Pathophysiology of white matter perfusion in Alzheimer's disease and vascular dementia

Abstract

Little is known about the contributors and physiological responses to white matter hypoperfusion in the human brain. We previously showed the ratio of myelin-associated glycoprotein to proteolipid protein 1 in post-mortem human brain tissue correlates with the degree of ante-mortem ischaemia. In age-matched post-mortem cohorts of Alzheimer's disease (n = 49), vascular dementia (n = 17) and control brains (n = 33) from the South West Dementia Brain Bank (Bristol), we have now examined the relationship between the ratio of myelin-associated glycoprotein to proteolipid protein 1 and several other proteins involved in regulating white matter vascularity and blood flow. Across the three cohorts, white matter perfusion, indicated by the ratio of myelin-associated glycoprotein to proteolipid protein 1, correlated positively with the concentration of the vasoconstrictor, endothelin 1 (P = 0.0005), and negatively with the concentration of the pro-angiogenic protein, vascular endothelial growth factor (P = 0.0015). The activity of angiotensin-converting enzyme, which catalyses production of the vasoconstrictor angiotensin II was not altered. In samples of frontal white matter from an independent (Oxford, UK) cohort of post-mortem brains (n = 74), we confirmed the significant correlations between the ratio of myelin-associated glycoprotein to proteolipid protein 1 and both endothelin 1 and vascular endothelial growth factor. We also assessed microvessel density in the Bristol (UK) samples, by measurement of factor VIII-related antigen, which we showed to correlate with immunohistochemical measurements of vessel density, and found factor VIII-related antigen levels to correlate with the level of vascular endothelial growth factor (P = 0.0487), suggesting that upregulation of vascular endothelial growth factor tends to increase vessel density in the white matter. We propose that downregulation of endothelin 1 and upregulation of vascular endothelial growth factor in the context of reduced ratio of myelin-associated glycoprotein to proteolipid protein 1 are likely to be protective physiological responses to reduced white matter perfusion. Further analysis of the Bristol cohort showed that endothelin 1 was reduced in the white matter in Alzheimer's disease (P < 0.05) compared with control subjects, but not in vascular dementia, in which endothelin 1 tended to be elevated, perhaps reflecting abnormal regulation of white matter perfusion in vascular dementia. Our findings demonstrate the potential of post-mortem measurement of myelin proteins and mediators of vascular function, to assess physiological and pathological processes involved in the regulation of cerebral perfusion in Alzheimer's disease and vascular dementia.

Keywords: Alzheimer’s disease; brain ischaemia; cerebral blood flow; dementia; neuropathology.

Figure 1

Scatterplot showing the relationship between ET1 concentration (A) and ACE activity (B) and the MAG:PLP1 ratio in the Bristol cohort. ET1 correlated positively with the MAG:PLP1 ratio (r = 0.465, P = 0.0005). There was no significant correlation between ACE activity and MAG:PLP.

Figure 2

Bar charts indicating the mean levels of ET1 (A), ACE activity (B), VEGF (C) and FVIIIRA (D) in parietal deep white matter from the Bristol cohort, in control, vascular dementia (VaD) and Alzheimer’s disease (AD) brains. ET1 concentration differed significantly across the three groups (P = 0.0008). Dunn’s test showed significant reductions in Alzheimer’s disease compared to both controls (P < 0.05) and vascular dementia (P < 0.001). ACE activity, VEGF and FVIIIRA were not significantly altered in vascular dementia or Alzheimer’s disease compared to controls. Error bars show the SEM. *P < 0.05, ***P < 0.001.

Figure 3

Scatterplots showing the relationship between VEGF concentration and the MAG:PLP1 ratio in the Bristol (A) and Oxford (B) cohorts. VEGF was correlated negatively with the MAG:PLP1 ratio in both cohorts (Bristol cohort: r = −0.315, P = 0.0015; Oxford cohort: r = −0.631, P < 0.0001).

Figure 4

Scatterplots showing the relationships between FVIIIRA and VEGF concentration (A and B) and MAG:PLP1 ratio (C and D) in the Bristol (A, C) and Oxford (B, D) cohorts. FVIIIRA correlated positively with VEGF concentration in both the Bristol (r = 0.205, P = 0.042) and Oxford (r = 0.253, P = 0.0267) cohorts. FVIIIRA tended to decline slightly with MAG:PLP, but not significantly in either cohort.

Figure 5

Scatterplots showing the relationships between vessel density, as measured by determining the area fraction immunopositive for FVIIIRA in histological sections of right parietal white matter, and both (A) FVIIIRA level and (B) VEGF concentration in the left parietal white matter. The vessel density correlated strongly with FVIIIRA concentration (r = 0.580, P < 0.0001) and tended to increase with VEGF concentration, although not significantly so. VIIIRA labelled microvessels in the white matter. Scale bar = 100 μm.