Chronic low-intensity exercise attenuates cardiomyocyte contractile dysfunction and impaired adrenergic responsiveness in aortic-banded mini-swine

Abstract

Exercise improves clinical outcomes in patients diagnosed with heart failure with reduced ejection fraction (HFrEF), in part via beneficial effects on cardiomyocyte Ca²⁺ cycling during excitation-contraction coupling (ECC). However, limited data exist regarding the effects of exercise training on cardiomyocyte function in patients diagnosed with heart failure with preserved ejection fraction (HFpEF). The purpose of this study was to investigate cardiomyocyte Ca²⁺ handling and contractile function following chronic low-intensity exercise training in aortic-banded miniature swine and test the hypothesis that low-intensity exercise improves cardiomyocyte function in a large animal model of pressure overload. Animals were divided into control (CON), aortic-banded sedentary (AB), and aortic-banded low-intensity trained (AB-LIT) groups. Left ventricular cardiomyocytes were electrically stimulated (0.5 Hz) to assess Ca²⁺ homeostasis (fura-2-AM) and unloaded shortening during ECC under conditions of baseline pacing and pacing with adrenergic stimulation using dobutamine (1 μM). Cardiomyocytes in AB animals exhibited depressed Ca²⁺ transient amplitude and cardiomyocyte shortening vs. CON under both conditions. Exercise training attenuated AB-induced decreases in cardiomyocyte Ca²⁺ transient amplitude but did not prevent impaired shortening vs. CON. With dobutamine, AB-LIT exhibited both Ca²⁺ transient and shortening amplitude similar to CON. Adrenergic sensitivity, assessed as the time to maximum inotropic response following dobutamine treatment, was depressed in the AB group but normal in AB-LIT animals. Taken together, our data suggest exercise training is beneficial for cardiomyocyte function via the effects on Ca²⁺ homeostasis and adrenergic sensitivity in a large animal model of pressure overload-induced heart failure.

NEW & NOTEWORTHY Conventional treatments have failed to improve the prognosis of heart failure with preserved ejection fraction (HFpEF) patients. Our findings show chronic low-intensity exercise training can prevent cardiomyocyte dysfunction and impaired adrenergic responsiveness in a translational large animal model of chronic pressure overload-induced heart failure with relevance to human HFpEF.

INTRODUCTION

Traditional therapeutic approaches commonly used to treat heart failure with reduced ejection fraction (HFrEF) have largely failed in heart failure with preserved ejection fraction (HFpEF) patients (for reviews, see Refs. 43, 48). Given the difficulties in treating HFpEF and the systemic nature of the disease, utilizing exercise therapeutically for this heart failure subtype may carry added clinical importance (23, 35). Several clinical studies have shown exercise to be a safe way to improve quality of life and cardiorespiratory fitness for both HFrEF and HFpEF patients (7, 10, 12, 16, 26, 39, 42, 44). In HFrEF, improvements to general health status following exercise training may be related to its effects on cardiomyocyte excitation-contraction coupling (ECC) including improvement in speed and amplitude of systolic Ca²⁺ transients and contraction and enhancement of sarco/endoplasmic reticulum Ca²⁺ ATPase (SERCA)-mediated diastolic sarcoplasmic reticulum (SR) Ca²⁺ reuptake and relaxation (25). In contrast, the effects of exercise training on cardiomyocyte function in a setting of HFpEF are largely unknown.

In heart failure, cardiomyocyte Ca²⁺-handling processes are impaired resulting in aberrant timing and efficacy of ECC, thereby upsetting the inherent stability and plasticity of ventricular contraction and relaxation. Thus, abnormalities in the ability to handle Ca²⁺ represent a primary underlying factor leading to clinical manifestations of heart failure, including contractile dysfunction and arrhythmia. In contrast with the extensive literature on cardiomyocyte function in HFrEF (1, 33, 57), much less is known regarding the mechanisms disrupting cardiomyocyte ECC in HFpEF as these patients often do not qualify for cardiac transplantation, limiting the ability to study human tissue. Thus the fundamental question is cardiac dysfunction in HFpEF intrinsic or extrinsic to the cardiomyocyte remains controversial.

The goal of this study was to investigate the effects of low-intensity exercise training on cardiomyocyte Ca²⁺ handling and contractile function in a Yucatan mini-swine model of chronic pressure overload-induced heart failure that exhibits the cardiac features of HFpEF including preserved ejection fraction at rest, diastolic dysfunction, cardiac hypertrophy, diminished left ventricular (LV) contractile reserve, increased LV fibrosis, and lung congestion (13, 14, 21, 34, 40, 41). We hypothesized chronic low-intensity exercise training would attenuate pressure overload-induced impairments to cardiomyocyte ECC and preserve cardiomyocyte functional reserve following adrenergic challenge.

