Posttranslational modulation of FoxO1 contributes to cardiac remodeling in post-ischemic heart failure

Atherosclerosis 249 (2016) 148e156

Contents lists available at ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

Posttranslational modulation of FoxO1 contributes to cardiac remodeling in post-ischemic heart failure € hr a, b, 1, Lorenzo De Angelis a, Maria Mavilio a, Ben Arpad Kappel a, b, 1, Robert Sto Rossella Menghini a, Massimo Federici a, c, * a b c

Department of Systems Medicine, University of Rome Tor Vergata, 00133 Rome, Italy Department of Internal Medicine I, University Hospital Aachen, Pauwelsstraße 30, 52074 Aachen, Germany Center for Atherosclerosis, Department of Medicine, Policlinico Tor Vergata, 00133 Rome, Italy

a r t i c l e i n f o

a b s t r a c t

Article history: Received 14 December 2015 Received in revised form 2 April 2016 Accepted 5 April 2016 Available online 7 April 2016

Objective: Forkhead box protein O1 (FoxO1) plays a key role in energy homeostasis, stress response and autophagy and is dysregulated in diabetes and ischemia. We investigated cardiac FoxO1 expression and posttranstranslational modifications after myocardial infarction (MI) and further tested if active posttranstranslational modulation of FoxO1 can alter cardiac remodeling in postischemic heart failure. Methods: Non-diabetic and diabetic C57BL/6 mice were subjected to MI by ligation of left anterior descending artery. In selected experiments we combined this model with intramyocardial injection of adenovirus expressing different isoforms of FoxO1. We used Millar catheter, histology, Western blot and metabolomics for further analyses. Results: We show that after MI total cardiac FoxO1 is downregulated and partly recovers after 7 days. This downregulation is accompanied by fundamental posttranslational modifications of FoxO1, particularly acetylation. Adenovirus experiments revealed smaller infarction size and improved heart function in mice expressing a constitutively deacetylated variant of FoxO1 compared to a wild type variant of FoxO1 in both non-diabetic (MI size:
13.4 ± 3.5%; LVDP: þ29.1 ± 9.4 mmHg; p < 0.05) and diabetic mice (MI size:
17.6 ± 3.7%; LVDP: þ10.9 ± 3.6 mmHg; p < 0.05). Metabolomics analyses showed alterations in metabolites connected to muscle breakdown, collagen/elastin and energy metabolism between the two groups. Conclusion: First, our results demonstrate that myocardial ischemia is associated with downregulation and posttranslational modification of cardiac FoxO1. Second, we show in a mouse model of postischemic heart failure that posttranslational modulation of FoxO1 alters heart function involving collagen and protein metabolism. Therefore, posttranslational modifications of FoxO1 could be an option to target remodeling processes in postischemic heart failure. © 2016 Elsevier Ireland Ltd. All rights reserved.

Keywords: FoxO1 Posttranslational modification Acetylation Diabetes Myocardial infarction Heart failure Cardiac remodeling

1. Introduction Patients with type 2 diabetes (T2DM) have a greater risk of suffering a myocardial infarction than non-diabetic individuals [1]. In addition, the diabetic heart is more vulnerable to ischemia with the outcome of myocardial infarction being worse than in patients without diabetes [2], thus contributing to the increased mortality and morbidity among patients with T2DM [3]. Although available

* Corresponding author. Department of Systems Medicine, University of Rome“Tor Vergata”, Via Montpellier 1, 00133 Rome, Italy. E-mail address: [email protected] (M. Federici). 1 Equally contributing authors. http://dx.doi.org/10.1016/j.atherosclerosis.2016.04.001 0021-9150/© 2016 Elsevier Ireland Ltd. All rights reserved.

antidiabetic drugs are effective in treating microvascular complications of T2DM, most agents have little effects on cardiovascular complications such as coronary heart disease and heart failure [4]. Therefore, the unmet demand for treatment options for diabetesrelated heart failure remains high. Alterations of the cardiac energy metabolism play an important role and contribute to the more vulnerable diabetic heart [5]. In T2DM glucose utilization of the heart is disturbed, resulting in a high influx of fatty acids in the ischemic diabetic heart exceeding its oxidative capacity. The consequently higher oxygen demand leads to intracellular acidosis, mitochondrial dysfunction and production of reactive oxygen species [6]. Recent findings suggest that posttranslational modifications of the cardiac proteome in T2DM contribute to heart failure [7].

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

Forkhead box protein O1 (FoxO1) is an ubiquitously expressed transcription factor in mammalian cells playing a key role in the regulation of metabolic homeostasis, oxidative stress, cell differentiation, proliferation, survival and autophagy [8,9]. Under conditions of diabetes, inflammation and ischemia FoxO1 is dysregulated [10e12]. The activity of FoxO1 is strictly controlled by posttranslational modifications. Insulin receptor signaling stimulates Akt-dependent phosphorylation of FoxO1 and attenuates its transcriptional activity by shuttling it from the nucleus to the cytoplasm [8,13]. The role of acetylated FoxO1 has yet not been fully elucidated [14]. Acetylation makes FoxO1 more sensitive to insulin-induced phosphorylation, therefore suggesting a reduction of its transcriptional activity [14,15]. However, it has recently been found that acetylated FoxO1 that is located in the cytoplasm regulates autophagy [16], indicating that posttranslational modification of FoxO1 goes beyond altering its transcriptional activities on target genes. On the basis that loss of FoxO1 in mice is embryonically lethal due to defective angiogenesis [17] the role of FoxO1 on the cardiovascular system has intensively been studied in in vivo-models of atherosclerosis as well as in endothelial cells and macrophages [18e21]. We recently reported that posttranslational modification of FoxO1 increases levels of asymmetric dimethylarginine in endothelial cells and confirmed this correlation in cohort of patients with unstable atherosclerosis [22]. The role of FoxO1 on cardiac metabolism, especially in T2DM, is however less clear. In cardiomyocytes FoxO1 regulates proliferation, apoptosis and autophagy [9,23e27]. In mice, lipid overload induces cardiac triglyceride accumulation via a FoxO1-dependent pathway [28] and activation of FoxO1 has been shown to contribute to diabetic cardiomyopathy [29,30]. Although FoxO1 is a critical mediator of energy metabolism, oxidative stress response, survival and autophagy in the heart, little is known on its role during myocardial ischemia and postischemic cardiac remodeling. Therefore, we conducted this study 1) to investigate how cardiac FoxO1 is chronologically expressed and posttranslationally modified following ischemia in non-diabetic and diabetic mice and 2) if active posttranscriptional modulation of FoxO1 has an impact on cardiac remodeling in a mouse model of post-ischemic heart failure.

