HTT transgene containing two loxP sites flanking the exon 1 were used and maintained on the FVB/N strain. NG2-Cre mice were backcrossed onto the FVB/N background and then crossed to BACHD mice to generate BACHDxNG2Cre (BN) mice [11]. Male and female mice were housed with littermates of mixed genotype in groups of 2–5 on a 12 h light/ dark cycle with free access to food and water. All experiments were performed with the approval of the Institutional Animal Care and Use Committee (IACUC no.151067) at the Biomedical Sciences Institute (ASTAR) and in accordance with their approved guidelines.

Transmission electron microscopy

Mice (3 per genotype, all males) were transcardially perfused with 2.5%glutaraldehyde and 2.5%paraformaldehyde (PFA) in phosphate buffered saline (PBS) before post-fixing the brains overnight at 4°C in the same buffer. Cervical spinal cord segments (C1-C3) were microdissected and post-fixed in 1%osmium tetroxide/1.5%potassium ferrocyanide solution for 1 h at room temperature (RT). After washing with 1%uranyl acetate in maleate buffer, the samples were dehydrated, infiltrated with epon, embedded, and polymerized at 60°C for two days. Ultra-thin slices (100 nm) were cut before imaging on a transmission electron microscope. For image analyses, axon and myelin fibres diameters were measured using ImageJ. 12 images with a total of more than 300 axons were analyzed for each mouse. Myelinated axons were divided into two categories and classified as ‘regular shaped axons’ when they presented a circular, quasi-circular or elliptical shape, while remaining axons were classified as ‘irregularly shaped axons’ (dystrophic axons).

Immunohistochemistry and stereological measurements

Cervical spinal cord segments (C1-C3) from 12-month-old animals were microdissected and post-fixed overnight in 4%PFA before switching to 30%sucrose in 1× PBS. Spinal cord segments were sliced coronally into 25μm coronal sections on a cryostat (Leica CM3050S) and were collected free-floating in PBS with 0.01%sodium azide. A series of sections spaced 300μm apart were used for the analysis. Sections were incubated with 1%phenylhydrazine in 1 x phosphate buffered saline (PBS) for 45 min at RT, subsequently washed 3 times with PBS and blocked in 5%normal goat serum (NGS) in PBS containing 0.1%Triton X-100 (TX) for 30 min at RT. Sections were then incubated with primary antibodies in 1%NGS in PBS-TX overnight at 4°C. Rabbit anti-Olig2 at 1:750 (Millipore AB9610), rabbit anti-GSTpi at 1:3000 (MBL, 311), rabbit anti-NG2 at 1:500 (Sigma-Aldrich, AB5320), and rabbit anti-Sox10 at 1:500 (Millipore, AB5727) were used. Sections were washed 3 times with PBS followed by incubation with biotinylated anti-rabbit antibody (1:200; Vectastain ABC HRP Kit, PK-4006; Vector Laboratories, Burlingame, CA, USA) for 1.5 h at room temperature with 1%NGS, 0.2%Triton X-100 in PBS. After three washes in PBS, sections were incubated in Vectastain Elite ABC reagent (Vector Labs Inc., Burlingame, CA, USA) for 2 h at RT and staining was visualized using 3,3^′-diamino- benzidine (DAB) (ImmPACT DAB Peroxidase Substrate, SK-4105; Vector Laboratories, Burlingame, CA, USA). Slices were mounted on slides and images of whole spinal cord cross-sections were captured using a Nikon Ni-E upright microscope. The GM and WM area of the spinal cord was measured by tracing the perimeter of the respective areas using ImageJ and the total number of cells was counted. For GM and WM area, Olig2 staining was used and half of the spinal cord was measured to estimate total area and total cell counts.

Spinal cord tissue collection for molecular analysis

Cervical spinal cord segments (C4-C6) from 12-month-old animals were microdissected, snap-frozen in liquid nitrogen and stored at –80°C until further use.