METHODS

Ethical approval, aortic banding, and exercise training.

All animal protocols were in accordance with the Principles for the Utilization and Care of Vertebrate Animals Used in Testing Research and Training and approved by the University of Missouri Animal Care and Use Committee. Male Yucatan miniature swine (29–32 kg; 8 mo old) were assigned into three groups: sedentary control (CON), aortic-banded sedentary (AB), and aortic-banded low-intensity exercise trained (AB-LIT; n = 7 for all groups). Aortic bands were placed around the ascending aorta (proximal to the brachiocephalic artery), and a systolic transstenotic gradient of ~70 mmHg was established (74 ± 2 and 72 ± 1 mmHg for AB and AB-LIT, respectively, P = NS) over a period of ≈6 mo as previously reported (41). Two months after aortic banding, AB-LIT animals started chronic exercise training for 17 wk using low-intensity treadmill training. The exercise training protocol included treadmill running 55 min/day, 3 days/week, for 17 wk with gradually increasing intensity until finally consisting of the following: 1) a 5 min warm up at 1.5 mph; 2) 45 min at 2.5 mph; and 3) a 5 min cool down at 1.5 mph. Dissection of vital tissues and removal of skeletal muscle (deltoid) for analysis of citrate synthase activity (50) occurred at the time of death, 24 h after the last training bout. Animals were fed a standard diet averaging 15–20 g/kg once daily, and water was provided ad libitum.

Transthoracic echocardiography.

Transthoracic echocardiography was performed while animals were under inhaled isoflurane anesthesia (0.5%) in the supine/right lateral position 6 mo postbanding. Short axis two-dimensional and M-mode images were recorded at the mid-papillary level using a 1.5- to 4-MHz transducer on a GE Vividi Ultrasound system. Fractional shortening (FS%) and LV diastolic chamber dimension and wall thickness [LV internal diastolic dimension (LVIDd), LV diastolic wall thickness (LV WTd), interventricular septum diastolic wall thickness (IVS WTd), and relative left ventricular and septal diastolic wall thickness normalized to LVIDd (relative LV + IVS WTd)] were calculated from M-mode recordings as previously reported (20). From an apical four-chamber view, velocity time integral (VTI), a measure directly proportional to stroke volume (SV), was measured at the level of the aortic annulus using pulse wave Doppler. Stroke volume was determined using the equation SV = π(o^r) × VTI.

Cardiomyocyte Ca²⁺ and contractile measurements.

Midmyocardial cardiomyocytes were enzymatically isolated from LV wedge preparations of excised hearts as previously described (19, 36). Isolated myocytes were plated on laminin-coated glass coverslips and dye loaded for 20 min at room temperature (22–24°C) with 5 μM fura-2-AM (Thermo/Molecular Probes) in physiological saline solution (PSS) containing the following (in mM): 135 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 10 d-glucose, 10 HEPES, and 1 NaHCO₃, pH 7.4 with NaOH. After a 40- to 60-min wash period in PSS, myocytes were placed in a temperature-controlled bath mounted on an inverted fluorescence microscope (IX71; Olympus America, Center Valley, PA). Myocytes were superfused with PSS via an inline solution heater and viewed using a ×40 oil-immersion objective (UApo340: 1.35 NA; Olympus). Action potentials (0.5 Hz) were triggered with electrical field stimulation (S48; Grass Instruments, Warwick, RI) via platinum electrodes placed at the edges of the bath. Epifluorescence microcopy was performed using an IonOptix Ca²⁺ and Contractility Recording System (IonOptix, Milton, MA) with alternating excitation at 340 and 380 nm, and fluorescence emission was measured from 510 to 565 nm. Background-subtracted fluorescence emission at each wavelength (F₃₄₀ and F₃₈₀) was obtained, and the fura-2 ratio (R = F₃₄₀/F₃₈₀) was used as an index of intracellular Ca²⁺ concentration ([Ca²⁺]_i). The fura-2 ratio was acquired simultaneously with sarcomere length (L_s) at 250 Hz. Myocytes were equilibrated to physiological temperature (35–37°C) for 5 min, followed by electrical field stimulation at 0.5 Hz for assessment of Ca²⁺ transients and sarcomere shortening. In addition to measurements obtained under untreated conditions, a subset of cardiomyocytes were treated with 1 μM dobutamine (±) to assess cardiomyocyte adrenergic functional reserve. With dobutamine, cardiomyocytes experienced a gradual inotropic effect that transitioned into “Ca²⁺ overload” with incidence of spontaneous Ca²⁺ transients and contractile activity. Therefore, the action-potential-induced Ca²⁺ transient and shortening response, which preceded the onset of spontaneous activity, was used for analysis. Action-potential-induced Ca²⁺ transients and L_s shortening were analyzed offline using Ionwizard 6.3 software (IonOptix) with the following parameters of Ca²⁺ handling and contractile function assessed: 1) diastolic [Ca²⁺]_i; 2) peak systolic [Ca²⁺]_i; 3) Ca²⁺ transient amplitude (Δratio), calculated as peak systolic fura-2 ratio – diastolic fura-2 ratio; 4) rate of rise of the Ca²⁺ transient (ΔR/Δt) in arbitrary units per second; 5) diastolic L_s; 6) systolic L_s; 7) shortening magnitude, calculated as diastolic L_s – systolic L_s; and 8) maximum rate of shortening (ΔL/Δt, in μm/s). For presentation of sarcomere length vs. Ca²⁺ relationships during ECC, traces of Ca²⁺ and sarcomere length from each cardiomyocyte within each group were temporally aligned based on the time of electrical field stimulation, and the average of responses of all cells were plotted with Ca²⁺ on the x-axis and sarcomere length on the y-axis.

