Imbalance in fatty-acid-chain length of gangliosides triggers Alzheimer amyloid deposition in the precuneus

Abstract

Amyloid deposition, a crucial event of Alzheimer's disease (AD), emerges in distinct brain regions. A key question is what triggers the assembly of the monomeric amyloid ß-protein (Aß) into fibrils in the regions. On the basis of our previous findings that gangliosides facilitate the initiation of Aß assembly at presynaptic neuritic terminals, we investigated how lipids, including gangliosides, cholesterol and sphingomyelin, extracted from synaptic plasma membranes (SPMs) isolated from autopsy brains were involved in the Aß assembly. We focused on two regions of the cerebral cortex; precuneus and calcarine cortex, one of the most vulnerable and one of the most resistant regions to amyloid deposition, respectively. Here, we show that lipids extracted from SPMs isolated from the amyloid-bearing precuneus, but neither the amyloid-free precuneus nor the calcarine cortex, markedly accelerate the Aß assembly in vitro. Through liquid chromatography-mass spectrometry of the lipids, we identified an increase in the ratio of the level of GD1b-ganglioside containing C20:0 fatty acid to that containing C18:0 as a cause of the enhanced Aß assembly in the precuneus. Our results suggest that the local glycolipid environment play a critical role in the initiation of Alzheimer amyloid deposition.

Fig 1. Aß assembly on reconstituted membrane of lipids extracted from SPMs isolated from autopsied brains.

A-D, Representative immunohistochemical staining of the amyloid-free precuneus (A), the calcarine cortex of the brain providing A (B), the amyloid-bearing precuneus (C) and the calcarine cortex of the brain providing C (D). E, Procedure of AFM of the reconstituted lipid bilayer composed of SPM lipids. The monolayer of SPM lipids was transferred to the POPC-coated mica by horizontal deposition (the reconstituted lipid bilayer). The lipid bilayer was incubated with Aß42, and then the surface topography of the bilayer was imaged by AFM. F, Concentration-dependent Aß assembly on reconstituted membranes prepared from lipids extracted from the amyloid-bearing precuneus. Surface topographic studies of lipid bilayer were performed by AFM using SPM lipids extracted from the amyloid-bearing precuneus. AFM images were taken after Aß incubation at 20 nM, 200 nM, and 2 μM for 12 hr. To illuminate the area of Aß accumulation, original AFM images (upper) were binarized on the basis of the heights of membrane before its incubation with Aß42. White pixels in binarized AFM images (lower) were identified as Aß accumulation. Scale bar, 100 μm (A–D), 500 nm (E).

Fig 2. Aß assembly in the presence of the lipids extracted from SPMs isolated from autopsy brains.

A, Representative original (upper) and binarized (lower) AFM images of Aß mixtures incubated on the reconstituted membranes prepared from extracted lipid samples; P1 from the amyloid-free precuneus, C1 from the calcarine cortex of the brain providing P1, P2 from the amyloid-bearing precuneus, and C2 from the calcarine cortex of the brain providing P2. The SPM sample containing same amount of proteins was used for each analysis. AFM images were taken after Aß incubation at 20 μM for 15 min. B, Area of Aß42 assembly on the membrane. C, Area of Aß42 assembly on the membrane of P2 in the presence of 4396C or normal IgG. The mean of two independent experiments is shown in B and C. D, Western blots of insoluble (Ppt) and soluble (Sup) Aß42 following incubation with liposomes prepared from P2 and C2. The insoluble Aß42 assemblies were solubilized in formic acid prior to Western blot analysis. The blot was immunostained with the anti-Aß antibody 6E10. E, TEM image of Aß42 mixtures incubated with liposomes prepared from P2. Typical amyloid fibrils with 6–10 nm diameter were observed (arrowheads). Asterisks indicate liposomes. Scale: 500 nm (A), 100 nm (E).

Fig 3. Lipid analyses of cholesterol and sphingomyelin of SPMs.

