Long-Term Caloric Restriction Improves Cardiac Function, Remodeling, Adrenergic Responsiveness and Sympathetic Innervation in a Model of Post-Ischemic Heart Failure

Abstract

Background

Caloric restriction (CR) has been described to have cardioprotective effects and improve functional outcomes in animal models and humans. Chronic ischemic heart failure (HF) is associated with reduced cardiac sympathetic innervation, dysfunctional β-adrenergic receptor (β-AR) signaling and decreased cardiac inotropic reserve. We tested the effects of a long-term CR diet, started late after myocardial infarction (MI) on cardiac function, sympathetic innervation and β-AR responsiveness in a rat model of post-ischemic HF.

Methods and Results

Adult male rats were randomly assigned to MI or Sham operation and four weeks later were further randomized to a one-year CR or normal diet. One year of CR resulted in a significant reduction in body weight, heart weight and heart weight/tibia length ratio when compared to normal diet in HF groups. At the end of the study period, echocardiography and histology revealed that HF animals under the CR diet had ameliorated left ventricular remodeling compared to HF rats fed with normal diet. Invasive hemodynamic showed a significant improvement of cardiac inotropic reserve in CR HF rats compared to HF normal diet animals. Importantly, CR dietary regimen was associated with a significant increase of cardiac sympathetic innervation and with normalized cardiac β-AR levels in HF rats when compared to HF rats on the standard diet.

Conclusions

We demonstrate, for the first time, that chronic CR, when started after HF established, can ameliorate cardiac dysfunction and improve inotropic reserve. At the molecular level, we find that chronic CR diet significantly improves sympathetic cardiac innervation and β-AR levels in failing myocardium.

INTRODUCTION

Despite improvements in diagnosis, management and treatment, heart failure (HF) still represents one of the greatest health care challenges worldwide for its high morbidity, mortality and economic impact^{1, 2}. This syndrome represents the final common clinical event of numerous cardiovascular (CV) diseases, with coronary artery disease-dependent myocardial infarction (MI) being its most common cause in the western countries¹. Despite some improvement in HF survival, the death rate remains high and current therapies are only able to slow disease progression. Hence, there is an urgency to identify new therapies to add to the HF therapeutic armamentarium in order to reduce the enormous mortality rate that this disease still poses.

Dietary interventions, including caloric restriction (CR), have been proven to counteract some CV risk factors, such as diabetes, hypertension and obesity that are known to promote atherosclerosis and the progression of cardiac dysfunction^3–7. Indeed, pre-clinical and clinical studies have demonstrated that CR exerts several therapeutic effects on atherosclerosis by reducing circulating cholesterol levels, systemic inflammation and vascular fibrosis^8–10. Furthermore, CR has also been shown to curb the maladaptive cardiac effects of hypertension in both genetic and surgical models of the disease, reducing the development of left ventricular (LV) hypertrophy and diastolic dysfunction^11–15.

Importantly, the beneficial actions of CR on global metabolism are widely recognized and include a reduction of visceral fat mass, fasting insulin levels, and insulin resistance in diabetic and obese patients⁷. Further, CR has been described to exert direct cardioprotective effects^{3, 12, 16–19}. Accordingly, intermittent fasting (IF), a specific CR regimen, protects the heart from ischemic injury, reduces cardiac remodeling and ameliorates LV function, when initiated before MI-induction¹⁷. In the same setting (IF started before MI), this CR regimen was associated with an increase in circulating adiponectin levels and a reduction in cardiac and systemic inflammation²⁰. Moreover, short-term (6-week-long) IF has been shown to improve LV remodeling and survival in a rat model of post-ischemic HF via pro-angiogenic and anti-apoptotic mechanisms²¹. Interestingly, the beneficial effects of CR have been also studied in obese patients with HF at preserved ejection fraction (HFpEF) in a randomized clinical-trial, where CR combined with exercise training was able to increase peak oxygen consumption²². Of note, all of the aforementioned studies have shown the beneficial effects of CR on the post-MI heart only in the short-term. From a clinical point-of-view, however, it is evident that for chronic diseases such HF, a long-term and sustained CR regimen is potentially warranted to improve outcomes. Hence, the effects of a long-standing CR diet on cardiac function and structure in a model of ischemic HF has not been properly investigated. Thus, in the present study we have tested the effects of a 12-month CR regimen in a rat post-MI model where CR was started after HF establishment.