Western blotting.

LV tissue was homogenized in lysis buffer (150 mM NaCl, 50 mM Tris-base, pH 8.0, 0.1% Triton-X 100, 2 mM EDTA, and 10 μl/ml HALT protease and phosphatase inhibitor). After centrifugation, supernatants were collected and the protein content was measured according to the method by Bradford (3). Protein (1 μg) was treated with Laemmli sample buffer containing 5% β-mercaptoethanol and heated for 5 min [15 min for phospholamban (PLB)] at 95°C. Protein was loaded into 4–20% polyacrylamide gels (Genscript), separated by SDS-PAGE at 140 V for 50 min, and transferred at 34 V for 60 min to a polyvinylidene difluoride membrane. Membranes were then blocked for 1 h with 5% milk in Tris-buffered saline and 0.1% Tween 20 (TBST). Blots were incubated overnight (25°C) with the primary antibody against SERCA2a (1:1,000; ThermoFisher Scientific), phospholamban (PLB; 1:1,000; ThermoFisher Scientific), sodium-calcium exchanger (NCX; 1:1,000; ThermoFisher Scientific), Na⁺-K⁺-ATPase (1:1,000; ThermoFisher Scientific), ryanodine receptor (RyR; 1:1,000; ThermoFisher Scientific), Ser16 phosphorylated PLB (1:1,000; Badrilla), Thr17 phosphorylated PLB (1:1,000; Badrilla), and calsequestrin (CSQ; 1:10,000; ThermoFisher Scientific). The following day, membranes were washed five times for 10 min in TBST and then incubated for 1 h with a horseradish peroxidase-conjugated secondary antibody (anti-mouse polyclonal IgG, 1:1,000; ThermoFisher Scientific; or anti-rabbit IgG, 1:1,000; Cell Signaling). Membranes were again washed five times for 10 min in TBST, visualized via enhanced chemiluminescence (Luminata Forte; Millipore), and quantified by densitometry with Kodak 4000R Imager and Molecular Imagery software.

Statistical analysis.

All data analysis was performed using SigmaPlot version 12.3 (SysStat Software, San Jose, CA). To examine phenotypic changes within populations of cardiomyocytes obtained from the respective animal groups and conditions, the following number of cells (n) from number of animals (N) were sampled: untreated baseline conditions, n = 53/N = 6 CON, n = 37/N = 5 AB, and n = 72/N = 7 AB-LIT; dobutamine, n = 48/N = 6 CON, n = 32/N = 5 AB, and n = 58/N = 7 AB-LIT. The value of n was utilized for subsequent statistical analysis. Group comparisons were made using one-way ANOVA, with Student Newman-Keuls post hoc analysis. Data are reported as individual observations, means of all responses, or means ± SE, with significance denoted at the P < 0.05 and P < 0.10 levels (8, 53).

