Impact of heart rate on cross-bridge cycling kinetics in failing and nonfailing human myocardium

Abstract

The force-frequency relationship (FFR) is an important regulatory mechanism that increases the force-generating capacity as well as the contraction and relaxation kinetics in human cardiac muscle as the heart rate increases. In human heart failure, the normally positive FFR often becomes flat, or even negative. The rate of cross-bridge cycling, which has been reported to affect cardiac output, could be potentially dysregulated and contribute to blunted or negative FFR in heart failure. We recently developed and herein use a novel method for measuring the rate of tension redevelopment. This method allows us to obtain an index of the rate of cross-bridge cycling in intact contracting cardiac trabeculae at physiological temperature and assess physiological properties of cardiac muscles while preserving posttranslational modifications representative of those that occur in vivo. We observed that trabeculae from failing human hearts indeed exhibit an impaired FFR and a reduced speed of relaxation kinetics. However, stimulation frequencies in the lower spectrum did not majorly affect cross-bridge cycling kinetics in nonfailing and failing trabeculae when assessed at maximal activation. Trabeculae from failing human hearts had slightly slower cross-bridge kinetics at 3 Hz as well as reduced capacity to generate force upon K⁺ contracture at this frequency. We conclude that cross-bridge kinetics at maximal activation in the prevailing in vivo heart rates are not majorly impacted by frequency and are not majorly impacted by disease.

NEW & NOTEWORTHY In this study, we confirm that cardiac relaxation kinetics are impaired in filing human myocardium and that cross-bridge cycling rate at resting heart rates does not contribute to this impaired relaxation. At high heart rates, failing myocardium cross-bridge rates are slower than in nonfailing myocardium.

INTRODUCTION

The force-frequency relationship (FFR) is an important regulatory mechanism that increases the force-generating capacity as well as speeds up the contraction and relaxation kinetics in human cardiac muscle as the heart rate increases (3, 29). Hearts of large mammals, including humans, increase their cardiac output primarily via increasing their heart rate, and patients with heart failure typically become symptomatic as they physically exert themselves (e.g., exercising or walking a few blocks) (15). This clinical observation is supported by the fact that failing hearts have been reported to have no increase or even a decrease in developed force as frequency is increased (8, 9, 12–14, 20–23, 26). There may be multiple pathological mechanisms at play, including impaired calcium handling at the level of L-type calcium channels and SERCA, decreased intracellular sodium concentration, and increased calcium sensitivity (2, 6, 20, 21, 31, 32).

The rate of cross-bridge cycling, which has been reported to affect cardiac output, could be potentially dysregulated and contribute to blunted or negative FFR in heart failure. One study found decreased cross-bridge cycling kinetics in skinned right ventricular trabeculae isolated from patients with dilated cardiomyopathy (11). However, to our knowledge, the relationship between stimulation frequency and the rate of cross-bridge cycling has not been elucidated, especially in the context of intact human tissue and heart failure. We recently developed and herein use a novel method for measuring the rate of tension redevelopment (K_tr). This allows us to obtain an index of the rate of cross-bridge cycling in intact cardiac trabeculae at physiological temperature and assess physiological properties of cardiac muscles while preserving posttranslational modifications representative of those that occur in vivo (17, 18). Here, we set out to determine the frequency-dependent changes in cross-bridge cycling kinetics in intact nonfailing and failing left ventricular human trabeculae.

METHODS

Human heart procurement.

Procurement of all of the human hearts used in this study has been approved by The Ohio State University Institutional Review Board. Failing hearts were obtained from patients with end-stage heart failure who received a heart transplant (n = 15). All patients gave informed consent. Nonfailing hearts were obtained from collaboration with the Lifeline of Ohio Organ Procurement (n = 16) after they were rejected for transplantation. Immediately after the hearts were removed from patients/donors, they were coronary perfused with ice-cold cardioplegic solution containing (in mM) 110 NaCl, 0.5 CaCl₂, 16 KCl, 16 MgCl₂·6H₂O, and 10 NaHCO₃ (pH 7.4). Within ~20 min after explantation, the hearts were transported into the laboratory in ice-cold cardioplegic solution. Hearts from patients with hepatitis B or C or HIV, those labeled as high risk because of intravenous drug abuse, or those from patients under the age of 18 yr were excluded.

Human trabeculae preparation.

