Rat Hair Metabolomics Analysis Reveals Perturbations of Unsaturated Fatty Acid Biosynthesis, Phenylalanine, and Arachidonic Acid Metabolism Pathways Are Associated with Amyloid-β-Induced Cognitive Deficits

Abstract

Hair is a noninvasive valuable biospecimen for the long-term assessment of endogenous metabolic disturbance. Whether the hair is suitable for identifying biomarkers of the Alzheimer's disease (AD) process remains unknown. We aim to investigate the metabolism changes in hair after β-amyloid (Aβ_1-42) exposure in rats using ultra-high-performance liquid chromatography-high-resolution mass spectrometry-based untargeted and targeted methods. Thirty-five days after Aβ_1-42 induction, rats displayed significant cognitive deficits, and forty metabolites were changed, of which twenty belonged to three perturbed pathways: (1) phenylalanine metabolism and phenylalanine, tyrosine, and tryptophan biosynthesis-L-phenylalanine, phenylpyruvate, ortho-hydroxyphenylacetic acid, and phenyllactic acid are up-regulated; (2) arachidonic acid (ARA) metabolism-leukotriene B4 (LTB4), arachidonyl carnitine, and 5(S)-HPETE are upregulation, but ARA, 14,15-DiHETrE, 5(S)-HETE, and PGB2 are opposite; and (3) unsaturated fatty acid biosynthesis- eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), FA 18:3 + 1O, and FA 18:3 + 2O are downregulated. Linoleic acid metabolism belonging to the biosynthesis of unsaturated fatty acid includes the upregulation of 8-hydroxy-9,10-epoxystearic acid, 13-oxoODE, and FA 18:2 + 4O, and downregulation of 9(S)-HPODE and dihomo-γ-linolenic acid. In addition, cortisone and dehydroepiandrosterone belonging to steroid hormone biosynthesis are upregulated. These three perturbed metabolic pathways also correlate with cognitive impairment after Aβ_1-42 stimulation. Furthermore, ARA, DHA, EPA, L-phenylalanine, and cortisone have been previously implicated in the cerebrospinal fluid of AD patients and show a similar changing trend in Aβ_1-42 rats' hair. These data suggest hair can be a useful biospecimen that well reflects the expression of non-polar molecules under Aβ_1-42 stimulation, and the five metabolites have the potential to serve as novel AD biomarkers.

Keywords: Arachidonic acid; Hair metabolism; L-Phenylalanine; Targeted/untargeted metabolomics; Unsaturated fatty acid.

Fig. 1

Experimental design. An overview workflow of the metabolomics analysis in AD. Hair samples from Aβ_1-42-induced rats (n = 10 rats) and sham rats (n = 10 rats) were collected and subjected to untargeted metabolomics analysis. The AD rat model was verified based on motor activity, spatial learning and memory tests, and histological analyses. After hair sample decontamination, homogenization, and extraction, the hair extracts were analyzed with UHPLC-HRMS. The raw MS data were converted to peak lists using MS-DIAL for feature detection and alignment. The significantly altered features were identified by MS/MS analysis to explore chemical markers, and then the metabolic pathways were analyzed

Fig. 2

Learning and memory performance and motor activity for sham and Aβ_1-42-induced rats. The spontaneous alternation in the A Y-maze, B the latency to elapse, C numbers of working memory errors, and D numbers of reference memory errors in the radial maze tests, and E the latency to fall off the rotarod and F the maximum speed reached during the test in the rotarod tests were presented. Data are present mean ± SD. The number of animals used was n = 10 for each experimental group. *p < 0.05 versus the Sham group. ^#p < 0.05 versus the Pre-OP Aβ  group. Pre-OP, preoperation; Post-OP, postoperation

Fig. 3

Histological analysis in Aβ_1-42-induced rats and sham rats and correlation analysis between behavioral test and histological effects. Histology of hippocampus from Aβ_1-42-induced and sham rats using HE staining (A, B), NeuN + Fluoro-Jade B staining (C, D), thioflavin-S staining (E, F), and NeuN + TUNEL staining (G, H). Correlation analysis between Y maze test and (I) HE/ (J) NeuN + Fluoro-Jade B/ (K) Thioflavin-S/ (L) NeuN + TUNEL staining. The number of animals used was n = 10 for each experimental group

Fig. 4

Differential metabolites and the correlation between metabolites and histological analysis and behavioral test in Aβ_1-42-induced rat. Volcano plot analysis illustrated the identified features between Aβ_1-42-induced and sham rats in A positive and B negative mode. The orange spot represents the features displayed with larger magnitude fold changes (x-axis, |log₂(Aβ_1-42/Sham) |≥ 1.0) and statistical significance difference (y-axis, -log P-value ≥ 1.33). The correlations between C Y maze and 8-hydroxy-9,10-epoxystearic acid, D Y maze and FA 18:2 + 4O, E Thio-S staining and LysoPC(17:0), F Thio-S staining and sphingosine, G Thio-S staining and 13-OxoODE, H Thio-S staining and l-phenylalanine, I FJB staining and 8-hydroxy-9,10-epoxystearic acid, and J TUNEL staining and 8-hydroxy-9,10-epoxystearic acid. The number of animals used was n = 10 for each experimental group

Fig. 5

Classification, pathway analysis, and correlations of the differential hair metabolites between Aβ_1-42-induced and sham rats. A Thirty-three differential hair metabolites were classified. B Pathway analysis based on differential metabolites using online software MetaboAnalyst (www.metaboanalyst.ca). The correlations between C TUNEL staining and LTB4, D HE staining and 14,15-DHET, E TUNEL staining and PGB2, F HE staining and 5(S)-HPETE, G TUNEL staining and 5(S)-HPETE, H FJB staining and 5(S)-HETE, I FJB staining and DHEA, and J TUNEL staining and DHEA. The number of animals used was n = 10 for each experimental group

Fig. 6

Overview of the altered metabolites and perturbed pathways in Aβ_1-42-induced rats. The changes affected by Aβ_1-42 are indicated by the pink and green boxes. Pink box: upregulation in Aβ_1-42-induced rats compared with sham rats; green box: downregulation in Aβ_1-42-induced rats compared with sham rats; blue box: not analyzed in UHPLE-HRMS analysis. (ALA, α-linolenic acid; DGLA, dihomo-γ-linolenic acid; L-PHE, l-phenylalanine; PC, phosphatidylcholine; PEA, phenethylamine; PGB2, prostaglandin B2; PPA, phenylpyruvic acid; LPC: lysophosphatidylcholine; LTB4, leukotriene B4; LA, linoleic acid)