METHODS

The data, analytic methods, and study materials will be made available from the corresponding authors to other researchers upon request for purposes of reproducing the results or replicating the procedure. Expanded Methods are available in the Supplemental Material.

Experimental design

All animal procedures and experiments were performed in accordance with the guidelines of the Institutional Animal Care Committee of University Federico II of Naples, Italy. Two-month-old, male Sprague-Dawley rats (n=44) were randomly assigned to different surgical groups: 34 rats underwent surgically-induced MI and 10 rats received Sham operation (Figure 1A). Mortality rate was ~23.5% at 4-weeks post-MI. Four weeks post-MI, a time point when cardiac dysfunction was established, HF rats underwent echocardiographic evaluation and were further randomized to a one-year CR (IF dietary regimen that consists in a complete fasting day followed by a normal feed day), or normal diet (ND, standard rat diet) (Figure 1A). Sham-operated rats followed a ND regimen throughout the study period. Thus, the final population that started the dietary protocol consisted of 3 groups: Sham-operated rats fed with ND (Sham, n=10), HF rats fed with ND (HF-ND, n=13) and HF rats fed with CR dietary regimen (HF-CR, n=13). Twelve months after the beginning of the feeding protocol, echocardiography and hemodynamic analysis were performed in all groups. Then, rats were euthanatized for histological and molecular analysis. An expanded Methods section appears in Supplemental Material.

Figure 1. Long-term caloric restriction decreases body weight and accordingly heart weight in ischemic HF.

A, Overall design of the 13-months study. B, Body weight (BW) changes during the study period for all groups at 0-, 6- and 12-months after the beginning of the dietary protocol. C, Representative picture of HF-ND (right side) and HF-CR rats (left side) at the end of the study period (12 months after the beginning of the dietary protocol). Measures of (D) Heart weight (HW) and (E) HW normalized to tibia length (TL). F, Graph showing the direct correlation between HW and BW in HF groups (HF-ND group: black spots; HF-CR group: red spots). n. at the end of the study: Sham=10; HF-ND=11; HF-CR=12. Data are presented as mean±SEM *p<0.0001 vs HF-ND, **p<0.01 vs HF-ND, ***p<0.05 vs HF-ND #p<0.01 vs Sham, ##p< 0.001 vs Sham, ^p<0.0001 vs Sham. Two- or one-way ANOVA and Bonferroni test were used between groups.

In vivo experimental procedures

MI, echocardiography and hemodynamic analysis were performed as described^23–25.

Plasma membrane levels of β-AR levels and Real-time PCR

Membrane proteins were isolated from LV samples²⁶ to evaluate protein levels of β₁-AR, β₂-AR and β₃-AR. Real-time PCR was performed as descrybed²⁶.

Histological sectioning and staining

Masson’s Trichrome staining was used to evaluate LV fibrosis while capillaries were detected by Lectin Bandeiraea simplicifolia I staining. Immunofluorescence technique was performed using antibodies against Vesicular acetylcholine transporter (VaChT) and Dopamine β-hydroxylase (DβH)²⁷.

Statistical analysis

All values in the text and figures are presented as means ± SEM. Statistical significance was determined by one-way or two-way ANOVA with Bonferroni post-hoc analysis or T-test as appropriate. All data were analyzed using GraphPad-Prism software version 7. Probabilities of 0.05 or less were considered to be statistically significant.

RESULTS

Effect of long-term caloric restriction on body weight and heart weight

As expected, at 4-weeks post-surgery, before the start of dietary intervention, the average body weight (BW) was similar among all study groups (Figure 1B and Table). Six months after the beginning of the CR diet protocol, average BW became significantly lower (p<0.001) in HF-CR rats compared to HF-ND rats (Figure 1B and Table). Of note, at this time-point HF-ND BW was significantly higher compared to Sham BW (p<0.05) (Figure 1B and Table). Twelve months after the beginning of different dietary regimens, HF-CR rats showed a significantly lower BW compared to HF-ND and Sham animals (respectively p<0.001 and p<0.0001) that was so evident on physical exam (Figure 1B–C). To analyze whether the different dietary protocols were able to influence the heart mass, we measured heart weight (HW), HW to tibia length (TL) ratio (HW/TL) and HW to BW ratio (HW/BW) at the end of the study period, in all groups. HW and HW/TL were significantly increased in HF-ND group compared to Sham group, consistent with an HF phenotype (p<0.001), while, HF-CR animals showed lower HW and HW/TL compared to HF-ND rats (p<0.01 and p<0.05, respectively) (Figure 1D–E and Table). As expected, HW/BW was increased in HF rats compared to Sham (p<0.001 vs HF-ND, p<0.0001 vs HF-CR) (Table) while HW/BW was not statistically different between HF-ND and HF-CR groups. Hence, we decided to analyze the correlation between HW and BW in all HF rats and we found that there was a direct correlation between these parameters (R²=0.6546, p<0.0001), suggesting that the decrease in BW in rats treated with a restricted dietary regimen was paralleled by a proportional decrease in HW (Figure 1F). The latter observation could also explain the reason why we did not find any difference between HF-ND and HF-CR groups in terms of HW/BW ratio (Table). Contrarily, HW/TL was significantly different between the two HF groups, as mentioned above, indicating that TL is not influenced by the dietetic regimen.

