Oxidative Glial Cell Damage Associated with White Matter Lesions in the Aging Human Brain

Abstract

White matter lesions (WML) are common in brain aging and are associated with dementia. We aimed to investigate whether oxidative DNA damage and occur in WML and in apparently normal white matter in cases with lesions. Tissue from WML and control white matter from brains with lesions (controls lesional) and without lesions (controls non-lesional) were obtained, using post-mortem magnetic resonance imaging-guided sampling, from the Medical Research Council Cognitive Function and Ageing Study. Oxidative damage was assessed by immunohistochemistry to 8-hydroxy-2'-deoxoguanosine (8-OHdG) and Western blotting for malondialdehyde. DNA response was assessed by phosphorylated histone H2AX (γH2AX), p53, senescence markers and by quantitative Reverse transcription polymerase chain reaction (RT-PCR) panel for candidate DNA damage-associated genes. 8-OHdG was expressed in glia and endothelium, with increased expression in both WML and controls lesional compared with controls non-lesional (P < 0.001). γH2Ax showed a similar, although attenuated difference among groups (P = 0.03). Expression of senescence-associated β-galactosidase and p16 suggested induction of senescence mechanisms in glia. Oxidative DNA damage and a DNA damage response are features of WML pathogenesis and suggest candidate mechanisms for glial dysfunction. Their expression in apparently normal white matter in cases with WML suggests that white matter dysfunction is not restricted to lesions. The role of this field-effect lesion pathogenesis and cognitive impairment are areas to be defined.

Keywords: DNA damage; dementia; ischemic white matter; white matter disease; white matter lesions.

Figure 1

Immunohistochemistry to 8‐hydroxy‐2′‐deoxoguanosine showing strongly labeled nuclei (red arrows), unlabelled nuclei (blue arrows) and nuclei with faint reactivity (yellow arrow). Scale bar 50 μm.

Figure 2

Double‐labeling studies for cell specific markers (red fluorescent label). A. GFAP; D. oligodendrocyte‐specific protein; G. CD68. 8‐hydroxy‐2′‐deoxoguanosine detected with 3,3'‐diaminobenzidine (DAB) label visualized under light microscopy (B, E, H). Merged images shown in C, F and I. Arrows identify example double‐labeled cells. Scale bar 50 μm.

Figure 3

Box plots showing variation in 8‐hydroxy‐2′‐deoxoguanosine counts (A), malondialdehyde (MDA) protein by western blot (B) and γH2Ax counts among the three groups. D. Example Western blot for MDA showing expected size band at 64 kDa.

Figure 4

Immunohistochemistry showing nuclear expression of γH2Ax (A) and DNA protein kinase (B). Scale bars 50 μm. C and D. Western blots detecting bands of the expected sizes.

Figure 5

Expression of senescence markers. A. β‐gal expression in cells (arrows). B. β‐gal (blue) colocalized with oligodendrocyte‐specific protein (brown) indicating expression in oligodendrocytes (arrows). C. Colocalization of β‐gal with the astrocyte marker GFAP (arrows). D. β‐gal and CD68 double‐staining. β‐gal reactivity (arrows) does not colocalize with CD68 (open arrows). E. Colocalization of β‐gal with p16 (arrow).

Figure 6

Scatterplot showing the association between p16 cell count and area immunoreactivity for GFAP.

Figure 7

A. Variation in gene expression between control non‐lesional, control lesional (CL) and deep subcortical (here labeled DSCL) from RTqPCR array, Error bars represent standard deviations B. Expression of p53 (scale bar 50 μm). C. Boxplot showing variation in percentage of nuclei positive for p53 among the three groups, with higher values in the CL group.