149

To induce a non-insulin-dependent diabetes mellitus-like phenotype we combined high-fat diet with 60% kcal% fat (HFD) (D12492; Research Diets, New Brunswick, USA) with a single injection of streptozotocin (STZ) as previously described [31]: C57BL/6 mice received a HFD for 6 weeks followed by a single dose intraperitoneally-injected STZ (100 mg/kg, Sigma). HFD was continued for 2 more weeks before mice were subjected to myocardial infarction experiments. Glucose levels were measured in the fasting state with an OneTouch Lifescan Glucometer and only mice with fasting blood glucose >20 mmol/l were considered as having diabetes [32] and were used for further experiments, named HFD þ STZ mice in this manuscript. All other non-diabetic animals, called wild type (WT) mice, received normal chow food (10% calories from fat; GLP Mucedola Srl, Settimo Milanese, Italy). Animals were housed in shoebox-sized cages with free access to water and food in a 12 h day-night cycle. 2.3. Acute myocardial infarction model and adenoviral vector injection Ligation of the left anterior descending artery (LAD) was performed as a model of myocardial infarction as previously described and is explained in more detail in the supplementary data [33,34]. Green fluorescent protein-tagged ADV expressing a wild type variant of FoxO1 (FoxO1-WT) or constitutively deacetylated variant of FoxO1 (FoxO1-KR) have previously been described [22] and were injected directly into the myocardial border zone during surgery. A total of 50 mL containing 5  108 infectious particles was injected. In a preliminary experiment, we confirmed using Western blot that ADV express in the infarcted area (Suppl. Fig. S2). 2.4. Measurement of hemodynamic parameters with Millar catheter

2. Materials and methods

Hemodynamic parameters were measured with a Millar catheter (Millar Instruments, Houston, TX, USA) as previously described [33,34]. Mice were anesthetized with Ketamin/Rompun and 1Fr Millar Catheter (PVR 1035) was advanced across the right carotic artery through the aortic valve into the left ventricle. After hemodynamic stabilization signals were continuously recorded with and analyzed with LabChart (ADInstruments). A minimum of 20 representative loops was averaged.

2.1. Study design

2.5. Measurement of infarction size

In a first experiment diabetic and non-diabetic mice as described below were subjected to a model of acute myocardial infarction and sacrificed at different time points: baseline (sham operation), 1 h, 24 h and 7 days after myocardial infarction. In a second experiment we performed the myocardial infarction model combined with injection of adenoviral vector (ADV) expressing different isoforms of FoxO1. After 3 weeks heart function was assessed by using Millar catheter and mice sacrificed for further analyses (Study design: Suppl. Fig. S1). In another experiment, we repeated the myocardial infarction model with injection of ADV expressing different isoforms of FoxO1 in diabetic mice. (Study design: Suppl. Fig. S1).

After terminal anesthesia and neck dislocation animals were bled via cardiac puncture before injecting 1 mL of ice cold 1 M KCl into the ventricle. The heart was then removed, washed in PBS and transferred into 4% PFA for 48 h. After dehydration with ethanol, hearts were embedded in paraffin and 5 mm sections were taken 400 mm apart and stained with hematoxylin and eosin or Massons trichrome. The slides were then digitalized and infarction area was calculated according to the midline length measurement using ImageJ and expressed as a percentage of total heart circumference (NIH, USA; http://rsb.info.nih.gov/ij/).

2.2. Experimental animal procedures

Nuclear and cytosolic proteins were extracted by using the NEPER kit according to the manufacturers instructions (Thermo scientific #78833). Western blot was performed with the following antibodies: FKHR (H-128), Ac-FKHR (D-19)(acetylated at residues Lys 259, Lys 262 and Lys 271), actin (C-11), Sirt2 (H-95), Sirt3 (P-19) (all Santa Cruz Biotechnology), FoxO1 (#2880), Phospho-FoxO1 (phosphorylated at serine 256) (#9461), Sirt1 (#3931) (all Cell

All animal procedures were performed in accordance to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85e23, revised 1996), approved by the University Hospital of Tor Vergata Animal Care Facility. C57BL/6 male mice were purchased from Charles River.

2.6. Western blot

150

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

cytoplasm

nucleus

a Total FoxO1

Ponceau Total FoxO1 Actin

Ponceau WT

HFD+STZ

WT

baseline

b

HFD+STZ

WT

1h

HFD+STZ

WT

24h

nucleus

HFD+STZ

7 days

cytoplasm

Total Foxo1

1.0

§

§ §

0.5

§

§

§

§ §

§ 0.0

WT

baseline

1h

HFD+STZ

24h

7 days

§ vs. baseline

baseline

§

1h

24h

*§

7 days

* HFD+STZ vs. WT

Fig. 1. Cardiac FoxO1 is lost after ischemia in non-diabetic and diabetic mice. a) Representative Western blots of nuclear/cytosolic extracts of infarcted hearts of non-diabetic (WT) and diabetic mice (HFD þ STZ) and sacrifice at different time points: baseline (sham operation), 1 h, 24 h and 7 days after myocardial infarction. b) Quantitative analysis of Western blots reveals loss of nuclear and cytosolic FoxO1 in both WT and HFD þ STZ mice after myocardial infarction with a maximum after 24 h. We observed a recovery of total FoxO1 in both animal groups one week after myocardial infarction with little intergroup differences between WT and HFD þ STZ (x indicates p < 0.05 versus baseline by Kruskal-Wallis test, * indicates p < 0.05 in HFD þ STZ mice vs. WT mice by Mann-Whitney U test; data are mean ± SEM; n ¼ 4 per group).

