Cholesteryl hemiazelate identified in CVD patients causes in vitro and in vivo inflammation

Abstract

Oxidation of PUFAs in LDLs trapped in the arterial intima plays a critical role in atherosclerosis. Though there have been many studies on the atherogenicity of oxidized derivatives of PUFA-esters of cholesterol, the effects of cholesteryl hemiesters (ChEs), the oxidation end products of these esters, have not been studied. Through lipidomics analyses, we identified and quantified two ChE types in the plasma of CVD patients and identified four ChE types in human endarterectomy specimens. Cholesteryl hemiazelate (ChA), the ChE of azelaic acid (n-nonane-1,9-dioic acid), was the most prevalent ChE identified in both cases. Importantly, human monocytes, monocyte-derived macrophages, and neutrophils exhibit inflammatory features when exposed to subtoxic concentrations of ChA in vitro. ChA increases the secretion of proinflammatory cytokines such as interleukin-1β and interleukin-6 and modulates the surface-marker profile of monocytes and monocyte-derived macrophage. In vivo, when zebrafish larvae were fed with a ChA-enriched diet, they exhibited neutrophil and macrophage accumulation in the vasculature in a caspase 1- and cathepsin B-dependent manner. ChA also triggered lipid accumulation at the bifurcation sites of the vasculature of the zebrafish larvae and negatively impacted their life expectancy. We conclude that ChA behaves as an endogenous damage-associated molecular pattern with inflammatory and proatherogenic properties.

Keywords: atherosclerosis; cholesteryl hemiazelates; cholesteryl hemiesters; innate inflammatory responses; lipidomics.

Fig. 1

ChA levels are increased in plasma of CVD patients. Boxplots depicting ChA and ChE of 1,11-undecendioic acid (ChU) concentrations across patient groups. A: Kruskal–Wallis one-way analysis of variance was applied to test whether ChA concentration across samples originates from the same distribution. Unpaired two-sample Wilcoxon test was applied as a post hoc test to evaluate significant differences in lipid levels for a patient group versus the control group (indicated by “p”). The horizontal black dashed line indicates the global mean across all samples. Plasma concentration of ChA was obtained by shotgun lipidomics of 74 donors with ACS (including ST-elevation and non-ST-elevation myocardial infarction and unSAP), 71 donors with SAP, and 52 age-matched control cohort. ChA was quantitatively analyzable in 34.6% of the control group, 69% of the SAP group, and 83.8% of the ACS group. B: Lipidomes of the atheromata obtained from six different carotid artery endarterectomy specimens. 1) Includes ether phosphatidylcholines; 2) includes ether phosphatidylethanolamines; 3) includes cardiolipins, phosphatidic acids, phosphatidylglycerols, phosphatidylinositols, and phosphatidyl serines; 4) includes lysophosphatidic acids, lysophosphatidylcholines, ether-lysophosphatidylcholines; lysophosphatidylethanolamines, ether-lysophosphatidylethanol-amines, lysophosphatidylglycerols; lysophosphatidylinositols, and lysophosphatidylserines. C: ChEs detected in atheromata obtained from the six carotid artery endarterectomy specimens.

Fig. 2

ChA changes the expression of surface markers and the inflammatory profile of human primary monocytes (Mono) and MDMs. Peripheral blood was harvested from human volunteers. Mononuclear leukocytes (peripheral blood mononuclear cells) were isolated from whole blood using histopaque-1077. CD14⁺ cells were purified from total peripheral blood mononuclear cells by immunomagnetic separation. CD14⁺ monocytes were cultured in the presence of ChA (10 μM or 25 μM) or vehicle for 24 h (eight different donors) or for 7 days to obtain MDM (six different donors). Monocytes were differentiated using rM-CSF. After lipid treatment, a representative sample of cells was stained with monoclonal antibodies to determine surface expression of HLA-DR (A), CD86 (B), CD206 (C), and CD163 (D); data represent the mean fluorescence intensity (MFI) ± SEM for all subjects normalized to the control cells. Cell culture supernatants were analyzed by ELISA for IL-1β (E), IL-6 (F), and IL-10 release (G). Values represent the mean ± SEM of cytokine released (in pg/ml). Reactive oxygen species production by monocytes and MDM exposed to ChA (H). ∗P < 0.05; ∗∗P < 0.01; and ∗∗∗P < 0.001 using two-way ANOVA with Tukey’s multiple comparisons test. rM-CSF, recombinant macrophage colony-stimulating factor.