Protein analysis

Protein lysate of cervical spinal cord segment (C4-C6) from male and female mice were prepared using RIPA buffer (Sigma-Aldrich) with 1 mM PMSF (Sigma-Aldrich), 5μm Z-VAD (Promega), 1 mM NaVan (Sigma-Aldrich), and 1× Complete Protease Inhibitor Cocktail tablets (Roche). For immunoblotting 30μg of protein lysates were separated on 12%Bis-Tris protein gel (Novex) with 20X MES running buffer (Novex) and transferred on nitrocellulose membranes. Samples were successively blocked with Odyssey Blocking Buffer (LI-COR, P/N 927-40000) 1 h at RT, and then incubated with primary antibodies at 4°C overnight. The following primary antibodies were used: Ermin (1:1000; Merck), MBP (1:500; Millipore), MAG (1:2000; Millipore), Sept8 (1:500; Proteintech), β-actin (1:5000; Sigma-Aldrich) and Calnexin (1:5000; Sigma-Aldrich). Membranes were then incubated with secondary antibodies (1:10,000) 2 h at RT: Alexa-Fluor goat anti-rabbit 800, Alexa-Fluor goat anti-mouse 680, Alexa-Fluor goat anti-rabbit 680, and Alexa-Fluor goat anti-rat 800 (all from Life Technologies). The membrane was imaged using the LiCor Imaging System and Odyssey V3.0 software (LiCor), followed by intensity analysis with ImageJ.

White, but not grey, matter atrophy in the spinal cord is rescued by inactivation of mutant HTT in oligodendroglia

To determine whether GM and WM regions are atrophied in the spinal cord of 12-month-old BACHD mice, we measured the area of GM and WM in the cervical (C1-C6) regions of the spinal cord by histological assessment using the oligodendroglia marker Olig2 (Fig. 1A,C). The spinal total area, GM, and WM area were significantly lower in BACHD mice compared to WT controls (Fig. 1D-F), demonstrating atrophy of GM and WM in spinal cord of HD mice.

OPEN IN VIEWER

Fig. 1

OPC-intrinsic effects of mHTT on white, but not grey, matter spinal cord in HD mice. A) Spinal cord diagram. B) Schematic representation of Cre-mediated genetic reduction of mHTT expression in OPCs (NG2 + cells) in BACHD mice. C-F) Spinal grey matter (GM), white matter (WM), and total area were evaluated measuring half side of the spinal cord using Olig2 staining in 12-month-old mice. C) Images of cervical spinal cord sections at 12 months of age. WT image shows how area was measured. GM area is depicted in grey, total area is depicted in red, and WM area was calculated subtracting GM area from total area. Scale bar represents 200μm. D-F) Spinal total area is decreased in BACHD and BN mice compare to WT (D), but it also shows a significant increase in BN mice compared to BACHD (D). BACHD mice present significant decrease in (E) GM and (F) WM compared with WT mice. Spinal WM, but not GM, area is rescued in BN mice (E,F). N = 6–13 per genotype. Data represent means±SEM; ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001 by one-way ANOVA followed by Tukey’s test.

We have previously shown that callosal WM abnormalities in the BACHD are caused by intrinsic effects of mHTT in oligodendroglia [11]. To examine whether the spinal WM atrophy observed in BACHD is similarly caused by intrinsic mHTT-induced oligodendroglial defects, we used the previously described BACHDxNG2Cre (BN) mouse line where mHTT is specifically reduced in oligodendrocyte progenitor cells (OPC) (Fig. 1B) [11]. The total spinal area in BN mice was significantly lower than WT mice, but also significantly higher than BACHD mice, indicating partial rescue (Fig. 1D). To further explore the partial rescue in total spinal area, assessment of spinal GM and WM areas in BN mice was performed. We showed that white, but not grey, matter atrophy in the spinal cord of BACHD mice was rescued in BN mice (Fig. 1E,F). Consistent with this finding, we observed significant atrophy of the spinal GM region in BN mice compared to WT mice (Fig. 1E), but no differences in spinal WM area (Fig. 1F). Together these findings demonstrate that WM abnormalities in HD extend to the spinal cord and are primarily driven by cell-intrinsic defects of mHTT in oligodendroglia.