A, Cholesterol level of SPMs. A, The cholesterol level of SPMs was determined by the cholesterol oxidase method using an Amplex-Red cholesterol assay kit. B, Composition of sphingomyelins of SPMs. Extracted sphingomyelins from SPMs were analyzed by reverse-phase LC-MS using a C30 column. The proportion of each sphingomyelin species containing diverse ceramide structures in four major sphingomyelins (containing d18 or d20 as the sphingoid base and C18:0 in the fatty acid chain) is expressed as mean ± SEM. P1, C1, P2, and C2 indicate lipid samples extracted from SPMs of the amyloid-free precuneus, the calcarine cortex of the brain with the amyloid-free precuneus, the amyloid-bearing precuneus, and the calcarine cortex of the brain with the amyloid-bearing precuneus, respectively. The SPM sample containing same amount of proteins was used for each analysis.

Fig 4. Increased proportion of GD1(d20:1–20:0) ganglioside species in the amyloid-bearing precuneus.

A, Representative chromatograms of reverse-phase LC-MS of gangliosides using a C30 column. GM1(d18:1–¹³C16:0) [m/z 1532.88 (shown on the left side on the chromatogram)] was used as the internal standard (ISD). Ganglioside species containing diverse ceramide structures, such as d18:1 or d20:1 in the sphingoid base and 18:0 or 20:0 in the fatty acid chain, were separately detected in this system. The expression level of each ganglioside species was calculated by measuring each peak area above the baseline. B, Composition of gangliosides of SPMs. Gangliosides in the lipid samples extracted from SPMs were analyzed by reverse-phase LC-MS as indicated A. The proportion of each ganglioside species containing diverse ceramide structures in the detected gangliosides is expressed as mean ± SEM. *, p<0.05. P1, C1, P2 and C2 indicate lipid samples extracted from SPMs of the amyloid-free precuneus, the calcarine cortex of the brain with the amyloid-free precuneus, the amyloid-bearing precuneus, and the calcarine cortex of the brain with the amyloid-bearing precuneus, respectively.

Fig 5. Analyses of gangliosides by normal-phase LC-MS with regard to the glycan structure of gangliosides.

A, Scheme of synthesis pathway of a- and b-series of gangliosides. B, Representative chromatograms of normal-phase LC-MS of gangliosides using an NH₂ column. GM1(d18:1–¹³C16:0) [m/z 1532.88 (shown on the left side on the chromatogram)] was used as ISD. In this system, a-series (GD1a) and b-series (GD1b) GD1-gangliosides were separately detected on the basis of the diversity of the glycan structure. The expression level of each ganglioside species was calculated by measuring each peak area above the baseline.

Fig 6. Increased ratio of level of (d20:1–20:0) to that of (d20:1–18:0) in GD1b-ganglioside in the amyloid-bearing precuneus.

Ratios of the level of (d20:1–20:0) to that of (d20:1–18:0) in GD1a- and GD1b-gangliosides were calculated from the levels of a- and b-series of gangliosides obtained by normal-phase LC-MS using an NH₂ column. The ratio is expressed as mean ± SEM. *, p<0.05. P1, C1, P2 and C2 indicate lipid samples extracted from SPMs of the amyloid-free precuneus, the calcarine cortex of the brain with the amyloid-free precuneus, the amyloid-bearing precuneus, and the calcarine cortex of the brain with the amyloid-bearing precuneus, respectively.

Fig 7. Effect of the alteration in the ratio of level of (d20:1–20:0) to that of (d20:1–18:0) in GD1b-ganglioside on Aß assembly.

A, Addition of GD1b(d20:1–20:0) ganglioside to the lipids extracted from the amyloid-free precuneus significantly enhanced Aß assembly as determined by AFM. B, Addition of GD1b(d20:1–18:0) ganglioside to the lipids extracted from the amyloid-bearing precuneus significantly inhibited Aß assembly as determined by AFM. The subspecies of GD1b-ganglioside was purified from commercially available GD1b-gangliosides and added to the extent that its ratio equaled (center, × 1) or exceeded by four-fold (right, × 5) the original ratio in the lipids extracted from the amyloid-bearing precuneus. Scale bar, 500 nm.