Targeting long-chain acylcarnitine accumulation to protect cardiac mitochondrial homeostasis after complete revascularization

Abstract

Approximately 20% of acute myocardial infarction (AMI) patients with multivessel disease experience adverse outcomes after complete revascularization. We aim to investigate the underlying metabolic mechanism of ischemia-reperfusion injury responsible for abnormal hemodynamic stresses in high-risk patients undergoing complete revascularization. Elevated preoperative serum levels of long-chain acylcarnitine (LCAC) 16:1 are associated with an increased risk of poor prognosis following complete revascularization. Multi-omics analyses reveal that reperfusion injury activates fatty acid degradation, and carnitine palmitoyltransferase 1A (CPT1A) is identified as a key regulator of LCACs in the interaction network in porcine models. In the early stages of reperfusion injury in non-culprit lesions, the release and prolonged elevation of circulating LCACs primarily depend on the activation of endothelial CPT1A through hemodynamic injury, which can be reduced using an inhibitor (etomoxir). Excess LCACs enter cardiomyocytes via the organic cation transporter 2, leading to imbalanced mitochondrial quality control and causing cardiomyocyte death.

Keywords: CPT1A; complete revascularization; hemodynamic shear stress; long-chain acylcarnitine metabolism; non-culprit lesion; reperfusion injury.

Graphical abstract

Figure 1

Elevated LCAC concentrations were associated with MACEs in patients with acute myocardial infarction and multivessel disease (A) Enrollment and analysis process of patients with myocardial infarction and multivessel disease. (B) Untargeted metabolomic profiling of plasma from patients in the screening datasets (n = 39). Volcano plots highlighted the serum metabolites that increased (red) in myocardial injury, as compared to non-myocardial injury group. (C) Heatmap of LCACs in myocardial injury and non-myocardial injury groups in the screening datasets (n = 39); metabolites were rank ordered by fold change. (D) Forest plot of multivariable Cox proportional hazard model for LCACs in derivation cohorts (n = 216).Hazard ratio (HR) and p value calculated by a Cox proportional hazards model. ∗p < 0.05. (E and F) Concentrations of LCAC C16:1 in myocardial injury and non-myocardial injury/MACE and non-MACE groups in derivation cohorts (n = 216) or validation cohorts (n = 161), respectively. Data were presented as means ± standard deviations as indicated. Data were compared by unpaired Student’s t test. ∗p < 0.05; ∗∗p < 0.01. (G) Kaplan-Meier curves of LCAC C16:1 in derivation cohorts and validation cohorts (n = 377). MACEs, adverse cardiovascular events; LCAC, long-chain acylcarnitine.

Figure 2

Construction of a swine model of acute myocardial infarction with multivessel coronary disease to simulate reperfusion injury in non-culprit lesions (A) Flowchart of multivessel disease swine model to mimic reperfusion injury in non-culprit lesion. (B) Angiography shows that the model was successfully constructed. The red arrow on the left shows the culprit vessel, while the arrow on the right shows the non-culprit artery. The blue arrow on the left shows collected coronary vein blood. (C) Electrocardiogram at 30 min of infarction. (D) Serum high-sensitivity troponin concentrations were measured at baseline, 30 min after infarction, and at 30 and 90 min after reperfusion in both the SSR and COR groups (n = 6/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 and ∗∗p < 0.01 in a Student’s unpaired t test. (E) Malondialdehyde levels after recanalization in the SSR and COR groups (n = 6/group). Data were presented as means ± standard deviations as indicated. ∗∗∗p < 0.001 in a Student’s unpaired t test. (F) ATP production in heart tissue after recanalization in the SSR and COR groups (n = 6/group). Data were presented as means ± standard deviations as indicated. ∗∗∗p < 0.001 in a Student’s unpaired t test. (G) Representative images of TUNEL staining of heart tissue after recanalization in the SSR and COR groups. Scale bar: 50 μm. (H) Ratio of TUNEL-positive cells (%) in the SSR and COR groups after recanalization. Data were presented as means ± SD as indicated. Data were compared by unpaired Student’s t test (n = 6/group); ∗∗p < 0.01 in a Student’s unpaired t test. LAD, left anterior descending artery; LCX, left circumflex artery; SSR, single-stage revascularization group; COR, culprit-only revascularization group; hs-TnI, high-sensitivity troponin I; MDA, malondialdehyde; ATP, adenosine triphosphate.