Oligodendroglia-intrinsic effects of mHTT cause myelin deficits in the spinal cord of BACHD mice

To test whether the myelin deficits observed in the corpus callosum also extended to the spinal cord, we used electron microscopy to visualize myelinated fibers in the cervical area of the spinal cord at 12 months of age (Fig. 2A) and analyzed the g-ratio. G-ratio of myelinated axons, used as a measure of myelin thickness, was calculated as the ratio of axon diameter to myelinated fiber diameter (Fig. 2B). The mean g-ratio of myelinated axons was significantly higher in BACHD mice compared to WT mice (Fig. 2C,E,F), indicating that their myelin sheaths were thinner. Consistent with the increased spinal WM area observed in BN mice, we found thicker myelin sheaths in BN mice as shown by a decrease in g-ratio. These results indicate that selective reduction of mHTT in oligodendroglia rescues the myelin thickness deficits not only in brain but also spinal cord WM (Fig. 2D,E,F). To exclude that the changes observed in the g-ratio are due to differences in axonal diameters, we analyzed axonal diameter frequencies and found that distribution of axonal diameters was similar between genotypes (Supplementary Figure 1A,B), indicating that differences in g-ratio between BACHD and BN mice are indicative of changes in myelin thickness and not the result of changes in axonal caliber. We also analyzed the number of dystrophic axons and did not find significant differences among the genotypes (Supplementary Figure 1C).

OPEN IN VIEWER

Fig. 2

OPC-intrinsic effects of mHTT cause myelination abnormalities in the spinal cord of HD mice. A) EM images of myelinated axons in cervical spinal cord sections at 12 months of age. Scale bar represents 2μm. B) G-ratio calculation diagram. C-F) Higher g-ratios (thinner myelin sheaths) in BACHD mice are rescued in BN mice. N = 3 per genotype; ∼300 axons quantified per animal. Data represent means±SEM; ^**p < 0.01 by one-way ANOVA followed by Tukey’s test.

Inactivation of mutant HTT in oligodendroglia rescues deficits in myelin-related proteins in the spinal cord

In order to evaluate whether the rescue of WM atrophy and myelin thickness in the spinal cord seen with inactivation of mHTT in oligodendroglia in BN mice is associated with changes in myelin protein levels, we next examined the expression levels of the myelin-related proteins Ermin, myelin basic protein (MBP), myelin-associated glycoprotein (MAG), and Sept8 by immunoblotting. We found that Ermin, MBP, and MAG were significantly increased in BN versus BACHD mice (Fig. 3A-C), whereas Sept8 did not show any significant changes (Fig. 3D). This increase in the levels of myelin-related proteins is consistent with the rescue in myelin thickness observed in BN mice.

OPEN IN VIEWER

Fig. 3

Myelin protein expression in BACHD is mediated by OPC-intrinsic effects of mHTT. A-D) BN mice present significant increase in (A) Ermin, (B) MBP, (C) MAG, but not (D) Septin-8 protein expression levels compared to BACHD mice in the spinal cord at 12 months of age. N = 10–11/genotype in A,B or N = 8–9 in C or N = 13 in D. All bands were used for quantification. Data represent means±SEM; ^*p < 0.05; two-tailed unpaired Student’s test in B (p = 0.0273, t = 2.391, d.f. = 19), or one-tailed unpaired Student’s test in A (p = 0.0354, t = 1.909, d.f. = 20), C (p = 0.0462, t = 1.798, d.f. = 15), and D.

No effect on oligodendroglial populations in BACHD or BN mice

To investigate whether the spinal WM atrophy and myelination abnormalities in BACHD mice and the rescue observed in BN mice are associated with changes in oligodendroglia populations in the spinal cord, we analyzed the density of relevant cellular populations by stereology assessment using different oligodendroglia lineage markers, namely Olig2 (a transcription factor expressed throughout the entire oligodendrocyte lineage), GST-pi (a mature myelinating oligodendrocyte marker), and NG2 (an OPC marker). Consistent with our previous observations in the brain, no changes in the density of these oligodendroglia populations were observed in the spinal cord of BACHD and BN mice at 12 months of age (Supplementary Figure 2).