Docosahexaenoic acid protects from dendritic pathology in an Alzheimer's disease mouse model

Abstract

Learning and memory depend on dendritic spine actin assembly and docosahexaenoic acid (DHA), an essential n-3 (omega-3) polyunsaturated fatty acid (PFA). High DHA consumption is associated with reduced Alzheimer's disease (AD) risk, yet mechanisms and therapeutic potential remain elusive. Here, we report that reduction of dietary n-3 PFA in an AD mouse model resulted in 80%-90% losses of the p85alpha subunit of phosphatidylinositol 3-kinase and the postsynaptic actin-regulating protein drebrin, as in AD brain. The loss of postsynaptic proteins was associated with increased oxidation, without concomitant neuron or presynaptic protein loss. n-3 PFA depletion increased caspase-cleaved actin, which was localized in dendrites ultrastructurally. Treatment of n-3 PFA-restricted mice with DHA protected against these effects and behavioral deficits and increased antiapoptotic BAD phosphorylation. Since n-3 PFAs are essential for p85-mediated CNS insulin signaling and selective protection of postsynaptic proteins, these findings have implications for neurodegenerative diseases where synaptic loss is critical, especially AD.

Figure 1. Dendritic Caspase Activation and Selective Drebrin Loss in Tg2576 Mouse and Selective Postsynaptic Marker Loss in Alzheimer’s Disease

(A–C) Fractin immunoreactivity was detected in dendrites (de) and occasionally in spines (s) but not in axons (ax) in the cortex of aged Tg2576 mice. Calibration bars, 500 nm. (D) Drebrin (Dr) in young (6–8 months) Tg(+) mice compared to old (16–18 months) Tg(+) mice in lysate from cortical membrane fractions. Values shown are percentage of young Tg(+) as the mean ± SEM. ^‡‡p < 0.01 versus young Tg(+). (E) Effect of transgene on fractin/actin ratio, drebrin (Dr), and synaptophysin (S) levels in membrane fraction of mouse cortex. Transgenic (+), n = 7–8, controls (−), n = 7–12. **p < 0.01 versus controls [Tg(−)]. (F) Drebrin (Dr), PSD-95 (PSD), and synaptophysin (S) in temporal cortex samples from AD patients (n = 10) and controls (n = 9). Values shown are percentage of control patients as the mean ± SEM. *p < 0.05, **p < 0.01 versus controls (Ctrl). (G) Diagram illustrating the generation of fractin from cleavage of actin by caspases.

Figure 2. Western Analysis of Brains from DHA-Depleted Tg2576 Mice Show Increased Caspase-Cleaved Actin and Reduced Post-synaptic Drebrin and PSD-95, without Altering Synaptophysin

N-3 polyunsaturated fatty acid (PFA) depletion (Low DHA, n = 6) in Tg2576 mice increased the cortical fractin/actin ratio and caused a selective and massive cortical loss of drebrin and PSD-95 compared to controls (Ctrl, n = 6–7), which was prevented by the addition of docosahexaenoic acid (High DHA, n = 6). (A) The fractin/actin ratio in cortical membrane fraction. †††p < 0.001 versus Tg(+) Ctrl Diet; ^○○○p < 0.001 versus Tg(+) n-3 PFA depleted (Low DHA). (B) Effect of n-3 PFA depletion (Low DHA, n = 6) and addition of DHA (High DHA, n = 6) on drebrin, PSD-95, and synaptophysin levels (Synap.) in the cortex of Tg2576 mice compared to controls (Ctrl, n = 6–7). Values shown are percentage of Tg(−) mice on control diet as the mean ± SEM. †p < 0.05, †††p < 0.001 versus Tg(+) Ctrl Diet and ^○○p < 0.01, ^○○○p < 0.001 versus Tg(+) Low DHA Diet.

Figure 3. No Significant Neuron Loss Was Associated with n-3 Polyunsaturated Fatty Acid Depletion in the Tg2576 Mouse

(A–E) Immunostaining with NeuN, an antibody that labels neuronal nuclei, revealed similar neuronal counts in the hippocampus or cortex between Tg2576 mice fed with low n-3 PFA diet [Tg(+) Low DHA] (A), Tg(−) fed with low n-3 PFA diet [Tg(−) Low DHA] (B), Tg(+) and Tg(−) mice fed with DHA-enriched diets (High DHA) ([C] and [D], respectively), and Tg2576 mice raised on control chow (E). Original magnification, 100×. (F) Image analysis quantification of NeuN-ir (number of neuronal nuclei per 1000 μm²) was performed on neuronal layers in cortex (entorhinal II, entorhinal III/IV, parietal II, parietal III/IV, parietal V/VI, frontal II, frontal III/IV, frontal V/VI) and hippocampus (CA1, CA2, CA3) at Bregmas −1.0 mm, 1.7mm, −2.7 mm, −3 mm (analyzed from anterior to posterior hippocampus with four consecutive sections analyzed per Bregma). There were no significant neuron density changes with APPswe transgene or with diet (n = 5 or 6 in each group). Since there were no treatment effects on neuronal nuclei densities in any layer and no interaction with regions, we show interaction bars rather than the breakdown of densities in layers and different Bregmas.

