Pyruvate dehydrogenase kinase regulates vascular inflammation in atherosclerosis and increases cardiovascular risk

Abstract

Aims: Recent studies have revealed a close connection between cellular metabolism and the chronic inflammatory process of atherosclerosis. While the link between systemic metabolism and atherosclerosis is well established, the implications of altered metabolism in the artery wall are less understood. Pyruvate dehydrogenase kinase (PDK)-dependent inhibition of pyruvate dehydrogenase (PDH) has been identified as a major metabolic step regulating inflammation. Whether the PDK/PDH axis plays a role in vascular inflammation and atherosclerotic cardiovascular disease remains unclear.

Methods and results: Gene profiling of human atherosclerotic plaques revealed a strong correlation between PDK1 and PDK4 transcript levels and the expression of pro-inflammatory and destabilizing genes. Remarkably, the PDK1 and PDK4 expression correlated with a more vulnerable plaque phenotype, and PDK1 expression was found to predict future major adverse cardiovascular events. Using the small-molecule PDK inhibitor dichloroacetate (DCA) that restores arterial PDH activity, we demonstrated that the PDK/PDH axis is a major immunometabolic pathway, regulating immune cell polarization, plaque development, and fibrous cap formation in Apoe-/- mice. Surprisingly, we discovered that DCA regulates succinate release and mitigates its GPR91-dependent signals promoting NLRP3 inflammasome activation and IL-1β secretion by macrophages in the plaque.

Conclusions: We have demonstrated for the first time that the PDK/PDH axis is associated with vascular inflammation in humans and particularly that the PDK1 isozyme is associated with more severe disease and could predict secondary cardiovascular events. Moreover, we demonstrate that targeting the PDK/PDH axis with DCA skews the immune system, inhibits vascular inflammation and atherogenesis, and promotes plaque stability features in Apoe-/- mice. These results point toward a promising treatment to combat atherosclerosis.

Keywords: CVD; PDK; atherosclerosis; immunometabolism; inflammation.

Figure 1

PDK1 and PDK4 positively correlate with inflammation and the vulnerability index in human atherosclerotic plaques, and PDK1 is associated with the risk of future cardiovascular (CV) events. (A) Heat map of RNA-Seq gene expression data from human carotid plaques (CPIP biobank) showing correlations between the different PDK isoforms and genes related to inflammation or plaque stabilization. The genes were manually categorized as pro- inflammatory and plaque destabilizing or anti-inflammatory and plaque stabilizing. Black cells represent nonsignificant correlations or R = 0. Spearman correlations, corrected P < 0.05 was considered significant; n = 78. (B) Spearman correlation analysis between PDK1–4 expression and the plaque vulnerability index on CPIP samples; n = 47. (C) Kaplan–Meier cardiovascular event-free survival curves after carotid endarterectomy stratified by tertiles of PDK1–4 expression; n = 70. (D) Proteomic analysis of different PDK isoenzymes in atherosclerotic plaques vs. adjacent tissue (n = 18/group) from BiKE. Data are expressed as mean ± SEM of normalized tandem mass tag (TMT) ratios. Two-tail paired Wilcoxon test was performed for statistical analysis. (E) Immunoblot analysis from human carotid plaques showing PDH phosphorylation sites in symptomatic (S) and asymptomatic (AS) patients from BiKE (n = 6 samples/group); Mann–Whitney U test analysis. (F) Quantification of cytokine production in cells isolated from human atherosclerotic plaques (pooled data from 3 independent experiments in triplicate); Kruskal–Wallis ANOVA with Dunn’s post-test analysis. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Figure 2

