Sirtuin 4 accelerates heart failure development by enhancing reactive oxygen species-mediated profibrotic transcriptional signaling

Abstract

Aims: Sirtuin 4 (SIRT4) is a mitochondrially-localized stress-responsive NAD⁺-dependent deacetylase predominantly regulating energy metabolism and reactive oxygen species (ROS) homeostasis. Overexpression of SIRT4 aggravates angiotensin-induced cardiac hypertrophy, however underlying mechanisms remain incompletely elucidated. To current study was designed to explore mechanisms underlying adverse effects of increased SIRT4 levels in the heart following pressure overload.

Methods and results: Mice with cardiomyocyte-specific overexpression of Sirt4 (cSirt4-Tg) or non-transgenic controls underwent transverse aortic constriction (TAC) or sham procedure. Cardiac structure, function and energy metabolism were assessed by echocardiography and working heart perfusions. Transcriptome analysis was performed using RNA sequencing. Nine weeks following TAC and thereafter, cSirt4-Tg mice displayed exacerbated cardiac dilation, dysfunction, and fibrosis compared to non-transgenic controls. This aggravation was accompanied by impaired rates of glycolysis and a blunted increase of mitochondrial respiratory capacity. More importantly, expression of numerous genes encoding collagens and profibrotic regulators was elevated. This profibrotic signaling was reversed by mitochondria-targeted antioxidant treatment using MitoQ, along with attenuation of cardiac dysfunction and reversal of structural remodeling. SIRT4 may drive oxidative stress and fibrotic signaling via increased NOX4 expression (>7-fold), and/or direct modulation of potential SIRT4 targets newly identified by Human Protein Microarray, including calcitonin gene-related peptide receptor component protein, cyclophilin A, and interleukin-2 receptor β.

Conclusions: SIRT4 overexpression accelerates heart failure development in response to pressure overload, predominantly by ROS-mediated enhancement of profibrotic transcriptional signaling.

Keywords: Fibrosis; Heart failure; Oxidative stress; Reactive oxygen species; SIRT4; Sirtuin.

Graphical abstract

Fig. 1

Cardiac remodeling, dysfunction, and fibrosis in cSirt4-Tg mice following TAC. A) Schematic representation of the generation of cSirt4-Tg mice. B) SIRT4 protein expression in heart, liver, kidney, and skeletal muscle from cSirt4-Tg mice (n = 2). Ejection fraction C) bi-weekly and D) at 11 weeks following Sham or TAC surgery of Control and cSirt4-Tg mice (n = 9–10). E) HW/BW ratio (n = 9–15), F) gene expression of hypertrophy markers in hearts (n = 9–10), G) LW/BW ratio (n = 9–14), and H) myocardial fibrosis (n = 6–8) of Control and cSirt4-Tg mice at 12 weeks following Sham or TAC surgery. Scale bar is 100 μm. 2-Way ANOVA: #, effect of genotype; §, effect of treatment; %, effect of interaction. *p < 0.05 versus Control, $p < 0.05 versus Sham (C); *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 (D—H) using Fisher's LSD test. BW, body weight; cSirt4-Tg, cardiomyocyte-specific overexpression of Sirt4; HW, heart weight; LW, lung weight; NPPA, natriuretic peptide A; NPPB; natriuretic peptide B; SIRT4, sirtuin 4; TAC, transverse aortic constriction. Representative M-mode images were scanned from a printout of echocardiography from a Vivid 7 Dimension micro-imaging system (D).

Fig. 2

Cardiac contractile function and energy substrate utilization in cSirt4-Tg mice following TAC. A) Cardiac power (n = 9–14), B) palmitate oxidation, C) glycolysis, D) glucose oxidation, E) ratio of glycolysis to glucose oxidation (n = 4–6), and F) MVO₂ (n = 9–14) measured in isolated working hearts of Control and cSirt4-Tg mice 12 weeks following Sham or TAC surgery. Gene expression of proteins related to G) FAO, and H) GLOX and I) sirtuins in hearts of Control and cSirt4-Tg mice at 12 weeks following Sham or TAC surgery (n = 9–10). J) Oxygen consumption rates, K) ATP synthesis and L) ATP/O ratio in the presence of glutamate/malate measured in mitochondria isolated from hearts 12 weeks following Sham or TAC surgery (n = 6–9). 2-Way ANOVA: #, effect of genotype; §, effect of treatment; %, effect of interaction. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 using Fisher's LSD test. ATP, adenosine triphosphate; ATP/O, ATP to oxygen ratio; CPT, carnitine palmitoyltransferase; cSirt4-Tg cardiomyocyte-specific overexpression of Sirt4; FAO, fatty acid oxidation; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GLOX, glucose oxidation; GLUT, glucose transporter; HADHB, hydroxyacyl-CoA dehydrogenase beta; HK2, hexokinase 2; LCAD, long-chain acyl-CoA dehydrogenase; MCAD, medium-chain acyl-CoA dehydrogenase; MVO₂, myocardial oxygen consumption; PDK4, pyruvate dehydrogenase kinase 4; PPARα, peroxisome proliferator-activated receptor α; TAC, transverse aortic constriction.