Figure 4. Confocal Microscopy of the Brains of Tg2576 Mice, Illustrating that Dietary DHA Depletion Increases Punctate Casapase-Cleaved Actin, which Is Associated with Reduced Drebrin

Fractin (green) and drebrin (red) double labeling in the cortex of mice showing drebrin loss and fractin labeling in the cortex of Tg2576 mice fed a low n-3 PFA diet. (A) Periplaque green punctate fractin labeling in the cortex of a Tg2576 mouse on the low n-3 polyunsaturated fatty acids diet is shown against a reduced red drebrin labeling in the vicinity of the plaque and throughout the neuropil. (B) Fractin peptide blocks fractin labeling in an adjacent section from the same Tg(+) animal. (C and D) Representative examples showing no fractin labeling and no drebrin loss in the same region of cortex in Tg(−) mice on control diet (C) or on low n-3 PFA-enriched diet (D). Magnification, 500×.

Figure 5. Brains of DHA-Depleted Tg2576 Mice Show Loss of p85 Expression and Phospho-BAD, Corresponding to Increased Protein Oxidation

Docosahexaenoic acid (High DHA) increased p85 expression, increased phosphorylation of BAD, and reversed cortical oxidative damage induced by n-3 PFA depletion (Low DHA) in Tg2576 [Tg(+)] mice. (A) The p85α regulatory subunit of phosphatidylinositol 3 (PI3)-kinase in cortex of Tg(+) mice and in temporal cortex samples from human control and AD patients. **p < 0.01 versus human controls, ^†††p < 0.001 versus Tg(+) Ctrl, and ^○○○p < 0.001 versus Tg(+) on n-3 PFA depletion (Low DHA). (B) Quantitative real-time RT-PCR measurements of p85α subunit mRNA in cortex of Tg(−) and Tg(+) mice fed with control diet or low DHA diet. *p < 0.05 versus Tg(+) Ctrl. (C) Image analysis quantification of phospho-BAD in cortex and hippocampus combined. ^○○p < 0.01 versus Tg(+) on n-3 PFA depletion (Low DHA). (D and E) Representative examples of phospho-BAD immunostaining in hippocampus (CA1, CA2, and CA3) and parietal cortex (PC) of two mice. The two mice were depleted of N-3 (D) but the mouse in (E) received a diet that was enriched in DHA. Magnification bar, 0.3 mm. (F) The effect of DHA on cortical oxidized proteins levels. ^†††p < 0.001 versus Tg(+) Ctrl, and ^○○p < 0.01 versus Tg(+) on n-3 PFA depletion (Low DHA). (G) Proposed mechanistic pathway for the transgene- and DHA-dependent effect on postsynaptic markers. Aβ overexpression, combined with low n-3 polyunsaturated fatty acid dietary intake, generates an autocatalytic vicious cycle in the postsynaptic dendrites leading to a further increase of oxidative stress and decrease of DHA. This could lead to decreased PI3-kinase activity, caspase activation, further oxidative damage, and consequent breakdown of dendrite spine F-actin filaments and postsynaptic damage.

Figure 6. Morris Water Maze Memory Acquisition Deficits in Tg2576 Mice Fed with a DHA-Depleted Diet

Cognitive function in Tg(+) and Tg(−) mice on low n-3 polyunsaturated fatty acid diets without DHA (Low DHA, n = 6) and with DHA (High DHA, n = 6) was assessed with the Morris water maze at 21–22 months with six blocks of visible platform followed by the 12 blocks of hidden platform testing (A). Consistent with the absence of sensorimotor deficits, there were no significant differences in the latency to find a visible platform in the latter half of visible training, although the Tg(−), DHA group performed better during the first three blocks (B). Consistent with no treatment differences in anxiety levels, there was no difference in thigmotaxis during visible training (C) nor were there differences in swim speed (D). In contrast, there were significant differences in hidden platform testing. Tg(+) mice on the low DHA diet did not appear to learn, exhibiting higher latencies from blocks 8 through 12 than the other groups (E). Learning curves of the other groups were slight due to age and fatigue, limiting further training. Nevertheless, combined block analysis (2 × 2 ANOVA: treatment × combined block) in hidden training revealed some learning and an impairment in the Tg(+) low DHA group, which was significantly different than all other groups (F). The same results were obtained by repeated ANOVA analysis p < 0.005 (data not shown). 2 × 2 ANOVA (treatment × transgene) of thigmotaxis during hidden platform testing demonstrated a significant Tg(+) effect (p = 0.0001) and a significant Tg × treatment interaction (p < 0.0013), showing that the Tg(+) mice on the low DHA diet showed excessive thigmotaxis (G). Probe retention analysis (percent path in different quadrants) confirmed a significant transgene effect (p < 0.009). Unlike the Tg(+) mice, the Tg(−) groups spent more paths in the target (T) quadrant (40%–50%) and less time in Opposite quadrant (<10%, [H]). The DHA diet failed to restore the transgene-dependent retention deficit (H).