Table.

Long-term effects of caloric restriction on physical parameters and left ventricular function.

	Sham	HF-ND	HF-CR
Physical data
BW(g) basal	350±13.8	352±6.90	349±9.57
BW(g) 6-months	475±19.6	541±16.3*	440±12.7†
BW(g) 12-months	538±11.5	574±16.6	444±12.0‡,§
HW(g)	1.098±0.061	1.453±0.040‡	1.213±0.056∥
HW/BW(g/Kg)	2.045±0.113	2.540±0.061‡	2.724±0.077#
HW/TL(g/mm)	0.2407±0.014	0.3166±0.008‡	0.2695±0.011**
AW(g)	0.026±0.0029	0.037±0.0032*	0.028±0.0017**
AW/BW(g/Kg)	0.049±0.0056	0.066±0.0055	0.061±0.0032
AW/TL(mg/mm)	5.8±0.64	8.1±0.69*	6.1±0.35**
Echocardiography (end of the study)
EF%	68.6±3.09	30.7±0.92#	33.5±0.99#
LVIDd, mm	7.82±0.21	10.20±0.28#	9.14±0.30*,**
LVIDs, mm	5.04±0.44	8.63±0.23#	7.62±0.26#,**
LVVol-d, µl	329±19.8	595±34.6‡	470±35.4*,**
LVVol-s, µl	127±25.7	437±19.7#	336±36.2‡,**
LVAWd, mm	1.57±0.11	0.99±0.03#	0.97±0.02#
LVAWs, mm	2.54±0.28	1.03±0.02#	1.04±0.02#
LVPWd, mm	1.53±0.13	1.89±0.09*	1.85±0.06
LVPWs, mm	2.43±0.14	2.599±0.09	2.46±0.14
HR, bpm	349±10	315±9	319±10
Hemodynamic data, basal
HR, bpm	298±15	293±6	305±7
LV+dP/dt, mm Hg/s	7258±644	4135±299‡	5480±429*
LV−dP/dt, mm Hg/s	−7386±1191	−3128±173††	−4659±573*
LVEDP, mmHg	5.934±0.97	11.61±0.81‡	9.9±0.68*
LVESP, mmHg	130.1±4.07	99.32±5.63*	125.6±9.86**
Hemodynamic data, ISO (333 ng/Kg BW)
HR, bpm	361±9	352±8	359±6
LV dP/dt, mm Hg/s	13651±1814	6620±480‡	11017±657**
LV −dP/dt, mm Hg/s	−9870±1391	−3961±146‡	−6723±609*,**
LVEDP, mmHg	5.33±1.21	9.77±1.93	6.72±1.88
LVESP, mmHg	139.8±7.23	109.8±4.17††	142.5±4.95∥

Body weight (BW), heart weight (HW), HW/BW ratio and HW/tibia length ratio (HW/TL), adrenal gland weight (AW), AW/BW ratio and AW/TL ratio were measured in all groups. Left ventricular (LV) ejection fraction (EF), internal diameter at diastole (LVIDd) and systole (LVIDs), anterior wall in diastole (LVAWd) and systole (LVAWs), posterior wall in diastole (LVPWd) and systole (LVPWs), Volume in systole (LVVol-s) and diastole (LVVol-d), and heart reate (HR) were evaluated in all groups. In vivo HR, LV + dP/dt, − dP/dt, end-diastolic pressure (EDP) and end-systolic pressure (ESP) were assessed under basal conditions and after maximal isoproterenol stimulation (ISO 333ng/Kg BW). n= 6 to 13, Data are presented as mean±SEM

^*p<0.05 vs Sham;

^†p<0.001 vs HF-ND;

^‡p<0.001 vs Sham;

^§p<0.0001 vs HF-ND;

^∥p<0.01 vs HF-ND;

^#p<0.0001 vs Sham;

^**p<0.05 vs HF-ND;

^††p<0.01 vs Sham.