Figure 3

Multi-omics data to identify regulator of the LCAC response to reperfusion injury (A) KEGG metabolic pathway enrichment of differential metabolites (SSR vs. COR) of coronary veins in after 30 and 90 min of reperfusion (n = 6/group). (B) Trend expression of differential metabolites of coronary veins at baseline, 30 min after infarction, and at 30 and 90 min after reperfusion in SSR group (n = 6/group). (C) Volcano plot of differential proteins (SSR vs. COR) in infarcted heart tissue (n = 6/group). (D) KEGG pathway enrichment analysis of differential proteins (SSR vs. COR). (E) Heatmap of differential proteins of fatty acid degradation and the oxidative phosphorylation pathways in the SSR and COR groups (n = 6/group). (F) GSEA of fatty acid degradation and the oxidative phosphorylation (SSR vs. COR). (G) Protein-metabolism interaction network and subnetwork interacting with LCACs. (H) Immunostaining for CD31/cTNT (green) and CPT1A (red) on infarcted cardiac tissue sections from the SSR and COR groups. Scale bar, 50 μm. KEGG, Kyoto Encyclopedia of Genes and Genomes; SSR, single-stage revascularization group; COR, culprit-only revascularization group; FC, fold change; GSEA, gene set enrichment analysis; NES, normalized enrichment score; LCAC, long-chain acylcarnitine; MCC, maximal clique centrality.

Figure 4

Shear stress-dependent endothelial CPT1A regulates LCAC metabolism in vitro (A) CPT1A protein expression was quantified in HUVECs and NCMs that were exposed to FSS or HR (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗∗p < 0.01 in a Student’s unpaired t test. (B) Concentrations of LCAC C16:1 in culture medium of HUVECs and NCMs that were exposed to FSS or HR (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗∗p < 0.01 in a Student’s unpaired t test. (C) Protein CPT1A expression quantitation treated with FSS/siRNA-CPT1A in HUVECs (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 and ∗∗p < 0.01 in a Student’s unpaired t test. (D) Representative fluorescence image of CPT1A treated with FSS/siRNA-CPT1A in HUVECs. Scale bar, 100 μm. (E) Concentrations of LCAC C16:1 in culture medium treated with FSS/siRNA-CPT1A in HUVECs (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗∗p < 0.01 in a Student’s unpaired t test. (F) Protein CPT1A expression quantitation treated with FSS/ETO in HUVECs (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 in a Student’s unpaired t test. (G) Representative fluorescence image of CPT1A treated with FSS/ETO in HUVECs. Scale bar, 100 μm. (H) Concentrations of LCAC C16:1 in culture medium treated with FSS/ETO in HUVECs (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 and ∗∗p < 0.01 in a Student’s unpaired t test. HUVEC, human umbilical vein endothelial cell; NCMs, mouse neonatal cardiomyocytes; FSS, fluid shear stress; HR, hypoxia reoxygenation; CON, control; siRNA, siRNA-CPT1A; ETO, etomoxir; LCAC, long-chain acylcarnitine.

Figure 5

Endothelial cell-derived LCACs affect cardiomyocytes, leading to cell apoptosis and mitochondrial dysfunction in vitro (A) Schematic diagram of the in vitro experimental protocol: (1) HUVECs subjected to FSS stimulation were subsequently co-cultured with ACMs or (2) ACMs were treated with various concentrations of LCAC C16 (0, 50, 100, and 150 μM). (B) Apoptotic cells were labeled with annexin V-FITC (green fluorescence) in ACMs from the co-culture and control groups. Scale bar, 100 μm. (C) Representative photomicrographs of JC-1 staining in ACMs from the co-culture and control groups. Scale bar, 100 μm. (D) Quantification of the JC-1 fluorescence ratio in ACMs from the co-culture and control groups (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗∗∗p < 0.001 in a Student’s unpaired t test. (E) ATP levels of ACMs from the co-culture and control groups (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗∗p < 0.01 in a Student’s unpaired t test. (F) Mito NAD/NADH ratio of ACMs from the co-culture and control groups (n = 3/group). Data were presented as means ± standard deviations as indicated.∗∗p < 0.01 in a Student’s unpaired t test. (G–I) Oxygen consumption rate of ACMs from the co-culture and control groups using the seahorse system. The basal respiration and maximal respiration were assessed (n = 3/group). Data were presented as means ± standard deviations as indicated.∗∗p < 0.01 in a Student’s unpaired t test. (J) Apoptotic cells were labeled with annexin V-FITC (green fluorescence) in ACMs treated with various concentrations of LCAC C16 (0, 50, 100, and 150 μM). Scale bar, 100 μm. (K) Representative photomicrographs of JC-1 staining in ACMs treated with various concentrations of LCAC C16 (0, 50, 100, and 150 μM). Scale bar, 100 μm. (L) Quantification of the JC-1 fluorescence ratio in ACMs treated with various concentrations of LCAC C16 (0, 50, 100, and 150 μM) (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 in a Student’s unpaired t test. (M) ATP levels of ACMs treated with various concentrations of LCAC C16 (0, 50, 100, and 150 μM) (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗∗p < 0.01 and ∗∗∗p < 0.001 in a Student’s unpaired t test. (N) Mito NAD/NADH ratio of ACMs treated with various concentrations of LCAC C16 (0, 50, 100, and 150 μM) (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗∗p < 0.01 and ∗∗∗p < 0.001 in a Student’s unpaired t test. (O–Q) Oxygen consumption rate of ACMs treated with various concentrations of LCAC C16 (0, 50, 100, and 150 μM) using the Seahorse system. The basal respiration and maximal respiration were assessed (n = 3/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 in a Student’s unpaired t test. (R and S) Molecular docking results of OCTN2 and palmitoylcarnitine (binding energy = −6.5 kcal/mol) and acetyl carnitine (binding energy = −5.0 kcal/mol). HUVEC, human umbilical vein endothelial cell; ACMs, adult rat cardiomyocytes; FSS, fluid shear stress; CON, control; OCR, oxygen consumption rate; LCAC, long-chain acylcarnitine; OCTN2, type 2 organic cation transporter.