Signaling Technology), Sirt6 (ab62739) (abcam). 2.7. Metabolomics For Metabolomics analysis the serum of non-diabetic WT mice subjected to LAD ligation and ADV-injection (FoxO1-WT vs. FoxO1-KR) was collected in the fasted state 3 weeks after LAD ligation (each group n ¼ 5). Metabolomics assays were performed in service at Metabolon Laboratory and are described in detail in the supplementary data. 2.8. Statistical analysis Results of the experimental studies are mean ± SEM. Statistical analyses were performed with GraphPAD Prism 6.0 or by Metabolon® using the unpaired Student's t-test if values followed Gaussian distribution or otherwise Mann-Whitney U test or Kruskal-Wallis test. Values of p < 0.05 were considered statistically significant. Heatmap of metabolomics analysis was created using MetaboAnalyst 3.0 (http://www.metaboanalyst.ca) [35]. 3. Results 3.1. Loss of total cardiac FoxO1 after ischemia in non-diabetic and diabetic mice To investigate chronological FoxO1 expression after cardiac ischemia, experimental myocardial infarction using ligation of LAD was carried out in WT and HFD þ STZ mice. Animals were sacrificed at different time points: baseline (sham operation), 1 h, 24 h and 7

days after surgery (each group n ¼ 4). The area of myocardial infarction was used for further analyses. Western blots were performed from nuclear and cytosolic extracts. Since we found several commonly used loading controls such as actin (Fig. 1a), GAPDH and tubulin to be modulated, Ponceau S staining was used as loading control as previously described (Fig. 1a) [36,37]. Our results show a massive loss of total cardiac FoxO1 after ischemia in both WT and HFD þ STZ animals (Fig. 1a, b). 1 h after LAD-ligation we noticed a significant reduction of total FoxO1 only in the cytoplasm of both animal groups (Fig. 1a, b), whereas 24 h following myocardial ischemia we found a massive decrease of nuclear and cytoplasmic FoxO1 expression (24 h vs. baseline; nucleus: WT:
77 ± 7%, HFD þ STZ:
84 ± 6%; cytoplasm: WT:
91 ± 13%, HFD þ STZ:
88 ± 11%; n ¼ 4 per group, all p < 0.05, Fig. 1a, b). After one week, we observed that total FoxO1 was in part recovered with a higher extend in the nucleus compared to the cytoplasm (7 days vs. baseline; nucleus: WT:
32 ± 6%, HFD þ STZ:
31 ± 6%; cytoplasm: WT:
59 ± 13%, HFD þ STZ:
81 ± 14%; n ¼ 4 per group, all p < 0.05, Fig. 1a, b). 7 days after myocardial ischemia was further the only time point where we observed a difference between WT and HFD þ STZ mice and that only in the cytoplasm (HFD þ STZ vs. WT:
54 ± 13%, n ¼ 4 per group, p < 0.05, Fig. 1a, b). 3.2. Cardiac FoxO1 is posttranslationally modified by diabetes and ischemia Since we did not observe fundamental changes of total cardiac FoxO1 between WT and HFD þ STZ mice, we next investigated the

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

151

cytoplasm

nucleus

a P-FoxO1 Ac-FoxO1

Total FoxO1 P-FoxO1

Ac-FoxO1 Total FoxO1 WT

HFD+STZ

WT

baseline

HFD+STZ

WT

1h

b

WT

HFD+STZ

7 days

c cytoplasm

*§

1.0

§ 0.5

*§

§ §

§

* 0.0

baseline

1h

nucleus

24h

7 days

WT

§

baseline

HFD+STZ

§ § 1h

24h

7 days

§ vs. baseline

Acetylated / total FoxO1

nucleus

Phosphorylated / total FoxO1

HFD+STZ

24h

cytoplasm

* §

1.0

*

§

0.5

* 0.0

baseline

* 1h

§

§ § 24h

7 days

* baseline

* 1h

24h

7 days

* HFD+STZ vs. WT

Fig. 2. Diabetes and ischemia contribute to posttranslational modulation of FoxO1 in the heart. a) Representative Western blots of nuclear/cytosolic extracts of myocardial infarction (MI) area show alterations in posttranslational modulation of FoxO1 between non-diabetic mice (WT) and diabetic mice (HFD þ STZ) (P-FoxO1: phosphorylated FoxO1; Ac-FoxO1: acetylated FoxO1). b) and c) Quantitative analysis of Western blotting showing posttranslational modulation [b): phosphorylation; c) acetylation] of FoxO1 between HFD þ STZ vs. WT mice in the MI area at baseline (sham operation), 1 h, 24 h and 7 days after MI (x indicates p < 0.05 versus baseline by Kruskal-Wallis test, * indicates p < 0.05 in HFD þ STZ mice vs. WT mice by Mann-Whitney U test; data are mean ± SEM; n ¼ 4 per group).

posttranslational modulation of FoxO1. Ratios of phosphorylated and acetylated FoxO1 (normalized to total FoxO1) were analyzed in the infarcted hearts of WT and HFD þ STZ animals at different time points. At baseline ratios of phosphorylated FoxO1 differed between WT and HFD þ STZ animals only in the cytoplasm, but not in the nucleus (Fig. 2a, b). After 1 h FoxO1 phosphorylation decreased in both groups, but to a higher extend in the nucleus of WT animals (p < 0.05, n ¼ 4 per group, Fig. 2a, b). After 24 h in both groups phosphorylation was significantly reduced compared to baseline levels in both nucleus and cytoplasm (all p < 0.05, n ¼ 4 per group, Fig. 2a, b). After one week, WT animals exhibited similar FoxO1 phosphorylation compared to baseline, whereas analysis in HFD þ STZ animals revealed a slight reduction of nuclear FoxO1 phosphorylation (p < 0.05, n ¼ 4 per group, Fig. 2a, b) and a remarkable increase of FoxO1 phosphorylation in the cytoplasm (p < 0.05, n ¼ 4 per group, Fig. 2a, b). Investigating FoxO1 acetylation, we observed a notable increase in hearts of HFD þ STZ animals compared to WT littermates in both nucleus and cytosol at baseline as well as 1 h and 24 h after myocardial infarction (p < 0.05, n ¼ 4 per group, Fig. 2a, c). 24 h after LAD ligation FoxO1 acetylation ratio was about three times higher in the nucleus of HFD þ STZ animals compared to baseline levels and also increased to a similar extend compared to WT littermates at the same time point (p < 0.05, n ¼ 4 per group, Fig. 2a, c). In the WT group FoxO1 acetylation also increased shortly after cardiac ischemia, but to a lower extend (p < 0.05, n ¼ 4 per group,

Fig. 2a, c). After 7 days, in both groups acetylation levels were found to be the lowest and the intergroup difference was lost.

