Amyloid Plaque-Associated Oxidative Degradation of Uniformly Radiolabeled Arachidonic Acid

Abstract

Oxidative stress is a frequently observed feature of Alzheimer's disease, but its pathological significance is not understood. To explore the relationship between oxidative stress and amyloid plaques, uniformly radiolabeled arachidonate was introduced into transgenic mouse models of Alzheimer's disease via intracerebroventricular injection. Uniform labeling with carbon-14 is used here for the first time, and made possible meaningful quantification of arachidonate oxidative degradation products. The injected arachidonate entered a fatty acid pool that was subject to oxidative degradation in both transgenic and wild-type animals. However, the extent of its degradation was markedly greater in the hippocampus of transgenic animals where amyloid plaques were abundant. In human Alzheimer's brain, plaque-associated proteins were post-translationally modified by hydroxynonenal, a well-known oxidative degradation product of arachidonate. These results suggest that several recurring themes in Alzheimer's pathogenesis, amyloid β proteins, transition metal ions, oxidative stress, and apolipoprotein isoforms, may be involved in a common mechanism that has the potential to explain both neuronal loss and fibril formation in this disease.

Keywords: Alzheimer’s disease; Oxidative stress; polyunsaturated fatty acids.

Figure 1

Fate of [1-¹⁴C]-ARA in WT mouse brain. (A) Specific radioactivity (DPM/gram tissue wet mass) of cerebellum and hippocampus, divided by the whole brain specific radioactivity, to yield “relative specific radioactivity”. (B) Percentage of total brain radioactivity that partitioned to the middle and upper (hydrophilic) phases of a Bligh-Dyer extract. There were no significant differences among these results. (C) The percentage radioactivity that eluted with the retention time characteristic of ARA. Error bars indicate standard deviations, and white numerals on the bars indicate the number of days after ICV injection. *Indicates a difference with 0.01 < P < 0.05. **Indicates a difference with P < 0.01. P values were determined by t test.

Figure 2

Fate of [U-¹⁴C]-ARA in WT and J20 TG mouse brain. (A) The relative specific radioactivity in the cerebellum and hippocampus of WT (n = 16) and TG (n = 20). There were no significant differences. (B) Percentage of total brain radioactivity that partitioned to the middle and upper (hydrophilic) phases of a Bligh–Dyer extract. There were no significant differences. (C) Activity in 2 min HPLC fractions from the stock [U-¹⁴C]-ARA solution used for ICV injection, subjected to extraction and saponification as for tissue samples. Arbitrary vertical scale. Intervals mark the six pooled fractions into which the 2 min fractions were groups for analysis. Characteristic compounds eluting into these fractions are listed in Table 1. (D) Representative activity in 2 min HPLC fractions from the hippocampus of a TG mouse. (E) Total ion chromatogram for MRM transitions of known oxidized ARA derivatives. (F) Averaged percentages of total recovered activity in the six pooled fractions for whole brain (n = 6WT/7TG), cerebellum (n = 9WT/8TG), and hippocampus (n = 9TG/8WT).

Figure 3

Fate of [U-¹⁴C]-ARA and [1-¹⁴C]-ARA in WT and 5xFAD TG mouse brain. (A) Relative specific radioactivity in the cerebellum and hippocampus of WT (n = 10) and TG (n = 7) mice injected with [U-¹⁴C]-ARA, and WT (n = 4) and TG (n = 4) injected with [1-¹⁴C]-ARA. (B) Percentage of total brain radioactivity in WT and 5xFAD mouse brain that partitioned to the middle and upper (hydrophilic) phases of a Bligh-Dyer extract. There were no significant differences. (C) Activity in pooled fractions I and V from WT (black) and 5xFAD (gray) mice as a percentage of total recovered radioactivity. The data represents an average percentage of total recovered activity from whole brain (n = 7WT/8TG), cerebellum (n = 7WT/7TG), and hippocampus (n = 8WT/5TG). (D) Averaged percentages of total recovered activity in the six pooled fractions for whole brain, cerebellum, and hippocampus (animal numbers as in panel C). *Indicates a difference with 0.01< P < 0.05. **Indicates a difference with P < 0.01. P values were determined by t test.

Figure 4

Immunofluorescence. The brain from the temporal cortex of an 89 year old woman with Braak stage V/VI Alzheimer's disease at autopsy after a 9 h postmortem interval was stained with anti-Aβ (left) or anti-HNE-His (right). The center images are merged left and right images. Large square images are 50 × 50 μm; other images to same scale.

Figure 5

Correlation analysis. The fraction of total radioactivity that elutes as unmodified ARA (fraction V) is shown versus relative specific activity for (A) hippocampus and (B) cerebellum. Data for WT, TG, J20 and 5xFAD mice are aggregated.