One-way ANOVA and Bonferroni test were used between groups.

In vivo cardiac function over the course of the study period and biomarkers of heart failure

Four weeks after surgically-induced MI, echocardiographic studies revealed that LV ejection fraction (LVEF) was significantly reduced in both groups of HF rats compared to Sham, while internal diameter at diastole and systole (LVIDd and LVIDs) were significantly higher in HF rats compared to Sham, which are expected post-MI findings (Figure 2A–C). At the same time-point, cardiac function and structure, evaluated by echocardiography, were not different between the two HF groups, confirming a similar HF level between groups before the randomization to dietary interventions. Twelve months after the beginning of the CR or normal dietary programs in the HF rats, these groups still showed significantly impaired LV systolic function and increased LV diameters and volumes compared to Sham rats (Figure 2D–F and Table). However, impairment in cardiac systolic function, as expressed by LVEF, and adverse LV remodeling, as measured by parameters of ventricular dilatation, further progressed in HF-ND rats (p<0.0001 vs Sham at the end of the study), as observed by comparing measurements performed at 4 weeks post MI induction to measures obtained at the end of the study period. Importantly, 12 months of a CR diet regimen resulted in blunted HF progression, in fact, HF-CR rats showed increased cardiac function and reduced LV diameters and volumes compared to HF-ND rats (Figure 2E–F and Table).

Figure 2. Long-term caloric restriction decreases left ventricular dilatation in ischemic HF.

Ejection fraction (EF) (A), LV internal diameter at diastole (LVIDd) (B) and systole (LVIDs) (C), and heart rate (HR) (D) as measured by echocardiography before the dietary protocol (4 weeks after MI). EF (E), LVIDd (F), LVIDs (G) and HR (H) as measured by echocardiography at the end of the study period (12 months after the dietary protocol started). n = 6 to 11 per group. Data are presented as mean±SEM. * p<0.0001 vs Sham, **p<0.001 vs Sham, #p<0.05 vs HF-ND, ^p<0.05 vs Sham. One-way ANOVA and Bonferroni test were used between all groups.

At the end of the study period, LV catheterization and hemodynamic analysis showed a significant reduction in LV contractility and relaxation in HF-ND group compared to Sham confirming HF-dependent cardiac dysfunction and adverse remodeling (Figure 3 and Table). Of note, HF-CR rats showed a significant decrease in LV relaxation index compared to Sham animals (p<0.05), while +dP/dt was not different between HF-CR and Sham. Importantly, upon infusion of a maximal dose of isoprotenerol (ISO) (333 ng/Kg BW), rats undergoing CR showed a robust improvement of both LV contractility and relaxation compared to control diet HF rats. In fact, HF-CR group showed a significant increase in +dP/dt and a significant decreased in −dP/dt compared to HF-ND group (p<0.001 and p<0.05, respectively) (Figure 3 and Table). This latter finding suggests that a chronic CR diet is able to improve inotropic reserve of the failing heart.

Figure 3. Long-term caloric restriction enhances in vivo LV β-AR inotropic reserve in ischemic HF.

Average LV +dP/dt (A) and LV −dP/dt values (B) in the experimental groups (Sham, n=5; HF-ND, n=7; HF-CR, n=6) evaluated under basal conditions and after maximal isoproterenol stimulation (ISO). Data are presented as mean±SEM. *p<0.05 vs Sham at basal, **p<0.001 vs Sham at basal, #p<0.0001 vs Sham after ISO, ^p<0.001 vs HF-ND after ISO, ^^p<0.05 vs Sham and HF-ND after ISO. Two-way ANOVA analysis and Bonferroni test were used among groups.