The expression pattern of PDK isoenzymes in human atherosclerotic plaques. (A–D) Dot plots depicting mRNA levels of PDK isoenzymes in common immune and vascular cell populations defined by their cell-type marker genes in human plaques (public data sets analysed using PlaqView; www.plaqview.com). Dot size depicts the percentage fraction of cells expressing a gene, and dot colour intensity depicts the degree of expression of each gene. (A) Data from carotid plaques, n = 18; (B) data from carotid plaques, n = 3; (C) data from carotid plaques, n = 3; (D) data from coronary plaques, n = 35. (E, F) Plaque material from human carotid endarterectomies was sectioned and representative samples were stained by immunofluorescence to detect PDK1–4 isoenzymes (columns), which are shown in red. Immunofluorescence costaining was performed for the markers for (E) CD3⁺ T-cells, (F) CD68⁺ macrophages, and (G) alpha-smooth muscle cell actin (αSMA), which are shown in green. Slides were stained with the nuclear counterstain DAPI. Pictures were acquired with 40 times magnification. Scale bar = 30 μm. NK, NK cells; Mono, monocytes; Mφ, macrophages; DC, dendritic cells; EC, endothelial cells; SMC, smooth muscle cells; FB, fibroblasts.

Figure 3

Targeting PDK reduces atherosclerosis and vascular inflammation and increases plaque stability features. (A) Percentage lesion areas of Sudan IV en face stained aortas. The right panels are representative samples from the control and DCA-treated groups (0.1 and 1 mg/mL; ∼17 and 170 mg/kg/day). (B) The per cent lesion area in oil-red-O-stained cross-sections of the aortic root for each group. Right panels show representative micrographs from each group; (A, B) n = 10, 9, and 14 for control, DCA 0.1 mg/mL (∼17 mg/kg/day), and 1.0 mg/mL (∼170 mg/kg/day), respectively; Kruskal–Wallis ANOVA with Dunn’s post-test analysis. Immunohistochemical analysis of (C) VCAM-1 and (D) CD68− staining and (E) CD4− and (F) Foxp3-positive cells and (G) αSMA immunostaining in the aortic root of controls and DCA-treated mice. (H) Collagen evidenced by sirius red staining of controls and DCA-treated mice. Right panels show representative pictures for each group. Original magnification, 40×. Scale bar = 100 μm. (I) Vulnerability index as defined in methods. (J) Percentage of necrotic core area. Right panels show representative pictures for each group. Original magnification, 40×. Scale bar = 100 μm. All results show the mean ± SEM and are pooled data from 2 independent experiments. (C–E, F–I) n = 10 and 13 for controls and 1 mg/mL for the DCA-treated group, respectively; (F) n = 10 and 12 for controls and the treated group, respectively; (J) n = 10 and 8 for controls and the DCA-treated group, respectively. (C–J) 1 mg/mL (∼170 mg/kg/day) DCA, Mann–Whitney U test analysis. ^#P = 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Figure 4

Inhibition of PDK shifts immune response and proliferation. Aortic mRNA quantification of (A) M1 and M2 macrophage signature markers, (B) CD4 T-cell signature markers and (C) aortic mRNA quantification of innate-immunity cytokines in controls and DCA-treated mice; (A–C) n = 5 and 7 for controls and DCA-treated groups, respectively; Mann–Whitney U test analysis. (D) Expression of the NLRP3 inflammasome-related proteins IL-1β, pro-caspase 1, and cleaved caspase 1 and the loading control vinculin in aortas from controls and DCA-treated mice (n = 3 samples/group). (E) mRNA quantification of M1 and M2 macrophage signature genes in the spleen (n = 7 and 8 for control and DCA-treated groups, respectively); Mann–Whitney U test analysis. (F) M2/M1 ratio in the spleen of controls and DCA-treated mice (n = 7 and 8 for control and DCA-treated groups, respectively); Mann–Whitney U test analysis. Right panels show representative plots for M2 (CD206+) and M1 (CD11c+) macrophages within gated CD11b+ F4/80+ splenocytes. (G) mRNA quantification of CD4+ T-cell signature markers in the spleen. (H) Treg/Th17 ratio. Right panels show representative flow cytometry plots for Treg (Foxp3⁺) and Th17 (RorγT⁺) T-cells from the spleen of controls and DCA-treated mice. Representative plots show percentages among CD4⁺ T-cells. (I) Quantification of splenocyte proliferation in vitro from controls and DCA-treated mice, in response to anti-CD3 and anti-CD28 stimulation for 48 h. (J) Cytokines in supernatants from the proliferation assay; (G–J) n = 7 and 8 for control and DCA-treated groups, respectively; Mann–Whitney U test analysis. (A–J) 1 mg/mL (∼170 mg/kg/day) DCA. The results show the mean ± SEM and are representative of two independent experiments. #P = 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Figure 5