RESULTS

Aortic-banded animals (AB and AB-LIT) exhibited normal fractional shortening and SV as assessed in vivo using echocardiography (Table 1). Global cardiac hypertrophic remodeling was observed in both the AB and AB-LIT groups and associated with a decrease in LV cardiomyocyte length:width ratio, consistent with concentric cellular remodeling (Table 1). With the use of echocardiography, an increase in relative diastolic wall thickness (predominately in the septum) was observed in parallel with no change in LV internal diastolic dimension in AB animals (vs. CON), confirming our postmortem findings. Changes in relative diastolic wall thickness were attenuated in the AB-LIT group. Decreases in the cardiomyocyte length:width ratio in AB (vs. CON) were due to increased cardiomyocyte width, while the decrease in ratio in AB-LIT (vs. CON) was due to a decrease in cardiomyocyte length. An increase in lung weight in the AB group compared with CON, indicative of pulmonary congestion, was attenuated in AB-LIT animals. No significant differences in citrate synthase activity were found in the deltoid muscle between groups.

Table 1.

Postmortem heart weight and body weight, in vivo echocardiographical measurements (fractional shortening, stroke volume, LVIDd, LV WTd, IVS WTd, and relative LV + IVS WTd), length, width, and length:width ratio of enzymatically isolated cardiomyocytes, lung weight, and deltoid citrate synthase activity of CON, AB, and AB-LIT groups

	CON	AB	AB-LIT
Post mortem heart weight, g	207.5 ± 9.8	256.3 ± 9.2*	257.4 ± 7.0*
Post mortem body weight, kg	46.1 ± 2.2	47.4 ± 2.4	47.7 ± 1.3
In vivo fractional shortening	46 ± 2	47 ± 2	47 ± 2
In vivo stroke volume, ml	67 ± 4	65 ± 3	62 ± 3
LVIDd, mm	49 ± 2	47 ± 1	50 ± 1
LV WTd, mm	7.4 ± 0.4	8.3 ± 0.7	8.2 ± 0.4
IVS WTd, mm,	11.0 ± 0.4	12.2 ± 0.2†	11.8 ± 0.4
Relative LV + IVS WTd	0.38 ± 0.01	0.44 ± 0.02*	0.40 ± 0.01#
Cardiomyocyte length, µm	214 ± 4	207 ± 4	191 ± 3*#
Cardiomyocyte width, µm	38 ± 1	43 ± 1*	36 ± 1#
Length:width ratio	6.0 ± 0.2	5.4 ± 0.2*	5.6 ± 0.2*
Lung weight, g	267 ± 14	317 ± 21†	294 ± 10
Deltoid citrate synthase activity, µmol·g wet weight−1·min−1	20.4 ± 2.8	15.2 ± 2.0	21.5 ± 4.3

CON, control; AB, aortic-banded; AB-LIT, aortic-banded exercise trained; LVIDd, left ventricular internal diastolic dimension; LV WTd, left ventricular diastolic wall thickness; IVS WTd, interventricular septum diastolic wall thickness; relative LV + IVS WTd, left ventricular and septal diastolic wall thickness normalized to LVIDd calculated as (LV WTd + IVS WTd)/LVIDd.

^*P < 0.05 vs. CON.

^#P < 0.05 vs. AB.

^†P < 0.10 vs. CON.

To examine the role of the cardiomyocyte in functional changes observed at the whole organ level, intracellular Ca²⁺ (fura-2) and sarcomere length were simultaneously monitored in enzymatically isolated LV myocytes in response to electrical field stimulation (Fig. 1A). Compared with CON, cardiomyocytes of AB exhibited similar diastolic Ca²⁺ and sarcomere length (Fig. 1, B and C). However, systolic Ca²⁺ was reduced in AB (Fig. 1B) resulting in impaired systolic sarcomere shortening (Fig. 1C), reduced Ca²⁺ transient amplitude (Fig. 2A), and impaired shortening amplitude (Fig. 2B). Ca²⁺ transient and shortening rates were also slower in AB vs. CON (Fig. 2, C and D). In AB-LIT animals, a decrease in diastolic Ca²⁺ was observed compared with both CON and AB (Fig. 1B) along with a longer diastolic sarcomere length (Fig. 1C). Exercise training also attenuated the decrease in Ca²⁺ transient amplitude and rate (Fig. 2, A and C) observed in the AB group, but did not prevent aortic banding-induced decreases in shortening amplitude or rate (Fig. 2, B and D). Examination of the sarcomere length-Ca²⁺ relationship during ECC shows an upward and leftward shift during the systolic phase in the AB group, highlighting the impaired amplitude of the Ca²⁺ transient and associated shortening compared with CON (Fig. 2E). In AB-LIT animals, a longer sarcomere length is evident throughout the ECC cycle in addition to an attenuated leftward shift in the Ca²⁺ transient.

Fig. 1.