Further, we assessed cardiac expression of specific genes known to play a role in HF pathogenesis and progression. At this regard, we measured the mRNA levels of atrial natriuretic factor (ANF), β-Myosin heavy chain (β-MHC) and Endothelin-1 in the LV of all groups via real-time polymerase chain reaction (RT-PCR) (Supplemental Figure 1). Consistent with functional and structural data (see below), ANF and β-MHC mRNA levels were markedly increased in HF-ND rats group compared with Sham rats (p<0.05) (Supplemental Figure 1). Interestingly, ANF levels were significantly reduced in HF-CR rats to levels similar to Sham animals (p<0.05) (Supplemental Figure 1). In addition, we found a trend to increase in Endothelin-1 mRNA levels in HF rats treated with CR diet when compared to HF-ND group (Supplemental Figure 1).

Effects of caloric restriction on cardiac β-AR level and adrenal catecholamine synthesis

Based on our in vivo findings, indicating that CR was able to ameliorate cardiac β-AR responsiveness in HF rats, we measured by Western blot β₁-AR levels on plasma membranes isolated from the hearts of all groups. As expected, we observed a significant reduction in β₁-AR plasma membrane levels in HF-ND hearts compared to Sham hearts (p<0.05) (Figure 4A–B). Importantly, 1 year of CR diet was able to completely prevent HF-related cardiac β₁-AR down-regulation. Indeed, when compared with HF-ND group, HF-CR hearts showed a significant increase in β₁-AR plasma membrane density that was equivalent to that observed in Sham hearts (Figure 4A–B). No differences were found in β₂-AR and β₃-AR membrane levels between Sham, HF-ND and HF-CR groups (Supplemental Figure 2).

Figure 4. Long-term caloric restriction restores β₁-AR membrane level in ischemic HF.

Representative western blot (A) and quantitative data (B) showing β₁-AR protein levels in membrane extracted from cardiac lysates in the 3 experimental groups (Sham, HF-ND and HF-CR). Ponceau used as loading control. n = 6 to 10 per group. Data are presented as mean±SEM. *p<0.05 vs Sham, #p<0.01 vs HF-ND. One-way ANOVA analysis and Bonferroni test were used among groups.

Moreover, we measured adrenal mRNA levels of enzymes involved in the production of Catecholamines: Tyrosine hydroxylase (TH), Dopa decarboxylase (DDC), Dopamine β-hydroxilase (DβH) and Phenylethanolamine N-methyltransferase (PNMT). We found that CR treatment was able to decrease HF-related adrenal TH up-regulation to the levels observed in Sham animals (Supplemental Figure 3). In addition, there was a trend to decrease in adrenal mRNA levels of PNMT in the HF-CR group when compared to HF-ND group (Supplemental Figure 3). No differences were found between HF groups in adrenal DDC and DβH gene expression (Supplemental Figure 3). Accordingly, adrenal weight (AW) and AW/TL ratio were reduced in HF-CR rats to the levels observed in Sham animals (Table and Supplemental Figure 3). No differences in AW/BW ratio have been found between the 3 study groups (Table).

Effects of caloric restriction on cardiac fibrosis, capillary density and innervation

Masson’s Trichrome staining was performed on cardiac specimens obtained from all study groups at the end of the study period. Interstitial Fibrosis was measured in Sham animals as well as in the remote area (RA) and border zone (BZ) of HF groups (Figure 5A). As indicated in Figure 5B, HF-ND rats showed a vast amount of fibrosis that was negligible in Sham animals. Importantly, 12 months of a CR diet dramatically resulted in a robust reduction in cardiac fibrosis both in BZ and RA, as evident by comparing HF-CR to HF-ND hearts (p<0.0001) (Figure 5A–B). At a molecular level, we found that one year CR dietary treatment did not influence levels of pro-fibrotic genes like TGF-β, MMP-2 and TIMP-2 while it increased the levels of the anti-fibrotic TIMP-1 (P<0.05) in HF rats when compared to HF-ND rats (Supplemental Figure 4).

Figure 5. Long-term caloric restriction reduces left ventricular fibrosis in ischemic HF.

Masson-Trichrome staining denoting cardiac fibrosis. (A) Representative images of Sham as well as HF-ND and HF-CR in border zone (BZ). Black scale bar corresponding to 100 µm (B) Quantification of the % fibrosis in Sham as well as remote area (RA) and BZ of HF groups. N= 3 to 5 per group. Data are presented as mean±SEM. *p<0.0001 vs Sham, **p<0.05 vs Sham, #p<0.0001 vs HF-ND (BZ). One-way ANOVA analysis and Bonferroni test were used among all groups.