Fig. 3

ChA affects the cytokine profile of neutrophils. Neutrophils were harvested from human volunteers (five different donors) using a gradient with histopaque-1177 and 1009. Cells were cultured in the presence of ChA (25, 100, and 500 μM) or vehicle for 4 h. After treatment, neutrophils were stained with monoclonal antibodies to determine the expression of CD11b (A) and cytokines: IL-1β (B), TNF-α (C), IL-6 (D), and IL-10 (E); data represent the percentage of positive cells for each inflammatory marker normalized to the control cells. ∗P < 0.05; ∗∗P < 0.01 using one-way ANOVA with Dunnett’s multiple comparisons test.

Fig. 4

ChA induces infiltration of neutrophils and macrophages in zebrafish caudal vein. A: Schematic representation indicating with rectangles the sites of lipid microinjection and imaging. B: Representative images of the caudal vein of Tg (mpeg.mCherryCAAX SH378, mpx:EGFP i114) zebrafish larvae fed for 2 or 10 days with normal, FC (4% [w/w] 103 μmol per g of food, positive control), or ChA (3% [w/w], 52 μmol per g of food)-enriched diets and imaged by confocal microscopy. The feeding started at 5 days postfertilization. Images are Z-stacks of fluorescent green (neutrophils) and red (macrophages) cells and the respective bright field. Arrows point to neutrophils and macrophages that are in the close proximity. Dashed lines delineate the caudal vein. Scale bars represent 20 μm. C and D, quantification of neutrophils (green cells, C) and macrophages (red cells, D) after 2 (gray color) or 10 days of feeding (red color). Infiltration of neutrophils occurs earlier than macrophages. The results are shown as mean ± SEM of three independent experiments (in each independent experiment, at least five larvae were analyzed per condition). E: Recruitment of neutrophils (gray color) and macrophages (green color) into the vasculature of Tg (mpeg.mCherryCAAX SH378, mpx:EGFP i114) larvae 24 h after the microinjection of POPC (27 μM, vehicle, Ct), FC:POPC (50 μM FC), and ChA:POPC liposomes (50 μM ChA). The graphs represent the mean ± SEM of three independent experiments (at least three larvae in each independent experience were analyzed in each condition). F–J: Quantitative RT-PCR for vcam-1 (F), il-1β (G), tnf-α (H), il-6 (I), and il-10 (J) after 2 or 10 days of feeding with normal, FC-enriched, or ChA-enriched diets. The values represent the mean ± SEM normalized to the control of three independent experiments (n = 20–30 zebrafish larvae per group); ∗P < 0.05 ∗∗p ˂ 0.01; ∗∗∗P < 0.001 using one-way ANOVA with Dunnett’s multiple comparisons test.