Cardiomyocyte Ca²⁺ transients and sarcomere shortening during excitation-contraction coupling. A: example traces of action potential-induced Ca²⁺ transients (fura-2 ratio, top traces) and sarcomere length (bottom traces) in cardiomyocytes of control (CON), aortic-banded (AB), and aortic-banded exercise trained (AB-LIT) groups. Electrical field stimuli are indicated by arrows. B and C: summary data of diastolic (○, ◇, ☐) and systolic (●, ◆, ■) Ca²⁺ (B) and sarcomere length (C) in CON (○, ●), AB (◇, ◆), and AB-LIT (☐, ■). *P < 0.05 vs. CON. §P < 0.10 vs. CON and AB.

Fig. 2.

Summary data of cardiomyocyte Ca²⁺ transient and sarcomere shortening amplitude and rate. A and B: summary data of Ca²⁺ transient (A) and shortening (B) amplitude in cardiomyocytes of CON, AB, and AB-LIT groups. C and D: summary data of maximum Ca²⁺ transient rate (C) and maximum (most negative) shortening rate (D). E: sarcomere length-Ca²⁺ loop relationships during excitation-contraction coupling. Loops are averaged traces of all cardiomyocytes within CON (●), AB (◇), and AB-LIT (). Traces were aligned for averaging based on electrical stimulation, with error bars omitted for clarity. *P < 0.05 vs. CON. #P < 0.05 vs. AB.

A consistent observation in heart failure is a lack of cardiac functional reserve following stress. We therefore examined cardiomyocyte function following acute application of the adrenergic agonist dobutamine. In all groups, dobutamine application induced the expected increase in Ca²⁺ transient and shortening amplitude vs. untreated conditions. The maximum inotropic response was defined as the largest action-potential-induced shortening response before the onset of spontaneous contractile activity (Fig. 3A, denoted by arrow). The time to maximum inotropic response was delayed in AB (Fig. 3B) animals compared with CON and AB-LIT. Similar to observations in untreated conditions, cardiomyocytes of AB were similar to CON in diastolic Ca²⁺ and sarcomere length yet exhibited reduced systolic Ca²⁺ (Fig. 4B), impaired sarcomere shortening (Fig. 4C), decreased Ca²⁺ transient amplitude (Fig. 5A), and reduced shortening amplitude (Fig. 5B) vs. CON. The rate of the Ca²⁺ transient was also slower in AB vs. CON (Fig. 5C). Exercise training completely prevented the AB-induced decreases in cardiomyocyte Ca²⁺ transient amplitude (Fig. 5A), shortening amplitude (Fig. 5B), and Ca²⁺ transient rate (Fig. 5C). Sarcomere length-Ca²⁺ relationship analysis during the ECC cycle again showed an upward and leftward shift in AB during the systolic phase vs. CON, indicative of impaired Ca²⁺ transients and sarcomere shortening following dobutamine treatment (Fig. 5E). As with untreated conditions, cardiomyocytes of AB-LIT operated at longer sarcomere lengths throughout ECC in the presence of dobutamine. However, with dobutamine, the amplitude of Ca²⁺ transients and shortening were similar between AB-LIT and CON, resulting in similar shape of the sarcomere length-Ca²⁺ relationship but at extended sarcomere lengths evident as an upward, but not leftward, shift.

Fig. 3.

Time to maximum inotropic response following dobutamine challenge is prolonged with aortic banding and prevented by exercise training. A: example cardiomyocyte shortening traces (0.5 Hz) before and following addition of dobutamine (1 μM, indicated by bar). Maximum inotropic response was taken as the response before the onset of spontaneous contractile activity (indicated by arrow). B: summary data of time to maximum dobutamine-induced shortening response in cardiomyocytes of CON, AB, and AB-LIT groups. *P < 0.05 vs. CON. #P < 0.05 vs. AB.

Fig. 4.

Cardiomyocyte Ca²⁺ transients and sarcomere shortening during excitation-contraction coupling following dobutamine challenge. A: example traces of action potential-induced Ca²⁺ transients (fura-2 ratio, top traces) and sarcomere length (bottom traces) in dobutamine-treated cardiomyocytes of CON, AB, and AB-LIT groups. Electrical field stimuli are indicated by arrows. B and C: summary data of diastolic (○, ◇, ☐) and systolic (●, ◆, ■) Ca²⁺ (B) and sarcomere length (C) in CON (○, ●), AB (◇, ◆), and AB-LIT (☐, ■) in the presence of dobutamine (1 μM). *P < 0.05 vs. CON. §P < 0.10 vs. CON and AB.

Fig. 5.