After the human hearts were transported into the laboratory, the ventricles were placed into an ice-cold modified Krebs-Henseleit (KH) solution previously bubbled with 95% O₂-5% CO₂ containing (in mM) 137 NaCl, 5 KCl, 0.25 CaCl₂, 20 NaHCO₃, 1.2 maH₂PO₄, 1.2 mM MgSO₄, 10 mM dextrose, and 20 mM 2,3-butanedione monoxime at pH 7.4. Linear, thin left ventricular (LV) trabeculae were isolated. The isolated trabeculae were kept at 4°C before the experiments were started. The trabeculae were mounted on a custom-made bath and connected at one end to a force transducer (KG7A, Scientific Instruments Heidelberg, 500-Hz resonance frequency) and to a linear motor (Scientific Instruments Heidelberg) at the other end. To avoid buckling of the preparation during the K_tr, instead of the normally used basket-shaped extension to the force transducer, we used a small stainless-steel needle extension with an insect pin for mounting of the muscle, which lowered the resonance frequency to ~280–330 Hz but was still sufficiently fast to capture the K_tr. We used an electronic signaling anti-oscillation unit to improve the signal-to-noise ratio. The muscles were initially stimulated at 0.5 Hz in a 37°C KH solution bubbled with 95% O₂-5% CO₂. The concentration of CaCl₂ was gradually increased over 20 min to the final concentration of 2 mM. The trabeculae were stretched to their optimal length at which the increase in developed tension is equal to the increase in resting tension. After reaching optimal length, the LV trabeculae were stabilized for 15 min, and their dimensions (width, thickness, and length) were measured in the experimental bath on an inverted microscope.

Twitch kinetics experiments.

After stabilization of contractile parameters of the trabeculae, we tested the FFR. In a subset of the trabeculae (2 out of 7 in nonfailing; 4 out of 6 in failing), we measured twitch forces and kinetics at 0.5 Hz, 1 Hz, 2 Hz, and 3 Hz before K_tr experiments (Table 1). These twitch data were combined with data from other LV trabeculae that did not undergo K_tr maneuvers (1 per heart) to generate a bigger data set in which twitch forces and kinetics in nonfailing (n = 11) and failing (n = 12) hearts were measured at 0.5-Hz, 1-Hz, 2-Hz, and 3-Hz stimulation frequencies. Because of poor curve fitting, the rates of contraction (dF/dt and dF/dt/F) and relaxation (negative dF/dt and negative dF/dt/F) were only measured in 9 nonfailing and 10 failing trabeculae. Average dimensions of nonfailing (n = 11) trabeculae were 0.42 ± 0.03 mm in width, 0.28 ± 0.02 mm in thickness, and 3.3 ± 0.40 mm in length. Failing (n = 12) trabeculae were 0.66 ± 0.08 mm in width, 0.46 ± 0.05 mm in thickness, and 2.62 ± 0.18 mm in length.

Table 1.

Twitch characteristics of muscles used for kinetics analysis

	Dev F, mN/mm2	Rest F, mN/mm2	TTP, ms	dF/dt, mN·mm−2·s−1	Positive dF/dt/F, s−1	RT50, ms	RT90, ms	Negative dF/dt, mN·mm−2·s−1	Negative dF/dt/F, s−1
Nonfailing
0.5 Hz	14.1 ± 4.1	5.8 ± 0.8	215.5 ± 12.5	108.6 ± 31.7	7.0 ± 0.5	139.5 ± 6.9	287.0 ± 11.8	−89.1 ± 29.9	−5.5 ± 0.3
1 Hz	14.9 ± 4.6	5.3 ± 0.7	195.8 ± 9.5	131.1 ± 37.7	8.3 ± 0.5	128.1 ± 6.1	259.3 ± 11.7	−102.4 ± 33.4	−5.9 ± 0.3
2 Hz	17.9 ± 6.1	5.1 ± 0.6	160.6 ± 5.9†‡	204.4 ± 69.6	10.3 ± 0.4†	106.8 ± 4.6†‡	208.2 ± 8.7†‡	−149.4 ± 49.0	−7.3 ± 0.4†‡
3 Hz	14.2 ± 4.5	6.0 ± 0.8	142.6 ± 4.3†‡	185.5 ± 57.8	11.0 ± 0.6†‡	89.5 ± 5.1†‡	157.3 ± 6.0†‡§	−140.2 ± 41.7	−8.2 ± 0.6†‡
Failing
0.5 Hz	11.5 ± 2.2	7.4 ± 1.4	208.3 ± 9.5	84.5 ± 16.2	7.7 ± 0.6	146.5 ± 6.3	299.7 ± 10.1	−57.4 ± 12.3	−4.9 ± 0.2
1 Hz	10.2 ± 1.7	6.9 ± 1.3	191.5 ± 8.5	84.0 ± 14.6	8.7 ± 0.5	138.2 ± 4.2	278.2 ± 7.3	−54.6 ± 10.4	−5.4 ± 0.3
2 Hz	8.2 ± 1.6	7.0 ± 1.1	160.6 ± 5.4†‡	80.7 ± 15.2*	10.2 ± 0.4†	116.0 ± 3.9†‡	219.8 ± 7.2†‡	−51.1 ± 10.6*	−6.5 ± 0.3†
3 Hz	5.7 ± 1.3	8.0 ± 1.1	137.8 ± 4.6†‡	66.2 ± 14.8*	10.7 ± 0.5†‡	100.6 ± 4.0†‡	171.8 ± 4.0†‡§	−42.7 ± 10.4*	−6.5 ± 0.5†

Values are means ± SE; n = 11 nonfailing; n = 12 failing (1 trabecula per heart). For first derivative of force (dF/dt), first derivative of force normalized to developed force (dF/dt/F), negative dF/dt, and negative dF/dt/F, n = 9 nonfailing and n = 10 failing. Dev F, developed force; −dF/dt/F = maximal velocity of relaxation; dF/dt/F, maximal velocity of contracture; Rest F, resting force; RT₅₀, time from peak force to 50% relaxation; RT₉₀, time from peak force to 90% relaxation; TTP, time-to-peak force.

