Discovering novel plasma biomarkers for ischemic stroke: Lipidomic and metabolomic analyses in an aged mouse model

Abstract

Ischemic stroke remains a leading cause of mortality and long-term disability worldwide, necessitating efforts to identify biomarkers for diagnosis, prognosis, and treatment monitoring. The present study aimed to identify novel plasma biomarkers of neurodegeneration and inflammation in a mouse model of stroke induced by distal middle cerebral artery occlusion. Using targeted lipidomic and global untargeted metabolomic profiling of plasma collected from aged male mice 24 h after stroke and weekly thereafter for 7 weeks, we discovered distinct acute and chronic signatures. In the acute phase, we observed elevations in myelin-associated lipids, including sphingomyelin (SM) and hexosylceramide (HCER) lipid species, indicating brain lipid catabolism. In the chronic phase, we identified 12-hydroxyeicosatetraenoic acid (12-HETE) as a putative biomarker of prolonged inflammation, consistent with our previous observation of a biphasic pro-inflammatory response to ischemia in the mouse brain. These results provide insight into the metabolic alterations detectable in the plasma after stroke and highlight the potential of myelin degradation products and arachidonic acid derivatives as biomarkers of neurodegeneration and inflammation, respectively. These discoveries lay the groundwork for further validation in human studies and may improve stroke management strategies.

Keywords: 12-HETE; arachidonic acid; brain lipids; inflammation; lipids; lipoxygenase; myelin; neurodegeneration; neurofilament light; sphingolipids.

Fig. 1

NfL is a plasma biomarker of neurodegeneration after stroke in aged mice. A: Experimental design, 20- to 23-month-old male mice were subjected to distal middle cerebral artery occlusion + hypoxia (DH) stroke. Plasma was collected 24 h after stroke and weekly thereafter for 7 weeks. T₂-weighted MRI was performed on a subset of mice at 24 h and 7 weeks after stroke to assess acute and chronic pathological sequelae (i.e., infarct expansion, hippocampal edema, and ventricle enlargement). Representative T₂-weighted MR images captured 24 h after stroke illustrate an infarct centered on the somatosensory cortex, extending to the corpus callosum. Representative T₂-weighted MR images captured 7 weeks after stroke illustrate enlargement of the ipsilateral ventricle and hippocampus. B: Plasma NfL levels remain elevated for at least 7 weeks after stroke in aged mice (n = 5–10; ANOVA, ∗∗∗∗P < 0.0001; Dunnett’s multiple comparisons test, ∗∗P < 0.01, ∗∗∗∗P < 0.0001). Data are presented as mean ± SD. C: Plasma NfL levels significantly correlated with infarct volume 24 h after stroke (n = 13; Spearman r = 0.7637, ∗∗P < 0.01). D: Infarct volume 24 h after stroke significantly correlated with ventricle enlargement 7 weeks after stroke (n = 16; Spearman r = 0.5353, ∗P < 0.05). E: Infarct volume 24 h after stroke weakly correlated with hippocampal edema 7 weeks after stroke (n = 16; Spearman r = 0.4235, P > 0.05). Curved lines represent 95% confidence bands for the linear fit.

Fig. 2

Plasma lipidome 24 h after stroke in aged mice. A: Volcano plot illustrating significant differences in lipid abundance between plasma from naïve mice and mice 24 h after stroke (FDR-adjusted P < 0.05; FC > |2|). B: Hierarchical clustering heatmap depicting the top 30 plasma lipids ranked by t-test (Euclidean; Ward). C: Lipid classes, reported as summations of all species of CE, SM, and TAG, respectively, are significantly altered in the plasma 24 h after stroke (CE: n = 10–11; unpaired t-test, ∗P < 0.05; SM: n = 10–11; Mann-Whitney U test, ∗∗∗∗P < 0.0001; TAG: n = 10–11; unpaired t-test, ∗P < 0.05). D: SM are significantly elevated in the plasma 24 h after stroke (n = 10–11; unpaired t-tests, ∗∗∗∗P < 0.0001). E: HCER are significantly elevated in the plasma 24 h after stroke (n = 10–11; unpaired t-tests, ∗∗∗∗P < 0.0001). F: CE are significantly elevated in the plasma 24 h after stroke (n = 10–11; unpaired t-tests, ∗∗P < 0.01, ∗∗∗P < 0.001). G: TAG are significantly reduced in the plasma 24 h after stroke (n = 10–11; Mann-Whitney U tests, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001). Data are presented as box and whisker plots, with boxes extending from the 25^th to 75^th percentiles and whiskers extending from the minimum to maximum values.