Summary data of cardiomyocyte Ca²⁺ transient and sarcomere shortening amplitude and rate following dobutamine challenge. A and B: summary data of Ca²⁺ transient (A) and shortening (B) amplitude in dobutamine-treated cardiomyocytes of CON, AB, and AB-LIT groups. C and D: summary data of maximum Ca²⁺ transient rate (C) and maximum (most negative) shortening rate (D) in the presence of dobutamine (1 μM). E: sarcomere length-Ca²⁺ loop relationships during excitation-contraction coupling with dobutamine. Loops are averaged traces of all cardiomyocytes within CON (●), AB (◇), and AB-LIT (). Traces were aligned for averaging based on electrical stimulation, with error bars omitted for clarity. *P < 0.05 vs. CON. #P < 0.05 vs. AB.

We next examined Ca²⁺ handling protein levels in whole heart homogenates using Western blotting (Fig. 6). No changes were observed in SERCA levels among groups; however, the SERCA/PLB ratio was increased in AB-LIT (Fig. 6D, vs. CON and AB). Phosphorylation of PLB at the Ser16 site was reduced in AB (Fig. 6E, vs. CON and AB- LIT), while phosphorylation of PLB at the Thr17 site was reduced in both AB and AB-LIT (Fig. 6F, vs. CON). The level of the Na⁺-K⁺ ATPase was similar among groups, while protein of a primary Ca²⁺ extrusion pathway in the cardiomyocyte NCX was increased in AB-LIT (Fig. 6H, vs. CON and AB). RyR levels were increased in AB and AB-LIT vs. CON (Fig. 6I). Taken together, these findings suggest complex regulation of Ca²⁺ handling protein expression and baseline phosphorylation in response to both pressure overload and exercise training.

Fig. 6.

Protein expression of Ca²⁺ handling proteins. A: example Western blots of sarco/endoplasmic reticulum Ca²⁺ ATPase (SERCA; 100 kDa), phospholamban (PLB; 6 kDa), the Ser16 phosphorylation site of PLB (SER16-PLB; 6 kDa), the Thr17 phosphorylation site of PLB (THR17-PLB; 6 kDa), the Na⁺-K⁺-ATPase (NA/K; 100 kDa), sodium-calcium exchanger (NCX; 120 kDa), ryanodine receptor (RYR; 565 kDa), and calsequestrin (CSQ; 50 kDa). Summary data of protein expression (relative to CSQ) of SERCA (B), PLB (C), SERCA/PLB ratio (D), SER16-PLB (E), THR17-PLB (F), Na⁺-K⁺ ATPase (G), NCX (H), and RYR (I). Data shown are from whole heart homogenates of CON, AB, and AB-LIT groups; n = 5–7 animals per group. *P < 0.05 vs. CON. #P < 0.05 vs. AB. §P < 0.10 vs. CON and AB.

DISCUSSION

The present study in a large animal model with cardiac features of HFpEF reveals chronic low-intensity exercise training attenuates functional impairments in Ca²⁺ homeostasis induced by pressure overload. In rodent models in the absence of pathological stimuli, exercise training increases expression of SERCA (55), PLB phosphorylation (24), and myofilament sensitivity (11, 55) within cardiomyocytes, all of which augment ECC. Chronic exercise training also increased cardiomyocyte power output in healthy Yucatan mini-swine (22). However, these findings were observed in response to exercise intensities likely intolerable for heart failure patients. This study demonstrates positive adaptations to cardiomyocyte ECC using an exercise stimulus <50% of the absolute intensity typically utilized in comparable physiological swine exercise studies and at a reduced frequency of three exercise bouts per week (vs. 5 sessions per week in healthy animal models) (28).

Beneficial effects of exercise on cardiomyocyte ECC have previously been reported in animal models of heart failure, although they have been predominately observed in rodent models with more relevance to HFrEF. Exercise training initiated following myocardial infarction (MI) in mice (9) and rats (54) prevented MI-associated changes in SR Ca²⁺-handling protein expression and myofilament dysfunction, with restoration of Ca²⁺ transient amplitude and contraction during ECC. Mechanistically, we observed an increase in the SERCA/PLB ratio and preservation of normal PLB phosphorylation at Ser16 in the AB-LIT group. Exercise training also induced a significant increase in NCX levels, which favors diastolic Ca²⁺ extrusion and lower diastolic Ca²⁺ (9, 56). Theoretically, this effect of training could shift cellular Ca²⁺ flux toward Ca²⁺ extrusion and lower SR Ca²⁺ content, particularly when considered in parallel with the reduction in PLB phosphorylation of Thr17 in the AB-LIT group. Overall, these findings may partially explain how exercise training attenuated, but did not fully prevent, AB-induced decreases to Ca²⁺ transient amplitude in AB-LIT animals.