Moreover we evaluated cardiac angiogenesis at the end of the study period and found that capillary density was significantly increased (p<0.01) in HF-CR hearts compared with HF-ND in the RA (Supplemental Figure 5). There was a trend to increase in capillary density of the BZ of HF-CR hearts when compared to the BZ of HF-ND. (p=0.1452).

Furthermore, we evaluated adrenergic and cholinergic cardiac innervation by immunofluorescence in all study groups (Figure 6). HF-ND rats showed a significant decrease in both adrenergic and cholinergic nerves compared to Sham animals (respectively p<0.05 and p<0.0001) (Figure 6A, C). Remarkably, LV specimens from HF-CR animals revealed an enhancement in adrenergic innervation as suggested by Dβh staining (p<0.01) when compared to HF-ND group (Figure 6A, C). Of note, CR was not able to significantly affect cardiac cholinergic innervation.

Figure 6. Long-term caloric restriction improves left ventricular adrenergic fibers in ischemic HF.

Representative images (A) and quantitative data (B, C). Shown in (A) anti-vesicular acetylcholine transporter (VaChT), Dopamine β-hydroxylase (Dβh) stainings and merged color image (VaChT in red and Dβh in green) of the 3 experimental groups: Sham (top panels), HF-ND (middle panels) and HF-CR (bottom panels); images at full definition (300 dpi) and 40x magnification; yellow scale bar corresponding to 100 µm. Quantification of number of fibers/mm² of VaCHT staining (B) and Dβh staining (C) in 3 groups. N=5 per each group. Data are presented as mean±SEM. *p<0.5 vs Sham, #p<0.0001 vs Sham, ^p<0.01 vs HF-ND. One-way ANOVA analysis and Bonferroni test were used among all groups.

DISCUSSION

In the present study, we have demonstrated that a long-term CR regimen in post-MI rats, started when HF was already established, is able to: a) improve HF-related maladaptive remodeling; b) enhance cardiac function and inotropic reserve; c) restore HF-related β₁-AR down-regulation; d) reduce cardiac fibrosis; d) increase cardiac adrenergic innervation and capillary density.

Over the last decade, several pre-clinical and clinical studies have provided evidence indicating that CR is a safe and effective treatment for age-related CV diseases²⁸. Importantly, it has been recently published that a two-year long CR diet can positively influence quality of life, global health, mood and sleep quality in healthy non-obese adults²⁹. This latter finding is of particular interest since depression is known to negatively impact on morbidity and mortality of HF patients^{30, 31}. Importantly, restricted dietary regimens have been shown to exert cardioprotective effect as shown in several models of cardiac dysfunction such as ischemia/reperfusion, MI and LV hypertrophy^{4, 11, 17, 18, 28, 32–36}. The molecular mechanisms underlying these beneficial effects include, but are not limited to, a reduction in cardiac free-radical and reactive oxygen species production, a decrease in cardiac and systemic inflammation, an improvement in mitochondrial function, as well as an increase in adiponectin circulating levels and consequent activation of cardiac AMP-Activated Protein Kinase-endothelial nitric oxide synthase signaling pathway^{20, 28, 33, 34, 36–40}. However, whether CR can be an effective long-lasting treatment in chronic HF has not been studied, yet. This is of crucial importance since HF is a chronic disease and current treatments are only effective at slowing disease progression. Ahmet et al. have investigated the short-term effects of CR in an ischemic model of HF (up to 10 weeks post-MI) and have found that a CR diet is associated only to a trend-to decrease in the end-diastolic volume compared to controls¹⁸. This latter finding may be explained by the fact that a dietetic treatment needs a longer period to exert its beneficial effects on cardiac structure and function. On the other hand, Katare and colleagues found that short-term IF regimen preserved cardiac volumes and improved LV function after MI²¹. Interestingly, our study indicates that one year of CR diet is able to reduce both systolic and diastolic LV diameters thus positively impacting on cardiac remodeling (Figure 2E–F and Table). Although, we are aware that a dietary regimen should be started as early as possible in order to prevent CV diseases or to slow their progression, it is recognized that, in the clinical practice, patients are motivated to start a dietetic treatment only when a debilitating chronic disease manifests itself. Therefore, we decided to randomize HF-rats to ad libitum or CR diets 4 weeks post-MI, a time point where cardiac dysfunction is already established.