3.3. Overexpression of constitutively deacetylated FoxO1 reveals cardioprotective effects compared to a wild type variant of FoxO1 Since we found that FoxO1 is highly acetylated in the hearts of diabetic animals compared to non-diabetic animals and increased by ischemia, we tested the effects of posttranscriptional modulation of FoxO1 on the cardiac outcome using a model of postischemic heart failure. Following LAD ligation we injected an adenoviral vector expressing a constitutively deacetylated variant of FoxO1 (FoxO1-KR) or an adenoviral vector expressing a wild type form of FoxO1 (FoxO1-WT) serving as control. After 3 weeks, heart function was evaluated using Millar catheter prior to scarification. Millar catheter showed significantly improved heart function in the FoxO1-KR group shown by increased left ventricular developed pressure (LVDP), rate pressure product (LVDP  heart rate (HR)) and maximum rate of left ventricular pressure rise (DpDtmax) as well as decreased maximum rate of left ventricular pressure decline (
DpDtmin) (LVDP: þ29.1 ± 9.4 mmHg, LVDP  HR: þ10,934 ± 3977 mmHg  beats/min, DpDtmax: þ2291 ± 684 mmHg/s,
DpDtmin:
1595 ± 555 mmHg/ s, all p < 0.05; n ¼ 5e7 per group; Fig. 3c). Histology analysis of infarcted hearts of KR injected animals revealed significantly reduced infarction size compared to WT injected animals (
13.4 ± 3.5% of ventricle, p < 0.01; n ¼ 5 per

152

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

b

a Gomori

infarction size

H&E

% of ventricle

80 60

20 0

FoxO1-WT

**

40

FoxO1 WT

FoxO1 KR

FoxO1-KR

c rate pressure product

40 20

FoxO1 WT

FoxO1 KR

20000

10000

0

FoxO1 WT

FoxO1 KR

-dP/dtmin

6000

0

** mmHg/s

mmHg

60

mmHg*beats/min

*

80

0

*

30000

100

dP/dtmax

4000

mmHg/s

LVDP

2000

0

FoxO1 WT

FoxO1 KR

-2000

-4000

-6000

* FoxO1 WT

FoxO1 KR

Fig. 3. Constitutively deacetylated FoxO1 reveals protective effects in postischemic heart failure compared to a wild type variant of FoxO1. a) Representative sections of infarcted hearts of FoxO1-WT and FoxO1-KR injected non-diabetic animals (Gomori and hematoxylin and eosin staining (H&E)). b) Quantification of infarction size reveals significantly reduced infarction size in FoxO1-KR injected mice compared to FoxO1-WT injected littermates (**p < 0.01 by unpaired Student's t-test; data are mean ± SEM; n ¼ 5 per group). c) Millar catheter prior to sacrifice 3 weeks after LAD ligation shows improved heart function in the FoxO1-KR group (*p < 0.05, **p < 0.01 by unpaired Student's t-test; data are mean ± SEM; n ¼ 5e7 per group)(LVDP: left ventricular developed pressure; LVDP*HR: rate pressure product; DpDtmax: maximum rate of left ventricular pressure rise; -DpDtmin: maximum rate of left ventricular pressure decline).

group; Fig. 3a, b).

3.5. Constitutively deacetylated FoxO1 exhibits protective effects also in diabetic mice

3.4. Metabolomics analysis reveals changes in muscle breakdown, collagen/elastin turnover and energy metabolism

In a final step, we repeated the prior experiment in HFD þ STZ mice as previously described (Supp. Fig. S1). We injected FoxO1-KR and FoxO1-WT overexpressing ADV during MI procedure. Heart function and infarction size were assessed after 3 weeks. Millar catheter revealed similar results as seen in non-diabetic mice with greater left LVDP, LVDP  HR and DpDtmax, respectively reduced
DpDtmin (LVDP: þ10.9 ± 3.6 mmHg, LVDP  HR: þ7730 ± 1930 mmHg  beats/min, DpDtmax: þ1240 ± 391 mmHg/s,
DpDtmin:
747 ± 285 mmHg/s, all p < 0.05; n ¼ 7e8 per group; Fig. 5c). Histological analysis again showed reduced myocardial infarction size in the FoxO1-KR group compared to FoxO1-WT (
17.6 ± 3.7% of ventricle, p < 0.05; n ¼ 4e5 per group; Fig. 5a, b).

To reveal the affected pathways that are related to the improved cardiac phenotype in the FoxO1-KR group, we performed metabolomics analysis of serum collected from FoxO1-WT vs. FoxO1-KR injected animals (each group n ¼ 5). Several metabolites were found to be significantly altered in FoxO1-KR group compared to FoxO1-WT (Fig. 4a, b): We found alterations in metabolites associated to collagen/elastin turnover demonstrated by increase of trans-4-hydroxyproline (p < 0.05), muscle breakdown shown by decrease of 3-methylhistidine (p < 0.05) as well as altered energy metabolism revealed by decrease of 2-hydroxystearate (p < 0.05) and reduction of metabolites of the branched chain amino acid (BCAA) breakdown such as propionylcarnitine (p < 0.05) and a trend towards reduction of Nacetylisoleucin, valin and 3-methylcrotonylglycine (all 0.05 < p < 0.10).

4. Discussion The present study gives insights in cardiac FoxO1 regulation

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

a

collagen/elastin turnover

153

b

trans-4-hydroxyproline

fold change

3

FoxO1-KR

*

2

glycolate (hydroxyacetate)

1

propionylcarnitine

0

FoxO1 WT

3-methylhistidine

FoxO1 KR

uridine 2-hydroxystearate

muscle break down

energy metabolism

3-methylhistidine 1.4

7-Hoca

1.5

1.2

fold change

fold change

caprylate (8:0)