Inhibition of PDK reduces the pro-inflammatory response of the metabolite succinate in atherosclerosis. (A) Enzymatic PDH activity from protein extract of aortic tissue from controls and DCA-treated mice (n = 4 and 5, respectively); Mann–Whitney U test analysis. (B) Representative immunostainings of phosphorylated PDH E1a (serine 232) in aortic roots from control and DCA-treated mice. (C) 1H-NMR metabolomic analysis of aortic tissue from controls and DCA-treated mice (n = 4/group); Mann–Whitney U test analysis. Effect size differences and P-value between groups are shown. Quantification of (D) lactate and (E) succinate in the aortic tissue of controls and DCA-treated mice (n = 4/group); Mann–Whitney U test analysis. (A–G) 1 mg/mL (∼170 mg/kg/day) DCA. (F) Correlation between aortic levels of succinate (1H-NMR) and % of plaque in the aortic arch of controls and DCA-treated mice (n = 8); linear regression analysis. (G) Correlation between plasma levels of succinate and aortic mRNA levels of Il1b in controls and DCA-treated mice (n = 11); linear regression analysis. Effect of DCA on (H) succinate release (I) SDH activity (n = differentiated cells from six individual mice, treated in duplicates); Kruskal–Wallis ANOVA with Dunn’s post-test analysis. (J) Sirt3 mRNA levels (n = differentiated cells from four individual mice, treated in duplicates); Kruskal–Wallis ANOVA with Dunn’s post-test analysis. (K) IL-1β release by BMDMs upon NLRP3 priming and activation, with DCA added before and (^&) after LPS stimulation (n = differentiated cells from six individual mice, treated in duplicates); Kruskal–Wallis ANOVA with Dunn’s post-test analysis. (L) Upper panel shows the experimental timeline for NLRP3 inflammasome activation and DCA treatment. Bottom panels, analysis of the protein expression of pro-caspase 1 and cleaved caspase 1 in representative BMDM protein extracts. (M) Evaluation of transfection efficiency of THP-1 cells incubated with small interfering RNA against PDK1–4 (PDK1–4 siRNA) or scrambled negative siRNA control (control), as described in methods (n = 3, pooled data from two experiments, in duplicate wells); the percentage expression after siRNA transfection is shown on the top of the bars. (N) THP-1 cells transfected with PDK1–4 siRNA or scrambled control siRNA and differentiated with PMA (10 nM for 24 h) were stimulated with LPS and ATP for NLRP3 priming and activation and of IL-1β release (n = 6, pooled data from two experiments, in triplicate wells). #P < 0.07, *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 6

Inhibition of PDK reduces GPR91-mediated inflammasome activation. Quantification of IL-1β released from BMDMs upon NLRP3 inflammasome activation with LPS and ATP and/or pre-treated with DCA 1 h before ATP. (A) Effect of treatment with sodium succinate (NaSucc) 1 h prior to NLRP3 inflammasome activation with LPS and ATP and/or pre-treated with DCA. (B) Effect of treatment with 200 ng/mL pertussis toxin (PTX) for 2 h and NaSucc for 1 h before NLRP3 inflammasome activation with LPS and ATP and/or pre-treated with DCA. (C) Effect of treatment with 50 µM GPR91 inhibitor (GPR91i) for 2 h and NaSucc for 1 h before NLRP3 inflammasome activation with LPS and ATP and/or pre-treated with DCA. (D) Effect of treatment with NaSucc 1 h prior to NLRP3 inflammasome activation with LPS and ATP and/or pre-treated with DCA on the release of IL-1β by BMDMs obtained from GPR91KO mice. Results show mean ± SEM of IL-1β production (n = differentiated cells from (A–C) five and (D) six individual mice, treated in duplicate). (A–D) Kruskal–Wallis ANOVA with Dunn’s post-test analysis. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.