CR can be performed either through daily reduction of caloric intake (15–60% less than ad libitum caloric intake without altering the levels of vitamins, minerals and amino acids) or by IF through the alternation of a complete fasting day with a feed day. Recently, IF has been proposed also in humans to obtain a BW reduction, since it seems to be easier to be followed and sometimes less frustrating than “traditional” CR dietary regimens, where subjects are not authorized to have a freely eating day. In particular, IF protocols proposed for humans consists in: 1) restricting energy intake on 1–3 day/week and eating liberally on the non-restricted days; 2) alternate day fasting regimen which includes a fast day (75% energy restriction) alternating with a ad libitum diet day^{7, 41}. Of note, IF has been shown to induce a BW and fat mass reduction that was similar to those obtained with other CR diets⁴¹. Importantly, our study is the first indicating that a 12-month-long CR regimen is able to improve cardiac remodeling, as indicated by the decreased LV diameters/volumes and fibrosis observed in HF-CR compared to HF-ND rats. Moreover, our results indicate that the beneficial effects of CR on cardiac function are also evident on cardiac inotropic reserve, as indicated by the results of the in vivo hemodynamics after challenge with β-agonist isoproterenol. In order to confirm our results at molecular level, we have measured cardiac β₁-AR membrane density in all study groups. β-AR dysfunctional signaling is recognized to be relevant pathogenic mechanisms in HF and reduced β-AR responsiveness has been described in several models of HF, as well as, in patients with HF^42–45. Moreover, therapeutic intervention able to restore HF-related cardiac β-AR abnormalities has been shown to be therapeutically effective⁴⁴. In our study, HF control hearts show reduced β₁-AR plasma membrane density and, importantly, long-term CR is able to completely restore β₁-AR down-regulation at levels similar to those observed in Sham hearts. CR did not affect plasma membrane levels of β₂-AR and β₃-AR in our post-ischemic HF model.

The adrenal medulla (the central part of the adrenal gland) is the major source of cathecolamine cirluating levels. In fact, almost all circulating Epinephrine levels together with a significant amount of circulating Norepinephrine levels derive from the adrenal medulla⁴³. In our study, one year CR reduced adrenal TH levels as well as adrenal weight and adrenal weight/tibia length ratio to the levels of the Sham animals (Supplemental Figure 3). These data are particularly interesting since the rate-limiting step in Catecholamines’ biosynthesis is the hydroxylation of L-tyrosine to L-DOPA mediated by TH⁴⁶. Given these results, we may speculate that CR diet might be able to blunt, at list in part, HF-related sympathetic nervous system hyperactivity.

Furthermore, in our study, long-term CR diet is associated with a reduction in cardiac fibrosis and increased capillary density (Figure 5 and Supplemental Figure 5). These findings are consistent with the reduction in HW and in LV diastolic and systolic diameters and, overall, confirms the improvement in LV remodeling obtained with CR. Moreover, in accordance with previous data showing positive effects of IF on cardiac fibrosis and angiogenesis in the short period, our results confirm that this therapeutic modality is effective also in the long-term. At molecular level, long-term CR did not affect pro-fibrotic genes like TGF-β, MMP-2 and TIMP-2 while resulted in upregulated levels of an anti-fibrotic gene like TIMP-1⁴⁷ in our model of ischemic HF (Supplemental Figure 4). This protective molecular mechanism can explain, at least in part, the tremendous decrease in fibrosis in the BZ of HF-CR rats compared to HF-ND rats. However, we cannot exclude that a dietary treatment is also able to blunt pro-fibrosis mechanisms (like TGF-β pathway) in an early phase of the dietetic regimen.

Functional and structural data have been confirmed by CR-induced changes in HF biomarkers. Long-term dietary treatment was able to significantly reduce HF-related up-regulation of cardiac ANF and to increase cardiac levels of Endothelin-1. Importantly, even a modest reduction in endothelin-1 gene expression has been shown to decrease cardiac function⁴⁸.

Another pivotal characteristic of chronic HF is represented by cardiac sympathetic denervation that has been described both in pre-clinical models and in patients with HF and it has been related to the high risk of sudden cardiac death due to ventricular arrhythmias, observed in HF patients⁴⁹. During HF, a dramatic loss in cardiac nerve fibers and, in particular, in sympathetic fibers has been described^{50, 51}. This is due, at least in part, to cholinergic transdifferentiation of cardiac SNS cardiac nerve terminals⁵². As expected, we observed a dramatic loss of both cholinergic and adrenergic fibers in HF-ND rats compared to Sham animals. More importantly, CR resulted in a significant increase in sympathetic cardiac nerve fibers as evident by comparing HF-CR to HF-ND hearts. This result is consistent with previous findings that showed neuroprotective effects of CR in models of neuronal damage or cognitive impairment⁵³.