Fig. 3

Myocardial oxidative stress in cSirt4-Tg mice following TAC. A) 4-HNE levels from hearts of Control and cSirt4-Tg mice 12 weeks following Sham or TAC surgery with or without MitoQ supplementation (n = 5–7). B) Ejection fraction (n = 9–12), C) HW/BW (n = 5–12) and D) myocardial fibrosis (n = 6–7) of Control and cSirt4-Tg mice 12 weeks following TAC surgery with or without MitoQ supplementation. E) Acetylation of MnSOD at K68 normalized to total MnSOD expression in hearts from Control and cSirt4-Tg mice 12 weeks following TAC surgery with or without MitoQ supplementation. Scale bar is 100 μm. 2-Way ANOVA: #, effect of genotype; §, effect of treatment; %, effect of interaction. *p < 0.05; ***p < 0.001 using Fisher's LSD test. Ac, acetylated; BW, body weight; cSirt4-Tg, cardiomyocyte-specific overexpression of Sirt4; HNE, hydroxyneonal; HW, heart weight; MitoQ, mitoquinone; MnSOD, manganese superoxide dismutase; SIRT4, sirtuin 4; TAC, transverse aortic constriction. Data presented for vehicle-treated TAC mice in B—D are same as Fig. 1 D, E, H. Representative M-mode images were scanned from a printout of echocardiography from a Vivid 7 Dimension micro-imaging system (B).

Fig. 4

Transcriptional reprogramming of the cSirt4-Tg heart following TAC. Heatmap of DEGs, as determined by next-generation RNA-sequencing analysis in hearts from Control and cSirt4-Tg mice following A) Sham or B) TAC surgery and C) TAC surgery following MitoQ supplementation. D) Venn diagrams of overlapping DEGs following Sham or TAC surgery, with or without MitoQ supplementation. E) Pathway analysis of DEGs by Reactome_2016 in cSirt4-Tg hearts following relative to Controls. DEGs determined based on a Wald test (P < 0.05 & Fold change >1.5) with Benjamini-Hochberg post hoc adjustment. cSirt4-Tg, cardiomyocyte-specific overexpression of Sirt4; DEG, differentially expressed genes; MitoQ, mitoquinone; SIRT4, sirtuin 4; TAC, transverse aortic constriction.

Fig. 5

Transcriptional reprogramming of fibrotic signaling in cSirt4-Tg mice following TAC. Gene expression of proteins involved in A) collagen expression, B) TGF-β signaling pathway and Smad proteins, C) MMPs and TIMPs, D) fibrosis-associated molecule, E) Wnt/β-catenin axis and Fzd receptors (n = 5–6) and F) downstream EGR1–3 (n = 6) in hearts of Control and cSirt4-Tg mice 12 weeks following TAC surgery with or without MitoQ supplementation as determined by RNA-sequencing. 2-Way ANOVA: #, effect of genotype; §, effect of treatment; %, effect of interaction. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 using Fisher's LSD test. ACTA2, actin alpha 2, smooth muscle; CCN2, cellular communication network factor 2; CTGF, connective tissue growth factor; COL, collagen; cSirt4-Tg, cardiomyocyte-specific overexpression of Sirt4; FN1, fibronectin 1; FRZB, frizzled related protein; FZD, frizzled class receptor 1; MitoQ, mitoquinone; MMP, matrix metalloproteinases; LOXl, lysyl oxidase-like proteins; PLOD3, procollagen-lysine,2-oxoglutarate 5-dioxygenase 3; POSTN, periostin; SIRT4, sirtuin 4; SFRP, soluble frizzled-related proteins; SMAD, sma- and mad-related proteins; TAC, transverse aortic constriction; TFGB, transforming growth factor beta; TGFBI, TGFβ-induced protein; TGIF, TGFβ induced factor homeobox; TIMP, tissue inhibitors of matrix metalloproteinase; VIM, vimentin; WNT, wingless/integrated.

Fig. 6

Transcriptional reprogramming of ROS signaling in cSirt4-Tg mice following TAC. Changes in expression of genes encoding for proteins involved in A) ROS production as determined by RNA-sequencing (n = 5–6) and B) validation of Nox4 expression by RT-qPCR (n = 9–10); and C) antioxidant stress defense as determined by RNA-sequencing (n = 5–6) and D) validation of Gpx1 expression by RT-qPCR (n = 9–10) in hearts of Control and cSirt4-Tg mice 12 weeks following TAC surgery with or without MitoQ supplementation. 2-Way ANOVA: #, effect of genotype; §, effect of treatment; %, effect of interaction. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 using Fisher's LSD test. CAT, catalase; cSirt4-Tg, cardiomyocyte-specific overexpression of Sirt4; CYBA, cytochrome b-245, alpha chain; CYBB, cytochrome b-245, beta chain; GPX, glutathione peroxidase; MAOA, monoamine oxidase A; MAOB, monoamine oxidase B; MitoQ, mitoquinone; NOX4, NADPH oxidase 4; PRDX3, peroxiredoxin 3; SOD2, superoxide dismutase 2, mitochondrial; TAC, transverse aortic constriction; TMX, thioredoxin-related transmembrane protein; XDH, xanthine dehydrogenase.

Fig. 7

SIRT4 targets identified by Human Protein Microarray. A) Scan of Human Protein Microarray in the presence or absence (Control) of human recombinant SIRT4 protein. The pattern of red fluorescence spots (always in duplicate) on the control chip serves to facilitate specific protein identification following fluorescence excitation. Fluorescence spots specifically identified on the array incubated with SIRT4 represent potential SIRT4 interaction targets (exemplarily illustrated by the white arrow). B) List of proteins identified by protein-protein interaction with SIRT4 using a Human Protein Microarray. SIRT4, sirtuin 4. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)