Figure 6

LCAC C16 causes cardiac mitochondrial homeostasis disruption by inhibiting the expression of PPARGC1A (A)Transcriptome analysis volcano plot of mRNA expression ACMs treated with LCAC C16. (B) KEGG pathway enrichment results of differentially expressed genes of ACMs treated with LCAC C16. (C) GSEA of oxidative phosphorylation pathways. (D) GO biological process enrichment in LCAC C16-treated ACMs. (E and F) Gene Ontology analysis of MQC system-related biology process changes and heatmap of key genes in the mitochondrion organization from RNA sequencing analysis. (G) MQC system-related differentially expressed genes interaction network. (H and I) Expression of PGC-1α, an MQC system-related biomarker, in myocardial tissue and coronary venous blood from SSR and COR groups in a swine model of AMI (n = 6/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 and ∗∗p < 0.01 in a Student’s unpaired t test. (J) Comparison of preoperative and postoperative levels of LCAC C16:1 between high- and low-concentration groups in patients with AMI and multivessel disease (n = 5/group). Data were presented as means ± standard deviations as indicated. ∗∗p < 0.01 in a Student’s unpaired t test. (K) Correlation between postoperative serum LCAC C16:1 levels and PGC-1α expression in patients with AMI and multivessel disease (n = 10). R, correlation coefficient; Spearman correlation analysis. (L) Correlation between ratio of LCAC C16:1 levels and ratio of PGC-1α (Post/Pre) in high and low LCAC C16:1 concentration groups (n = 5/group).R, correlation coefficient; Spearman correlation analysis. R, correlation coefficient; Spearman correlation analysis. FC, fold change; KEGG, Kyoto Encyclopedia of Genes and Genomes; LCAC, long-chain acylcarnitine; GO-BP, Gene Ontology biological process; ACMs, adult rat cardiomyocytes; GSEA, Gene Set Enrichment Analysis; NES, normalized enrichment score; MQC, mitochondrial quality control; SSR, single-stage revascularization group; COR, culprit-only revascularization group; AMI, acute myocardial infarction; Pre, preoperative; Post, postoperative.

Figure 7

Modulation of CPT1A-LCAC pathway alters post-ischemic cardiac remodeling and cardiac injury (A) Myocardial ischemia-reperfusion injury was induced by ligation of the left anterior descending coronary artery in the middle of the mid-ant region (I-0H) for 90 min and then reperfusion for the indicated time points (R-0H, 2H, 6H, 1D, 3D, and 7D). Cardiac tissues were processed for immunofluorescence staining. Immunostaining of CD31, cTNT, and CPT1A (red) in infarcted hearts. Scale bar: 50 μm. (B) Transmission electron microscopy images (×8,000 and ×15,000) of representative mitochondrial areas in infarcted hearts from the LCAC, IR, and ETO groups. (C) ATP production in infarcted hearts from the LCAC, IR, and ETO groups (n = 5/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 in a Student’s unpaired t test. (D) Mito NAD/NADH ratio in infarcted hearts from the LCAC, IR, and ETO groups (n = 5/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 in a Student’s unpaired t test. (E) Representative M-mode echocardiographic images of LCAC, IR, and ETO groups. Scale bar in mm/s on the right, and time stamp in seconds at the bottom. (F and G) Echocardiographic quantifications of LCAC, IR, and ETO groups (n = 5/group). Shown in the statistical graph are left ventricular ejection fraction (EF) and left ventricular fractional shortening (FS). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 in a Student’s unpaired t test. (H and I) Heart tissue WGA staining to quantification of cross-sectional area of cardiomyocytes in LCAC, IR, and ETO groups (n = 5/group). Scale bar: 25 μm. Data were presented as means ± standard deviations as indicated. ∗p < 0.05 in a Student’s unpaired t test. (J and K) Heart tissue Masson trichrome staining of LCAC, IR, and ETO groups (scale bar: 2 and 25 μm). Quantification of cardiac fibrosis area from Masson trichrome-stained sections in LCAC, IR, and ETO groups (n = 4/group). Data were presented as means ± standard deviations as indicated. ∗p < 0.05 and ∗∗p < 0.01 in a Student’s unpaired t test. LAD, left anterior descending coronary artery; MI/R and IR, myocardial ischemia-reperfusion injury; LCAC, long-chain acylcarnitine; ETO, etomoxir; EF, ejection fraction; FS, fractional shortening; WGA, wheat germ agglutinin staining.