2-hydroxystearate

*

1.0 0.8

methyl-beta-glucopyranoside

*

1.0

2-aminooctanoate

0.5

gamma-glutamylglutamine 0.6

FoxO1 WT

0.0

FoxO1 KR

FoxO1 WT

FoxO1 KR

trans-4-hydroxyproline

BCAA catabolism

mead acid (20:3n9)

propionylcarnitine

2’-deoxycytidine

fold change

3

glucose 2

*

beta-hydroxypyruvate

1

0

class: FoxO1 WT

FoxO1 KR

FoxO1-WT FoxO1-KR

2

1

0

-1

-2

Fig. 4. Metabolomics serum analysis reveals alterations in muscle breakdown, collagen/elastin turnover and energy metabolism in FoxO1-KR-injected mice. a) Metabolomic analysis of the serum of non-diabetic mice 3 weeks after LAD ligation with injection of adenovirus overexpressing either a wild type variant of FoxO1 (FoxO1-WT) or a constitutively deacetylated variant of FoxO1 (FoxO1-KR) shows alterations in metabolites associated with collagen/elastin metabolism, energy metabolism, muscle breakdown and branchedchain amino acid metabolism (*p < 0.05, **p < 0.01 by unpaired Student's t-test; n ¼ 5 per group). b) Heatmap representing top 15 metabolites between FoxO1-KR and FoxO1WT by unpaired Student's t-test (n ¼ 5 per group).

after myocardial ischemia and provides evidence that posttranslational modulation of FoxO1 is involved in cardiac remodeling in a model of postischemic heart failure. We demonstrated that in the area of myocardial infarction total FoxO1 is massively downregulated and only partly recovers one week after LAD ligation. Further, we could show that posttranslational modification of cardiac FoxO1 differs between diabetic and non-diabetic mice at baseline levels and shortly after ischemia. Particularly, we found that FoxO1 acetylation ratio is strongly increased in hearts of diabetic mice compared to wild type littermates, but also rises after induction of ischemia in both animal groups. Our findings are in agreement to those of other groups who analyzed posttranslational modulation of FoxO1 in non-infarcted hearts of diabetic and non-diabetic mice, i.e. Wang et al. found that FoxO1 is highly acetylated in hearts of diabetic compared to wild type mice [32]. Activation of FoxO1 has been shown to be either protective [12,25] or detrimental [28e30] depending on the conditions and experimental setup that were used. While wildtype FoxO1 shuttles between nucleus and cytoplasm, it has been proposed that acetylated FoxO1 is predominantly nuclear [13]. However, several groups suggested that acetylation of FoxO1 reduces transcriptional activities on target genes [8,13,15], but has also been shown to exhibit effects independently from altering DNA binding affinity [16]. To elucidate if FoxO1 acetylation has positive or unfavorable effects in a model of postischemic heart failure, we overexpressed a

constitutively deacetylated variant FoxO1, which exhibited cardioprotective effects compared to a wild type variant of FoxO1 in both non-diabetic and diabetic mice. Our results therefore indicate that 1) active posttranslational modulation of FoxO1 is capable of altering cardiac remodeling and 2) that increased acetylation of FoxO1 in diabetic hearts might contribute to a worse outcome after MI in patients with T2D. In vivo FoxO1 is deacetylated by Sirtuin1 (SIRT1), a class III NAD þ -dependent histone deacetylase, but also by other sirtuins such as SIRT2 and SIRT3 [15]. In cardiomyocytes of patients with severe heart failure SIRT1 was found to be downregulated [38]. In accordance to our findings several groups reported cardioprotective effects of SIRT1 activation [23,32,39e41]. However, SIRT1 interacts with multiple other proteins [15] and the compounds that have frequently been used to activate SIRT1 exhibit many off-target activities [42]. In the adjacent area to myocardial infarction, we performed Western blots of several sirtuins (Suppl. Fig. S3). SIRT1 expression increased after MI, but no difference was observed between diabetic and non-diabetic mice. 1 h after MI, SIRT2 and SIRT3 were both significantly reduced in diabetic hearts (compared to non-diabetic), which might contribute to increase of FoxO1 acetylation. Another limitation of our study is that we did not fully uncover the exact molecular pathway of how constitutively deacetylated FoxO1 unveils its cardioprotective effects. Our metabolomics analysis between the FoxO1-KR and FoxO1-WT group may give a

154

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

b Gomori

a

infarction size

% of ventricle

100 80

40 20

H&E

0

FoxO1-WT

**

60

FoxO1 WT

FoxO1 KR

FoxO1-KR

c

20

FoxO1 WT

FoxO1 KR

20000

10000

0

FoxO1 WT

FoxO1 KR

-dP/dtmin

6000

mmHg/s

mmHg

40

mmHg*beats/min

*

60

0

**

30000

100 80

dP/dtmax

rate pressure product

0

**

4000

mmHg/s

LVDP

2000

0

FoxO1 WT

FoxO1 KR

-2000

*

-4000

-6000

FoxO1 WT

FoxO1 KR

Fig. 5. Constitutively deacetylated FoxO1 reveals cardioprotective effects in diabetic mice. a) Representative sections of infarcted hearts of FoxO1-WT and FoxO1-KR injected diabetic HFD þ STZ mice (Gomori and hematoxylin and eosin staining(H&E)). b) Quantification of infarction size as percent of total left ventricle circumference confirms reduction of infarction size in FoxO1-KR group compared to FoxO1-WT in diabetic HFD þ STZ animals (**p < 0.01 by unpaired Student's t-test; data are mean ± SEM; n ¼ 4e5 per group). c) Millar catheter 3 weeks after experimental myocardial infarction reveals better cardiac function in FoxO1-KR compared to FoxO1-WT injected mice as shown by increase of left ventricular developed pressure (LVDP), rate pressure product (LVDP*HR) and maximum rate of left ventricular pressure rise (DpDtmax) as well as decreased maximum rate of left ventricular pressure decline (
DpDtmin) (*p < 0.05, **p < 0.01 by unpaired Student's t-test; data are mean ± SEM; n ¼ 7e8 per group).

hint to underlying pathways. We found variations in markers of muscle breakdown (3-methylhistidine) [43,44], BCAA metabolism (propionylcarnitine, valin, N-acetylisoleucin, 3methylcrotonylglycin) [45], collagen/elastin metabolism (trans-4hydroxyprolin) [46,47] and energy metabolism (2hydroxystearate) [48]. Similar alterations in BCAA metabolism have been described in mice subjected to two different models of heart failure: myocardial infarction and transverse aortic constriction [49], thus it remains to be elucidated if FoxO1-KR has a direct molecular effect on cardiac muscle breakdown or alterations in metabolites only reflect an improved heart function. 2-hydroxystearate has been associated to incomplete fatty acid oxidation in the heart [48] and was reduced in the FoxO1-KR group. Mice that are homozygous for the constitutively deacetylated FoxO1 allele (FoxO1 KR/KR) present an altered hepatic glucose metabolism and use lipids as preferred substrate [50]. Hence, variations in substrate usage during myocardial ischemia might account for the superior outcome found in the FoxO1-KR group. Along with this finding, our metabolomics analysis revealed that glucose levels were slightly higher in the FoxO1-KR group (Fig. 4b).