Chronic low-intensity exercise training also preserved normal cellular Ca²⁺ transient and shortening reserve assessed by adrenergic challenge. Evidence of this concept is provided by the maintenance of Ca²⁺ transient amplitude and rate, as well as shortening amplitude and rate, in AB-LIT cardiomyocytes compared with CON with dobutamine treatment. Although dobutamine exhibits activity at both α- and β-adrenoreceptor subtypes (46), it was utilized for the present study given the rapid and robust inotropic effects on cardiomyocytes via the β₁-receptor (2). The time to maximum inotropic response in AB-LIT was also similar to CON, indicative of exercise-induced beneficial effects on β-adrenergic sensitivity (9, 29). Together, our data illustrate the beneficial effects of low-intensity exercise training on cardiomyocyte function in an experimental setting of pressure overload and suggest the preservation of cellular β-adrenergic sensitivity following chronic exercise training may counteract decreases in cardiac reserve highly prevalent in heart failure patients.

Both the AB and AB-LIT groups exhibited global hypertrophic cardiac remodeling and an associated decrease in cardiomyocyte length:width ratio. As expected, cardiomyocytes in the AB group displayed an increase in cell width and no change in cell length compared with CON, consistent with hypertrophic cellular remodeling and the addition of sarcomeres in parallel. However, chronic exercise training caused unexpected structural remodeling in AB-LIT animals characterized by a decrease in length, with no change in width, when compared with CON. This cellular morphological change contrasts the ≈6–12% increase in cardiomyocyte length due to series addition of sarcomeres, along with a commensurate increase in cross-sectional area, traditionally observed following chronic endurance treadmill exercise-training programs in nonpathological animal models (25, 37, 38). An additional intriguing finding was the discordance between global and cellular hypertrophic remodeling in AB-LIT animals, where an increase in heart weight was observed without an associated increase in cardiomyocyte length or width. Cardiomyocyte morphology was obtained from the midmyocardium of the LV free wall, and we cannot exclude that cardiomyocyte hypertrophy in other cardiac regions contributed to an increase in overall cardiac mass. Cardiomyocyte proliferation following the combined pressure overload and exercise stimuli could also account for an increase in heart size. However, endogenous mechanisms of cardiomyocyte renewal in the adult heart are controversial, and most investigations indicate the adult heart has a low capacity for cardiomyocyte proliferation (15). The sarcomere length-Ca²⁺ relationship analysis also revealed that cardiomyocytes operated at longer sarcomere length throughout the ECC cycle in AB-LIT animals under untreated conditions as well as during adrenergic challenge. Such findings point to sarcomeric ultrastructural remodeling and an extension of cardiomyocyte slack length, which is classically viewed to be dependent on the length of the giant structural protein titin (17, 45). Changes in titin isoform expression and phosphorylation status have been shown with both cardiac disease (30, 31) and following exercise training (6, 18), and it is plausible that titin-based extension of cardiomyocyte slack length in AB-LIT contributes to the longer sarcomere length during ECC. In total, our findings suggest exercise of a low absolute intensity under conditions of chronic pressure overload in this translational swine model promotes alternate cardiomyocyte remodeling processes that differ from previous observations made in healthy animals.

The present investigation adds to previous work from our laboratories, which illustrates cardiomyocyte Ca²⁺ transients and shortening are impaired in a translational large animal model with cardiac characteristics consistent with HFpEF. The extent by which cardiomyocyte ECC is altered in HFpEF remains controversial, and interpretation of cardiomyocyte data from rodent models with relevance to HFpEF is often complicated by an adaptive hypercontractile compensatory phase (49) that rapidly decompensates to a phenotype more closely resembling HFrEF (32). In this study, data indicate the reduction in Ca²⁺ transient amplitude and rate in the AB group was not associated with changes in SERCA or PLB protein level. These findings stand in contrast to alterations in cardiomyocyte ECC particularly in end-stage decompensated heart failure where expression of SERCA is decreased (33, 57). Our data indicate decreased PLB phosphorylation at both Ser16 (protein kinase A site) and Thr17 (Ca²⁺/calmodulin-dependent protein kinase II site) may play a role by impairing SR Ca²⁺ sequestration via increasing the inhibitory effect of PLB on SERCA. In the absence of changes in SERCA/PLB expression, decreased PLB phosphorylation would lead to greater association of PLB with SERCA, increased inhibition of SERCA-mediated SR Ca²⁺ uptake, decreased SR Ca²⁺ content, and a decrease in the Ca²⁺ transient during ECC.