Recently, a CR dietary program has been tested in obese older patients with clinically stable HF and preserved EF in a randomized clinical trial²². The authors found that CR was able to increase peak V̇O₂ and to decrease LV mass, circulating C-reactive protein and cholesterol levels²². This latter investigation is consistent with the results of the present study and open to the idea to test if CR could be effective also in patients with ischemic HF and whether this dietary intervention may have an additive effect when combined with aerobic exercise training.

From a clinical point of view, we can hypothesize that CR could be associated with other pharmacological medications used in HF such as β-blockers and ACE inhibitors. Long-term β-blocker treatment is associated with weight gain and new onset diabetes^{54, 55}. Consequently, a restricted dietary program could, at least in part, counteract the metabolic side effects of β-blockade in patients with HF. However, the concomitant utilization of β-blocker and CR diet needs further investigation since Azar et al. have shown that obese individuals treated with β-blockers exhibit significant impairments in weight loss and waist circumference reduction in response to a hypocaloric dietary intervention⁵⁶.

Overall the findings of the present study are of particular interest since dietary interventions represent a cost saving therapeutic modality, whose favorable properties seem to show effectiveness comparable to that obtained with pharmacological interventions currently used in the clinical practice.

Study limitations

Due to the long-term nature of our study, it is difficult to obtain information regarding the mechanisms that are involved in the beneficial effects of CR in HF. In fact, the functional, molecular and structural observations that we have found at the end of the study period (13-months post-MI) may be the result of molecular mechanisms acting at the initial phase of the dietetic regimen. Moreover, we recognize that the lack of a Sham group fed with a CR diet could represent a limitation of the present study, but our interest was mainly focused on the long-term effects of CR in a post-ischemic model of HF, where altered β-AR signaling, increased cardiac fibrosis and denervation represent relevant pathogenic mechanisms that are currently therapeutic targets to curb HF progression. At this regard, Ahmet and colleagues characterized the outcomes of different long-term diets (CR vs ad-libitum) on cardiovascular fitness during aging. They showed positive effects of a long-lasting CR on LV diastolic function and remodeling in 24-month-old un-operated animals¹⁸. Compared to ad-libitum rats, 24-month-old CR rats showed reduced levels of cardiac fibrosis, attenuated diastolic dysfunction, but unchanged LV systolic function.

Conclusions

In summary, the present study reports that long-term CR, started when HF was already established, reverses cardiac dysfunction, attenuates LV remodeling and improves cardiac inotropic reserve in an experimental model of post-ischemic HF. As a contributing mechanism, we found that CR diet significantly reduces cardiac fibrosis and improves sympathetic cardiac innervation and β-AR membrane density in HF. Importantly, the beneficial effects of CR are strongly effective after 12 months of this regimen, providing evidence that this therapeutic approach is applicable long-term, although started when cardiac dysfunction was already evident. The true clinical potential of CR for a place in the HF therapeutic armamentarium awaits clinical validation in patients.

WHAT IS NEW?

• We investigated for the first time the effects of a long-term caloric restriction (CR) regimen in an animal model of ischemic heart failure (HF).

• CR improved cardiac inotropic reserve, as indicated by the results of the in vivo hemodynamics, and restored cardiac β₁-adrenergic receptor plasma membrane density in a chronic HF model.

• CR significantly curbed the dramatic loss in cardiac sympathetic nerve fibers in the failing hearts.

• CR reverted cardiac maladaptive remodeling in ischemic HF reducing cardiac fibrosis and increasing capillary density.

WHAT ARE THE CLINICAL IMPLICATIONS?

• For chronic diseases such HF, there is an urgency to identify new therapies that are successful if used for a long period.

• We found that a long-term and sustained CR regimen is an effective treatment in a model of ischemic HF.

• The findings of the present study are of particular interest since dietary interventions represent a cost saving therapeutic modality, whose favorable properties seem to show effectiveness comparable to that obtained with pharmacological interventions currently used in clinical practice.

• The clinical potential of CR for a place in the HF therapeutic armamentarium awaits clinical validation in patients.