Fig. 3

Temporal profile of the plasma lipidome after stroke in aged mice. A: Heatmap depicting group averages of the top 30 plasma lipids altered 24 h after stroke compared to naïve controls. B: TAG60:12-FA22:6 is significantly elevated in the plasma 24 h, 5 weeks, and 7 weeks after stroke (n = 4–11; ANOVA, ∗∗∗∗P < 0.0001; Dunnett’s multiple comparisons test, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗∗P < 0.0001). C: HCER(16:0) is transiently elevated in the plasma 24 h after stroke (n = 4–11; Kruskal-Wallis test, ∗∗∗∗P < 0.0001; Dunn’s multiple comparisons test, ∗∗∗∗P < 0.0001). D: SM (16:0) is transiently elevated in the plasma 24 h after stroke (n = 4–11; ANOVA, ∗∗∗∗P < 0.0001; Dunnett’s multiple comparisons test, ∗∗∗∗P < 0.0001). E: TAG56:4-FA18:1 is transiently reduced in the plasma 24 h after stroke (n = 4–11; Kruskal-Wallis test, ∗∗∗P < 0.001; Dunn’s multiple comparisons test, ∗∗P < 0.01). Data are presented as box and whisker plots, with boxes extending from the 25^th to 75^th percentiles and whiskers extending from the minimum to maximum values.

Fig. 4

Plasma metabolome 24 h after stroke in aged mice. A: Volcano plot illustrating significant differences in metabolite abundance between plasma from naïve mice and mice 24 h after stroke (FDR-adjusted P < 0.05; FC > |2|). B: Hierarchical clustering heatmap depicting the top 30 plasma metabolites ranked by t-test (Euclidean; Ward). C and D: Lipid metabolism, represented by 3-hydroxyadipate, 12-HETE, and stearoylcarnitine (C18), and amino acid metabolism, represented by indolepropionate and glycine, are altered 24 h after stroke (12-HETE, indolepropionate, glycine: n = 10–12; unpaired t-tests, ∗∗∗∗P < 0.0001; 3-hydroxyadipate, stearoylcarnitine (C18): n = 8–12; Mann-Whitney U test, ∗∗∗∗P < 0.0001). Data are presented as box and whisker plots, with boxes extending from the 25^th to 75^th percentiles and whiskers extending from the minimum to maximum values.

Fig. 5

Temporal profile of the plasma metabolome after stroke in aged mice. A: Heatmap depicting group averages of the top 30 plasma metabolites altered 24 h after stroke compared to naïve controls. B and C: Lipid metabolism, represented by 12-HETE and eicosenoylcarnitine (C20:1), and amino acid metabolism, represented by isobutyrylcarnitine (C4) and N6-methyllysine, are dysregulated for at least 24 h after stroke in aged mice (12-HETE, N6-methyllysine: n = 4–12; ANOVA, ∗∗∗∗P < 0.0001; Dunnett’s multiple comparisons tests, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; eicosenoylcarnitine (C20:1): n = 4–12; Kruskal-Wallis test, ∗∗P < 0.01; Dunn’s multiple comparisons test, ∗∗P < 0.01; isobutyrylcarnitine (C4): n = 4–12; Kruskal-Wallis test, ∗∗∗∗P < 0.0001; Dunn’s multiple comparisons test, ∗∗P < 0.01, ∗∗∗∗P < 0.0001). Data are presented as box and whisker plots, with boxes extending from the 25^th to 75^th percentiles and whiskers extending from the minimum to maximum values.

Fig. 6

Temporal profile of the plasma metabolome in the subacute and chronic phases after stroke in aged mice. A: Hierarchical clustering heatmap depicting the top 30 plasma metabolites ranked by ANOVA (Euclidean; Ward). Isobutyrylcarnitine (C4), 12-HETE, and N6-methyllysine represent novel plasma biomarkers in the acute, subacute, and chronic phases after stroke. B and C: Lipid metabolism, represented by phosphoethanolamine, 14-HDoHE/17-HDoHE, and sphinganine, and nucleotide metabolism, represented by ADP, are significantly elevated in the subacute and chronic phases after stroke (phosphoethanolamine, ADP: n = 4–12; ANOVA, ∗∗∗∗P < 0.0001; Dunnett’s multiple comparisons tests, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗∗P < 0.0001; 14-HDoHE/17-HDoHE: n = 4–12; Kruskal-Wallis test, ∗∗∗P < 0.001; Dunn’s multiple comparisons test, ∗P < 0.05, ∗∗P < 0.01; sphinganine: n = 4–12; Kruskal-Wallis test, ∗∗P < 0.01; Dunn’s multiple comparisons test, ∗P < 0.05). Data are presented as box and whisker plots, with boxes extending from the 25^th to 75^th percentiles and whiskers extending from the minimum to maximum values.

Fig. 7

Biosynthesis and biological actions of 12-HETE. Arachidonic acid, released from cellular membranes, serves as the precursor for 12-HETE synthesis. The enzymatic conversion of arachidonic acid by 12-lipoxygenase (12-LOX), occurring predominantly in leukocytes, platelets, and vascular endothelial cells, results in the formation of 12-HETE. 12-HETE, a bioactive lipid mediator, acts as a signaling molecule in various physiological and pathological processes, such as angiogenesis, efferocytosis, and platelet activation, and plays a pivotal role in the inflammatory cascade by promoting leukocyte chemotaxis and adhesion to endothelial cells. Additionally, 12-HETE contributes to the synthesis of pro-inflammatory cytokines, exacerbating the inflammatory response. [Created with BioRender.com].