Alzheimer's disease manifests abnormal sphingolipid metabolism

Abstract

Introduction: Alzheimer's disease (AD) is associated with disturbed metabolism, prompting investigations into specific metabolic pathways that may contribute to its pathogenesis and pathology. Sphingolipids have garnered attention due to their known physiological impact on various diseases.

Methods: We conducted comprehensive profiling of sphingolipids to understand their possible role in AD. Sphingolipid levels were measured in AD brains, Cerad score B brains, and controls, as well as in induced pluripotent stem (iPS) cells (AD, PS, and control), using liquid chromatography mass spectrometry.

Results: AD brains exhibited higher levels of sphingosine (Sph), total ceramide 1-phosphate (Cer1P), and total ceramide (Cer) compared to control and Cerad-B brains. Deoxy-ceramide (Deoxy-Cer) was elevated in Cerad-B and AD brains compared to controls, with increased sphingomyelin (SM) levels exclusively in Cerad-B brains. Analysis of cell lysates revealed elevated dihydroceramide (dhSph), total Cer1P, and total SM in AD and PS cells versus controls. Multivariate analysis highlighted the relevance of Sph, Cer, Cer1P, and SM in AD pathology. Machine learning identified Sph, Cer, and Cer1P as key contributors to AD.

Discussion: Our findings suggest the potential importance of Sph, Cer1P, Cer, and SM in the context of AD pathology. This underscores the significance of sphingolipid metabolism in understanding and potentially targeting mechanisms underlying AD.

Keywords: Alzheimer’s disease (AD); ceramide (Cer); ceramide 1-phosphate (Cer1P); sphingomyelin (SM); sphingosine (Sph).

FIGURE 1

Sphingolipid modulation in AD brain tissues. Sphingolipid levels were assessed in the brains of control subjects (n = 6), the patients with Cerad-b (n = 7), and those with AD (n = 6). Statistical evaluation of differences was conducted using one-way analysis of variance (ANOVA) and multiple comparisons with Tukey correction. (A) Dihydrosphingosine (dhSph), dihydrosphingosine 1-phosphate (dhS1P), sphingosine (Sph), and sphingosine 1-phosphate (S1P) levels. (B) Total levels of the ceramide 1-phosphate (Cer1P), ceramide (Cer), hexosyl ceramide (HexCer), lactosyl-ceramide (LacCer), dihydroceramide (DH Cer), deoxy-ceramide (Deoxy-Cer), deoxy dihydroceramide (Deoxy-dhCer), and sphingomyelin (SM). Statistical significance was denoted as follows: * for p < 0.05, ** for p < 0.01, *** for p < 0.001. Red * indicates a significant increase, while blue * indicates a significant decrease.

FIGURE 2

Sphingolipid modulation in iPS cells. Sphingolipid levels were measured in normal model neurons (C), mutated Presenilin 1-inserted iPS cell (PS), and AD patient derived iPS cell (AD), each group N = 8. Statistical evaluation of differences was conducted using one-way analysis of variance (ANOVA) and multiple comparisons with Tukey correction. (A) DhSph, dhS1P, Sph, and S1P levels. (B) Total levels of the Cer1P, Cer, Hex Cer, LacCer, dhCer, Deoxy-Cer, Deoxy-dhCer, and SM. Statistical significance was denoted as follows: * for p < 0.05, ** for p < 0.01, *** for p < 0.001. Red * indicates a significant increase.

FIGURE 3

Multivariate analysis of sphingolipids in AD brain tissues and cells. Score plots illustrate distinct profiles of sphingolipids among AD brains (A), Cerad-B brains, and control brains, as well as among AD, PS, and control cells (C). Variable Importance in the Projection (VIP) plots, derived from the PLS-DA models, ranked sphingolipids according to their discriminative ability in brain tissue (B) and cell samples (D).

FIGURE 4

Representative correlations between sphingolipids and the AD-associated markers. The correlations between sphingolipids and AD-associated markers were assessed using Spearman rank correlation. (A,B) Sph correlation with the Braak score and SP. (C,D) Correlations of Aβ40 levels with total Deoxy-dhCer and total dhCer. (E,F) Correlations of Aβ42 levels with total Deoxy-Cer and total Cer.

FIGURE 5

Discriminant analysis using the machine learning model. Illustration of predictor importance in brain tissue (A) and cell samples (C) using SPSS Modeler’s Predictor Importance Chart in the SVM model. This chart delineates the relative importance of each predictor, highlighting the top 10 sphingolipids ranked by their importance. (B) Association of measured sphingolipids with study groups was examined using the Kendall rank correlation. The analysis considered age, gender, and clinical (pathological) data [Aβ levels, Braak stage, and senile plaque (SP) grade] as covariates of interest.

FIGURE 6

Schematic illustration of sphingolipid modulation in AD. Graphs depicting the modulation of sphingolipids in brain tissue (A) and cell samples (B) relevant to AD.