Decreases in Ser16 phosphorylation of PLB in AB animals, observed in parallel with the delayed time to maximum response after dobutamine exposure, are consistent with the concept of “β-adrenoreceptor desensitization,” which has been observed with advancing age and heart failure (4, 5, 52). When cardiomyocytes were challenged with the adrenergic agonist dobutamine, all groups exhibited an increase in Ca²⁺ transient amplitude and shortening. However, Ca²⁺ transient amplitude and myocyte shortening were still decreased in AB vs. CON similar to observations under untreated conditions, indicating that β₁-receptor activation was unable to rescue function in the AB group. β₁-adrenergic stimulation exerts its cellular effects via the G_s-cAMP-protein kinase A signaling cascade, with resulting phosphorylation of Ca²⁺ handling proteins including PLB at the Ser16 site. Sustained elevation in Ca²⁺ cycling following β₁-adrenergic stimulation also leads to a Ca²⁺ dependent increase in PLB phosphorylation at Thr17 via Ca²⁺/calmodulin-dependent protein kinase II, although this secondary response may be blunted in heart failure due to excessive phosphatase activity (27). Our findings suggest impaired cardiomyocyte Ca²⁺ transients and shortening in this translational heart failure model are associated with decreased PLB phosphorylation via multiple signaling pathways.

Limitations.

The large animal model of pressure overload-induced heart failure used in this investigation reflects a number of features consistent with HFpEF, including LV hypertrophy, increased lung weight, and normal resting fractional shortening and SV. Taken into consideration with other published studies from our laboratory demonstrating diastolic dysfunction, increased LV fibrosis, and reduced LV contractile reserve in this same model (13, 14, 20, 34, 40, 41), these findings indicate the relevance of this translational model to the pathophysiology of HFpEF. Nevertheless, it is difficult to precisely replicate HFpEF in animal models given the significant aging and sex components associated with the disease that are often accompanied by numerous clinical comorbidities (47, 48). From a practical perspective, aging pigs to true senescence is difficult given their relatively long lifespan (~15 yr) (51). Although this study focused on males, we have also examined aortic-banded ovariectomized female swine as a distinct translationally relevant HFpEF model (40) given the prevalence of HFpEF is twice as high in females (47, 48). Comorbidities common to HFpEF also complicate interpretation of experimental findings, and for this reason we consider a strength of this animal model is the ability to examine mechanisms of cardiac dysfunction independent of confounding disease processes. Systemic hypertension and aortic stenosis also contribute to the development of HFpEF, and while our surgical intervention creates cardiac pressure overload, it does not completely reflect the impact of systemic high blood pressure or decreases in arterial compliance seen in aortic stenosis. Finally, therapeutic exercise regimens are typically not prescribed concurrent with excessive afterload, as patients under the care of physicians are often treated to reduce existing hypertension. Thus, while our experimental exercise modalities do not exactly mimic existing clinical paradigms, our results suggest that exercise under these conditions is not necessarily detrimental and may exert positive cellular benefits on the heart.

In summary, our data provide evidence that chronic low-intensity endurance exercise training is an effective therapeutic alternative that beneficially impacts impaired cardiomyocyte Ca²⁺ homeostasis, contractile function, and β-adrenergic sensitivity in a large animal model of pressure overload-induced heart failure. Importantly, these results show intrinsic impairments to cardiomyocyte ECC in an experimental setting with relevance to human HFpEF and imply potentially important pathological mechanistic differences compared with HFrEF. Considering the lack of effective pharmacological therapies for HFpEF patients, our findings provide a strong rationale regarding clinical implementation of similar exercise programs for treating this heart failure subtype.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grants R01-HL-112998 (to C. A. Emter) and R01-HL-136292 (to T. L. Domeier) and National Institute on Aging Grant K01-AG-041208 (to T. L. Domeier).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.A.H., C.A.E., and T.L.D. conceived and designed research; J.A.H., M.D.L., T.D.O., B.S.F., and C.A.E. performed experiments; J.A.H., A.B.V., M.D.L., K.S.M., C.A.E., and T.L.D. analyzed data; J.A.H., A.B.V., T.D.O., B.S.F., K.S.M., C.A.E., and T.L.D. interpreted results of experiments; J.A.H., A.B.V., M.D.L., C.A.E., and T.L.D. prepared figures; J.A.H., A.B.V., C.A.E., and T.L.D. drafted manuscript; J.A.H., A.B.V., M.D.L., T.D.O., B.S.F., K.S.M., C.A.E., and T.L.D. edited and revised manuscript; J.A.H., A.B.V., M.D.L., T.D.O., B.S.F., K.S.M., C.A.E., and T.L.D. approved final version of manuscript.