This finding might reflect a shift from glucose to fatty acid oxidation in the hearts of FoxO1-KR animals. However, we cannot exclude that other tissues such as liver are infected by ADV and therefore contribute to changes in metabolomics as well as the improved cardiac phenotype through a liver-heart axis. Further, future studies must reveal, if other processes relevant for cardiac remodeling such as autophagy [16,51] and apoptosis [40], which have been previously associated to FoxO1 acetylation, are involved. 5. Conclusion In conclusion, we could show that after ischemia posttranslational modification of cardiac FoxO1 differs between nondiabetic and diabetic mice, and that active posttranslational modulation of FoxO1 is able to alter the cardiac outcome in a mouse model of postischemic heart failure. Our findings therefore suggest that posttranslational modification of FoxO1 contribute to cardiac remodeling processes after ischemia and that active modulation of FoxO1 might be an option to target remodeling processes in postischemic heart failure.

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

Author contributions MF designed the study and the experiments; B.K., R.S., L.A., M.M., R.M. carried out the experimental work; B.K., R.S. and M.F. wrote the manuscript; all authors reviewed the manuscript.

[15]

[16]

Conflict of interest The authors have no conflict of interest to disclose regarding the content of this manuscript.

[17]

[18]

Acknowledgements We thank Dr. Domenico Accili (Columbia University, New York, USA) for the gift of FoxO1 adenovirus. M.F. laboratory is funded by Fondazione Roma NCDs call year  Italiana di Diabetologia year 2013, FP-7 2014, Progetto Societa EURHYTHDIA (contract grant agreement no. 278397), FP7-FLORINASH (contract grant agreement no. 241913), PRIN 201223BJ89E and Associazione Italiana per la Ricerca sul Cancro (Grant AIRC IG13163). B.K. and R.S. are supported by a grant from the Deutsche Herzstiftung (DHS) and a PhD fellowship from the University of Rome Tor Vergata.

[19]

[20]

[21]

[22]

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.atherosclerosis.2016.04.001. References [1] V.L. Roger, Epidemiology of heart failure, Circ. Res. 113 (2013) 646e659, http://dx.doi.org/10.1161/CIRCRESAHA.113.300268. [2] K.J. Harjai, G.W. Stone, J. Boura, L. Mattos, H. Chandra, D. Cox, et al., Comparison of outcomes of diabetic and nondiabetic patients undergoing primary angioplasty for acute myocardial infarction, Am. J. Cardiol. 91 (2003) 1041e1045, http://dx.doi.org/10.1016/S0002-9149(03)00145-0. [3] T. Almdal, H. Scharling, J.S. Jensen, H. Vestergaard, The independent effect of type 2 diabetes mellitus on ischemic heart disease, stroke, and death: a population-based study of 13,000 men and women with 20 years of followup, Arch. Intern. Med. 164 (2004) 1422e1426, http://dx.doi.org/10.1001/ archinte.164.13.1422. [4] B. Kappel, N. Marx, M. Federici, Oral hypoglycemic agents and the heart failure conundrum: Lessons from and for outcome trials, Nutr. Metab. Cardiovasc. Dis. 25 (2015) 697e705, http://dx.doi.org/10.1016/j.numecd.2015.06.006. [5] W.C. Stanley, G.D. Lopaschuk, J.G. McCormack, Regulation of energy substrate metabolism in the diabetic heart, Cardiovasc. Res. 34 (1997) 25e33, http:// dx.doi.org/10.1016/S0008-6363(97)00047-3. [6] N. Fillmore, J. Mori, G.D. Lopaschuk, Mitochondrial fatty acid oxidation alterations in heart failure, ischaemic heart disease and diabetic cardiomyopathy, Br. J. Pharmacol. 171 (2014) 2080e2090, http://dx.doi.org/10.1111/ bph.12475. [7] A.R. Wende, Post-translational modifications of the cardiac proteome in diabetes and heart failure, Proteom. Clin. Appl. 10 (2016) 25e38, http:// dx.doi.org/10.1002/prca.201500052. [8] D. Accili, K.C. Arden, FoxOs at the crossroads of cellular metabolism, differentiation, and transformation, Cell. 117 (2004) 421e426, http://dx.doi.org/ 10.1016/S0092-8674(04)00452-0. [9] A. Sengupta, J.D. Molkentin, K.E. Yutzey, FoxO transcription factors promote autophagy in cardiomyocytes, J. Biol. Chem. 284 (2009) 28319e28331, http:// dx.doi.org/10.1074/jbc.M109.024406. [10] J. Nakae, M. Oki, Y. Cao, The FoxO transcription factors and metabolic regulation, FEBS Lett. 582 (2008) 54e67, http://dx.doi.org/10.1016/ j.febslet.2007.11.025. [11] A. Barthel, D. Schmoll, T.G. Unterman, FoxO proteins in insulin action and metabolism, Trends Endocrinol. Metab. 16 (2005) 183e189, http://dx.doi.org/ 10.1016/j.tem.2005.03.010. [12] P. Puthanveetil, A. Wan, B. Rodrigues, FoxO1 is crucial for sustaining cardiomyocyte metabolism and cell survival, Cardiovasc. Res. 97 (2013) 393e403, http://dx.doi.org/10.1093/cvr/cvs426. [13] L. Qiang, A.S. Banks, D. Accili, Uncoupling of acetylation from phosphorylation regulates FoxO1 function independent of its subcellular localization, J. Biol. Chem. 285 (2010) 27396e27401, http://dx.doi.org/10.1074/jbc.M110.140228. [14] D. Frescas, L. Valenti, D. Accili, Nuclear trapping of the forkhead transcription