Fig. 5

Inflammasome and cathepsin-B inhibition decrease ChA-induced myeloid cell infiltration. A–D: Zebrafish larvae were fed for 10 days with normal, 4% FC-enriched, or 3% ChA-enriched diets, and macrophages were isolated by fluorescence-activated cell sorting (FACS). A: Representative confocal images of macrophages stained with BODIPY 493/503 to visualize neutral lipids. B: Quantification of neutral lipid accumulation on isolated macrophages. C: Representative confocal images of macrophages stained with LysoTracker to visualize lysosomes/acidic organelles. Scale bars represent 5 μm. D: Quantification of the lysosome area on isolated macrophages. In B and D, the results are the mean ± SEM of two independent experiments (at least 10 cells were analyzed per condition); ∗∗∗P < 0.001 using one-way ANOVA with Dunnett’s multiple comparisons test. E: In vivo lysosomal imaging of macrophages (m-Cherry) from larvae fed with different diets. Lysosomes were stained with LysoTracker (green). Scale bars represent 5 μm. F–H: Effect of caspase 1 (Casp-1) and cathepsin B (Cath B) inhibitors on neutrophils and macrophage infiltration into the caudal vein of Tg (mpeg.mCherryCAAX SH378, mpx:EGFP i114) zebrafish larvae fed with normal, 10% FC-enriched, or 14% ChA-enriched food. F: Representative confocal images. Images are Z-stacks of fluorescent green (neutrophils) and red (macrophages) cells and the respective bright fields. Dashed lines delineate the caudal vein. Scale bars represent 20 μm. G and H: Quantification of Casp-1 and Cath B inhibitors on neutrophil (G) and macrophage (H) infiltration into the caudal vein. Three independent experiments were performed, and at least three larvae were analyzed per experiment (total of >10 larvae per condition); ∗P < 0.05; ∗∗∗P < 0.001 using two-way ANOVA with Dunnett’s multiple comparisons test.

Fig. 6

ChA induces lipid accumulation in the vasculature and is toxic to zebrafish larvae. Five days postfertilization, zebrafish larvae were fed for 10 days with normal (in gray), FC-enriched (in blue), or ChA-enriched (in red) diets. A: Confocal z-projection images of fluorescent lipid deposits (in red, indicated by the arrows) of caudal vein of AB larvae. For visualization of lipid structures, diets (normal, 2 or 4% FC-enriched, 3 or 6% ChA-enriched food) were supplemented with 10 μg/g of a red fluorescent CE. Scale bars represent 20 μm. B: Quantification of total lipid structure area in the zebrafish caudal vein. Fluorescent images of at least 10 larvae were quantified per condition. The results are shown as mean ± SEM; ∗∗∗P < 0.001 using one-way ANOVA with Dunnett’s multiple comparisons test. C: Z-projection of the caudal vein of fli1:EGFP larvae fed as above. Green fluorescence corresponds to endothelial cells (ECs), and red fluorescence corresponds to deposits of lipids localized at the bifurcation sites (arrows). Scale bars represent 20 μm. D: Evaluation of larvae survival as a function of FC (in blue) or ChA (in red) concentration in the diet. Lipid concentration is given by logarithm of 26 μmol/g of food (FC 1%, ChA 1.5%), 52 μmol/g of food (FC 2%, ChA 3%), 103 μmol/g of food (FC 4%, ChA 6%), 155 μmol/g of food (FC 6%, ChA 9%), and 207 μmol/g of food (FC 8%, ChA 12%). Results are mean of two independent experiments (each experiment with 40 larvae). The error bars indicate the SEM and ∗∗P < 0.01; ∗∗∗P < 0.001 using one-way ANOVA with Tukey’s multiple comparisons test.

Fig. 7

Working model of the proatherogenic properties of ChA. ChA is an end product of cholesteryl linoleate oxidation, generated in the arterial intima. Because of its amphiphilic properties, ChA can be detected in the plasma of CVD patients. The presence of ChA in circulation can imprint an inflammatory phenotype in the circulating monocytes and neutrophils (1), conditioning the immunological response in the arterial intima. ChA promotes the recruitment of innate immune cells, neutrophils, and monocytes into the vasculature (2). Here, neutrophils in the presence of ChA secret IL-1β, which can interfere with monocyte/macrophage priming. Monocytes differentiated in the presence of ChA are activated, increasing the secretion of inflammatory cytokines: IL-1β, IL-6, and IL-10 (3). In activated macrophages, ChA induces lipid accumulation (foam cells) and lysosomal dysfunction, conferring then the second signal necessary for IL-1β secretion mediated by inflammasome activation (4). IL-1β can initiate a propagation loop of the inflammation, increasing the macrophage secretion of IL-6 and TNF-α. On the other hand, dysfunctional lysosomes will decrease the clearance capacity of macrophages, leading to lipid accumulation in the arterial intima.