Supplementary Figure 1. Reduced Weight, and Age-Related Increase in GFAP Immunostaining

Supplementary Figure 1. Reduced weight, and age-related increase in GFAP immunostaining in Ercc1Δ/- brain

a, b) Brain weights in grams (a) and percentage of body weight (b) of wt and Ercc1Δ/-mice at 4, 8 and 16 weeks. Brains of Ercc1Δ/- mice show reduced weight proportional to their reduced body weight.

c-h) Parasagittal brain sections of wt and Ercc1Δ/- mice stained for Calbindin (c, d) or GFAP (e-h). Calbindin staining that outlines subpopulations of neurons is the same in the brain of 4 week old wild-type and Ercc1Δ/-mice (c, d). GFAP-immunohistochemistry revealed small areas of increased GFAP staining in the brain of 4 week old Ercc1Δ/-mice. In contrast, a large increase in GFAP staining indicative of abundant astrocytosis occurs in 4 week old Ercc1Δ/-mice. e'-h' are high magnifications of area indicated in brainstem of e-h, respectively.

Bars: 2 mm (g), 100 µm (e')

Supplementary Figure 2. Age related increase in GFAP-labeling and neuronal degeneration in Ercc1Δ/- cerebellar cortex

a-f) Low- (a-c) and high-magnification (d-f) double labeling confocal immunofluorescence images of calbindin and GFAP in cerebellar cortex of 4, 8 and 16 week old Ercc1Δ/- mice showing areas of reduced calbindin staining, indicative of Purkinje cell loss in the molecular (ml) and Purkinje cell layer of 16 week old Ercc1Δ/- mice (arrow head in c, f). In parallel clusters of intense GFAP immunostaining appear in the molecular layer (arrows in b and c) in the molecular layer. Increased GFAP-immunoreactivity also occurs in the granule cell layer (gl) of 16 week old Ercc1Δ/- mice cerebellar cortex. GFAP and calbindin staining in 4 weeks Ercc1Δ/- mice is the same as in wild-type. Asterisks in d indicate the cell bodies of Purkinje cells.

Bar: 100 µm

Supplementary Figure 3. Activated microglia cells surrounding motor neurons in Ercc1Δ/- spinal cord

Photomicrographs of microglia cells stained for the complement 3 receptor (CR3) in motor columns in transverse lumbar L4 sections of 16 weeks old wt (a) and Ercc1Δ/- (b) mice. Sections are counterstained for thionin to outline motor neuron cell bodies (arrow head in b). Arrow in b points to a cluster of activated microglia cells. Scale bars: 50 µm.

Supplementary Figure 4. Neuronal and non-neuronal cell death in Ercc1Δ/-spinal cord

a-d) Low- (a and b) and high-magnification (b', c, d) photomicrographs of active caspase 3 staining in transverse Ercc1Δ/-spinal cord sections showing a labeled motor neuron (b, b') and labeled apoptotic cells in the grey (c) and white (d) matter. Arrows in b point to irregular labeled structures, putatively reflecting remnants of dead cells.

e-g) High-magnification photomicrographs of TUNEL-positive cells in the grey (e) and white (f) matter of Ercc1Δ/- spinal cord. Values in bar graph in g represent means ± SE (n=5 mice/bar) of TUNEL-positive cells in 4 consecutive paraffin sections expressed as % of labeled cells in wild-type littermates of the same age. The mean absolute number of TUNEL-positive cells were 2.9, 1.2 and 1.3 cells/mm2 for 4, 8 and 16 week old wild-type mice, respectively. *, P < 0.05 compared to age -matched wild-type (Students t-test).

Scale bars: 50 µm (a), 10 µm (c, f)

Supplementary Figure 5. A small subset of ATF3 positive motor neurons in Ercc1Δ/- spinal cord is surrounded by phagocytosing microglia

Double-labeling confocal immunofluorescence showing an ATF3-positive motor neuron surrounded by Mac2-positive microglia cells (arrows). Scale bar: 20 µm