Metabolomics analysis reveals an effect of homocysteine on arachidonic acid and linoleic acid metabolism pathway

Abstract

An increase in serum homocysteine level has been associated with an increased risk of vascular disease; however, the biochemical mechanisms that underlie these effects remain largely unknown. The present study aimed to use high-performance liquid chromatography-mass spectrometry (HPLC‑MS) to demonstrate the effects of serum homocysteine on human blood metabolites. A total of 75 fasting serum samples were investigated in the present study. Using a threshold of 15 µmol/l serum homocysteine level, samples were divided into high‑ and low‑homocysteine groups, and the serum extracts were analyzed with an HPLC‑MS‑based method. A total of 269 features exhibited significant differences and correlation with serum homocysteine levels in the electrospray ionization‑positive [ESI(+)] mode, and 69 features were identified in the ESI(‑) mode between the two groups. The principal component analysis plot revealed a separation between the high‑ and the low‑homocysteine groups. Metabolite set enrichment analysis identified arachidonic acid metabolism and linoleic acid metabolism as the two pathways with significantly enriched differences. These results revealed that arachidonic acid and linoleic acid metabolism may be associated with serum homocysteine levels and may be involved in homocysteine-induced vascular disease.

Keywords: homocysteine; metabolomics; high-performance liquid chromatography-mass spectrometry; arachidonic acid; linoleic acid.

Figure 1.

Experimental workflow for the serum metabolomics of different homocysteine level in human. Serum samples were detected with ESI in the positive and negative modes. The program XCMS was used for nonlinear alignment of raw data and the extraction of peak intensities. The SDF between high- and low-serum homocysteine groups were selected based on 75 serum samples and following this, the correlation between the SDF values and serum homocysteine levels were analyzed. The different and correlated features selected were searched against the Metlin database, and the corresponding compounds that were matched in the Metlin database were further analyzed by principal component analysis and metabolic pathway enrichment analysis methods in MetaboAnalyst software. cHCY, homocysteine concentration; HPLC/ESI-MS, high-performance liquid chromatography/electrospray ionization-mass spectrometry; SDF, significantly different features.

Figure 2.

PCA of the high- and low-serum homocysteine groups in ESI(+)-MS and ESI(−)-MS modes. PCA plots were generated from serum sample data sets of two groups, high- vs. low-serum homocysteine. (A) PCA plot of the two sample groups in ESI(+)-MS mode. (B) PCA plot of the two sample groups in ESI(−)-MS mode. Green circles represent the low-serum homocysteine group; Red circles indicate the high serum homocysteine group. ESI(+/−)-MS, electrospray ionization-mass spectrometry in the positive/negative mode; PCA, principal component analysis.

Figure 3.

Arachidonic acid and linoleic acid metabolism signaling pathways. (A) The 17 compounds that were matched in Metlin database are annotated (filled black circles) in the arachidonic acid pathway. (B) The 4 compounds that were matched in Metlin database are annotated (filled black circles) in the linoleic acid metabolism signaling pathway.