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

155

factor FoxO1 via sirt-dependent deacetylation promotes expression of glucogenetic genes, J. Biol. Chem. 280 (2005) 20589e20595, http://dx.doi.org/ 10.1074/jbc.M412357200. H. Daitoku, J.I. Sakamaki, A. Fukamizu, Regulation of FoxO transcription factors by acetylation and protein-protein interactions, Biochim. Biophys. Acta - Mol. Cell Res. 1813 (2011) 1954e1960, http://dx.doi.org/10.1016/ j.bbamcr.2011.03.001. Y. Zhao, J. Yang, W. Liao, X. Liu, H. Zhang, S. Wang, et al., Cytosolic FoxO1 is essential for the induction of autophagy and tumour suppressor activity, Nat. Cell Biol. 12 (2010) 665e675, http://dx.doi.org/10.1038/ncb2069. T. Furuyama, K. Kitayama, Y. Shimoda, M. Ogawa, K. Sone, K. Yoshida-Araki, et al., Abnormal angiogenesis in Foxo1 (Fkhr)-deficient mice, J. Biol. Chem. 279 (2004) 34741e34749, http://dx.doi.org/10.1074/jbc.M314214200. L. Qiang, K. Tsuchiya, J.-Y. Kim-Muller, H. V Lin, C. Welch, D. Accili, Increased atherosclerosis and endothelial dysfunction in mice bearing constitutively deacetylated alleles of Foxo1 gene, J. Biol. Chem. 287 (2012) 13944e13951, http://dx.doi.org/10.1074/jbc.M111.332767. K. Tsuchiya, J. Tanaka, Y. Shuiqing, C.L. Welch, R.A. DePinho, I. Tabas, et al., FoxOs integrate pleiotropic actions of insulin in vascular endothelium to protect mice from atherosclerosis, Cell Metab. 15 (2012) 372e381, http:// dx.doi.org/10.1016/j.cmet.2012.01.018. K. Tsuchiya, A.S. Banks, C.-P. Liang, I. Tabas, A.R. Tall, D. Accili, Homozygosity for an allele encoding deacetylated FoxO1 protects macrophages from cholesterol-induced inflammation without increasing apoptosis, Arterioscler. Thromb. Vasc. Biol. 31 (2011) 2920e2928, http://dx.doi.org/10.1161/ ATVBAHA.110.219477. J. Tanaka, L. Qiang, A.S. Banks, C.L. Welch, M. Matsumoto, T. Kitamura, et al., Foxo1 links hyperglycemia to LDL oxidation and endothelial nitric oxide synthase dysfunction in vascular endothelial cells, Diabetes 58 (2009) 2344e2354, http://dx.doi.org/10.2337/db09-0167. R. Menghini, V. Casagrande, M. Cardellini, M. Ballanti, F. Davato, I. Cardolini, et al., FoxO1 regulates asymmetric dimethylarginine via downregulation of dimethylaminohydrolase 1 in human endothelial cells and subjects with atherosclerosis, Atherosclerosis 242 (2015) 230e235, http://dx.doi.org/ 10.1016/j.atherosclerosis.2015.07.026. C.-J. Chen, W. Yu, Y.-C. Fu, X. Wang, J.-L. Li, W. Wang, Resveratrol protects cardiomyocytes from hypoxia-induced apoptosis through the SIRT1-FoxO1 pathway, Biochem. Biophys. Res. Commun. 378 (2009) 389e393, http:// dx.doi.org/10.1016/j.bbrc.2008.11.110. J.B. Morris, B. Kenney, H. Huynh, E. a Woodcock, Regulation of the proapoptotic factor FOXO1 (FKHR) in cardiomyocytes by growth factors and alpha1adrenergic agonists, Endocrinology 146 (2005) 4370e4376, http://dx.doi.org/ 10.1210/en.2005-0162. Y. Ning, Z. Li, Z. Qiu, FOXO1 silence aggravates oxidative stress-promoted apoptosis in cardiomyocytes by reducing autophagy, J. Toxicol. Sci. 40 (2015) 637e645, http://dx.doi.org/10.2131/jts.40.637. H.J. Evans-Anderson, C.M. Alfieri, K.E. Yutzey, Regulation of cardiomyocyte proliferation and myocardial growth during development by FOXO transcription factors, Circ. Res. 102 (2008) 686e694, http://dx.doi.org/10.1161/ CIRCRESAHA.107.163428. J. Liu, Y. Tang, Z. Feng, C. Hou, H. Wang, J. Yan, et al., Acetylated FoxO1 mediates high-glucose induced autophagy in H9c2 cardiomyoblasts: regulation by a polyphenol -(-)-epigallocatechin-3-gallate, Metabolism 63 (2014) 1314e1323, http://dx.doi.org/10.1016/j.metabol.2014.06.012. P. Puthanveetil, Y. Wang, D. Zhang, F. Wang, M.S. Kim, S. Innis, et al., Cardiac triglyceride accumulation following acute lipid excess occurs through activation of a FoxO1eiNOSeCD36 pathway, Free Radic. Biol. Med. 51 (2011) 352e363, http://dx.doi.org/10.1016/j.freeradbiomed.2011.04.009. P.K. Battiprolu, B. Hojayev, N. Jiang, Z. V Wang, X. Luo, M. Iglewski, et al., Metabolic stress-induced activation of FoxO1 triggers diabetic cardiomyopathy in mice, J. Clin. Investig. 122 (2012) 1109e1118, http://dx.doi.org/ 10.1172/JCI60329. Y. Qi, Q. Zhu, K. Zhang, C. Thomas, Y. Wu, R. Kumar, et al., Activation of Foxo1 by insulin resistance promotes cardiac dysfunction and -myosin heavy chain gene expression, Circ. Hear. Fail. 8 (2014) 198e208, http://dx.doi.org/10.1161/ CIRCHEARTFAILURE.114.001457. J. Mu, J. Woods, Y.-P. Zhou, R.S. Roy, Z. Li, E. Zycband, et al., Chronic inhibition of dipeptidyl peptidase-4 with a sitagliptin analog preserves pancreatic betacell mass and function in a rodent model of type 2 diabetes, Diabetes 55 (2006) 1695e1704, http://dx.doi.org/10.2337/db05-1602. B. Wang, Q. Yang, Y. Sun, Y. Xing, Y. Wang, X. Lu, et al., Resveratrol-enhanced autophagic flux ameliorates myocardial oxidative stress injury in diabetic mice, J. Cell. Mol. Med. 18 (2014) 1599e1611, http://dx.doi.org/10.1111/ jcmm.12312. €hr, B.A. Kappel, D. Carnevale, M. Cavalera, M. Mavilio, I. Arisi, et al., R. Sto TIMP3 interplays with apelin to regulate cardiovascular metabolism in hypercholesterolemic mice, Mol. Metab. 4 (2015) 741e752, http://dx.doi.org/ 10.1016/j.molmet.2015.07.007. K.M. Shioura, D.L. Geenen, P.H. Goldspink, Assessment of cardiac function with the pressure-volume conductance system following myocardial infarction in mice, Am. J. Physiol. Heart Circ. Physiol. 293 (2007) H2870eH2877, http://dx.doi.org/10.1152/ajpheart.00585.2007. J. Xia, I.V. Sinelnikov, B. Han, D.S. Wishart, MetaboAnalyst 3.0emaking metabolomics more meaningful, Nucleic Acids Res. 43 (2015) W251eW257, http://dx.doi.org/10.1093/nar/gkv380.

156

B.A. Kappel et al. / Atherosclerosis 249 (2016) 148e156

[36] J.E. Gilda, A.V. Gomes, Stain-Free total protein staining is a superior loading control to b-actin for Western blots, Anal. Biochem. 440 (2013) 186e188, http://dx.doi.org/10.1016/j.ab.2013.05.027. [37] a. V. Gomes, D.S. Waddell, R. Siu, M. Stein, S. Dewey, J.D. Furlow, et al., Upregulation of proteasome activity in muscle RING finger 1-null mice following denervation, FASEB J. 26 (2012) 2986e2999, http://dx.doi.org/ 10.1096/fj.12-204495. [38] T.-M. Lu, J.-Y. Tsai, Y.-C. Chen, C.-Y. Huang, H.-L. Hsu, C.-F. Weng, et al., Downregulation of Sirt1 as aging change in advanced heart failure, J. Biomed. Sci. 21 (2014) 57, http://dx.doi.org/10.1186/1423-0127-21-57. [39] C.-P. Hsu, P. Zhai, T. Yamamoto, Y. Maejima, S. Matsushima, N. Hariharan, et al., Silent information regulator 1 protects the heart from ischemia/reperfusion, Circulation 122 (2010) 2170e2182, http://dx.doi.org/10.1161/ CIRCULATIONAHA.110.958033. [40] T.K. Sin, A.P. Yu, B.Y. Yung, S.P. Yip, L.W. Chan, C.S. Wong, et al., Modulating effect of SIRT1 activation induced by resveratrol on Foxo1-associated apoptotic signalling in senescent heart, J. Physiol. 592 (2014) 2535e2548, http://dx.doi.org/10.1113/jphysiol.2014.271387. [41] A. Cattelan, G. Ceolotto, S. Bova, M. Albiero, M. Kuppusamy, S. De Martin, et al., NAD(þ)-dependent SIRT1 deactivation has a key role on ischemiareperfusion-induced apoptosis, Vasc. Pharmacol. 70 (2015) 35e44, http:// dx.doi.org/10.1016/j.vph.2015.02.004. [42] M. Pacholec, J.E. Bleasdale, B. Chrunyk, D. Cunningham, D. Flynn, R.S. Garofalo, et al., SRT1720, SRT2183, SRT1460, and resveratrol are not direct activators of SIRT1, J. Biol. Chem. 285 (2010) 8340e8351, http://dx.doi.org/10.1074/ jbc.M109.088682. [43] H.C. Lukaski, J. Mendez, E.R. Buskirk, S.H. Cohn, Relationship between endogenous 3-methylhistidine excretion and body composition, Am. J. Physiol. 240 (1981) E302eE307. http://www.ncbi.nlm.nih.gov/pubmed/ 7212061 (accessed November 25, 2015).

[44] S.M. Garvey, J.E. Dugle, A.D. Kennedy, J.E. McDunn, W. Kline, L. Guo, et al., Metabolomic profiling reveals severe skeletal muscle group-specific perturbations of metabolism in aged FBN rats, Biogerontology 15 (2014) 217e232, http://dx.doi.org/10.1007/s10522-014-9492-5. [45] E.D. Bruder, H. Raff, Cardiac and plasma lipid profiles in response to acute hypoxia in neonatal and young adult rats, Lipids Health Dis. 9 (2010) 3, http:// dx.doi.org/10.1186/1476-511X-9-3. [46] P.C. Kelleher, Urinary excretion of hydroxyproline, hydroxylysine and hydroxylysine glycosides by patients with Paget's disease of bone and carcinoma with metastases in bone, Clin. Chim. Acta 92 (1979) 373e379. http:// www.ncbi.nlm.nih.gov/pubmed/436278 (accessed November 25, 2015). [47] A. de C. Nogueira, R.G.S. Vale, A.L.M. Gomes, E.H.M. Dantas, The effect of muscle actions on the level of connective tissue damage, Res. Sports Med. 19 (2011) 259e270, http://dx.doi.org/10.1080/15438627.2011.608046. [48] K.H. Moore, J.F. Radloff, F.E. Hull, C.C. Sweeley, Incomplete fatty acid oxidation by ischemic heart: beta-hydroxy fatty acid production, Am. J. Physiol. 239 (1980) H257eH265 (accessed November 25, 2015), http://www.ncbi.nlm.nih. gov/pubmed/7406063. [49] B.E. Sansbury, A.M. DeMartino, Z. Xie, A.C. Brooks, R.E. Brainard, L.J. Watson, et al., Metabolomic Analysis of Pressure-Overloaded and Infarcted Mouse Hearts CLINICAL PERSPECTIVE, Circ. Hear. Fail. 7 (2014) 634e642, http://dx.doi.org/ 10.1161/CIRCHEARTFAILURE.114.001151. [50] A.S. Banks, J.Y. Kim-Muller, T.L. Mastracci, N.M. Kofler, L. Qiang, R.A. Haeusler, et al., Dissociation of the glucose and lipid regulatory functions of FoxO1 by targeted knockin of acetylation-defective alleles in mice, Cell Metab. 14 (2011) 587e597, http://dx.doi.org/10.1016/j.cmet.2011.09.012. [51] N. Hariharan, Y. Maejima, J. Nakae, J. Paik, R.A. Depinho, J. Sadoshima, Deacetylation of FoxO by sirt1 plays an essential role in mediating starvationinduced autophagy in cardiac myocytes, Circ. Res. 107 (2010) 1470e1482, http://dx.doi.org/10.1161/CIRCRESAHA.110.227371.