^*P < 0.05 nonfailing vs. failing after two-way ANOVA with Bonferroni post hoc test between nonfailing and failing trabeculae;

^†P < 0.05 vs. corresponding 0.5-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test;

^‡P < 0.05 vs. corresponding 1-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test;

^§P < 0.05 vs. corresponding 2-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test.

Rate of K_tr experiments.

Nonfailing (n = 7) and failing (n = 6) LV trabeculae were stimulated at each frequency (0.5 Hz, 1 Hz, 2 Hz, or 3 Hz) for 3 min, and we then induced a K⁺ contracture via exposure to a modified KH solution containing 121.4 mM KCl, 20.5 NaCl, 10 mM dextrose, 20 mM NaHCO₃, 1.2 mM MgSO₄, 1.2 NaH₂PO₄, and 6 mM CaCl₂. At the maximal force of K⁺ contracture, we performed a motor maneuver previously developed in our laboratory in which the trabeculae were slacked by 20% of their length in 1 ms, held for 10 ms, and rapidly restretched back to their original length in 1 ms (18). The trabeculae were then relaxed in KH solution, and stimulation frequency was changed. This process was repeated for three other stimulation frequencies, and the order of the frequencies was randomized to exclude any time-dependent change of K_tr. Average dimensions of nonfailing trabeculae included in this subset were 0.38 ± 0.04 mm in width, 0.25 ± 0.03 mm in thickness, and 1.69 ± 0.5 mm in length. Average dimensions of failing trabeculae in the subset were 0.63 ± 0.09 mm in width, 0.42 ± 0.06 mm in thickness, and 2.64 ± 0.59 mm in length.

Data and statistical analysis.

Twitch force and kinetics were measured and analyzed via custom-written LabVIEW programs. The specific muscle forces were determined by normalizing the total tension to the cross-sectional area and expressed as mN/mm². The K_tr was determined by fitting the K_tr tracing to the equation F = F_max × (1−e^-ktr·t) + F_initial, using Origin 7 software (OriginLab). Statistical analysis was performed via ANOVA followed by post hoc tests with Bonferroni correction where appropriate. Statistical significance was set at two-tailed P < 0.05, and data in the figures are shown as means ± SE.

RESULTS

Developed tension and FFR in nonfailing and failing hearts.

Many laboratories in the past have shown that developed tension increases as frequency increases in nonfailing myocardium (20–23). On the other hand, developed tension has been reported to be unchanged or even negative as frequency increases in failing myocardium (2, 19, 21–23). Our data in intact trabeculae in this study support these observations. In general, nonfailing trabeculae displayed a positive FFR as frequency was increased from 0.5 to 3 Hz., whereas developed tension at 2 Hz was significantly smaller in failing trabeculae (P = 0.033) but did not reach statistical significance at 3 Hz (P = 0.079) (Fig. 1A). To give equal weight to each muscle and investigate the change in developed tension in each muscle, we normalized developed tension to that at 0.5 Hz. Failing trabeculae had lower normalized developed force at 2 Hz and 3 Hz compared with nonfailing trabeculae (P = 0.033 and 0.004, respectively) (Tables 2 and 3 and Fig. 1B).

Fig. 1.

Force-frequency relationship and resting tension in nonfailing and failing trabeculae. A: developed tension is lower in failing trabeculae at 2 and 3 Hz. B: developed tension normalized to that at 0.5 Hz shows negative force-frequency relationship in failing trabeculae. C: resting tension is higher in failing trabeculae across the tested frequencies, but not at specific frequencies. *P < 0.05 nonfailing vs. failing after two-way ANOVA with Bonferroni post hoc test between nonfailing and failing trabeculae. #P < 0.05 nonfailing vs. failing after two-way ANOVA between nonfailing and failing trabeculae. n = 11 nonfailing, and n = 12 failing (1 trabecula per heart). $P < 0.05 vs. corresponding 0.5-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test. ΦP < 0.05 vs. corresponding 1-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test.

Table 2.

Characteristics of nonfailing hearts

Heart	Age	Sex	Race	Cause of Death	Heart Weight, g	Ktr?	Twitch Kinetics?
364587	19	Male	Caucasian	Blunt injury/MVA	300	X	X*
266541	57	Male	Caucasian	ICB/ICH	635	X
402879	54	Male	Caucasian	ICB/ICH	474	X
984478	54	Female	African American	ICB/ICH	348	X
435578	20	Male	Caucasian	Drug overdose	324	X
460025	69	Female	Caucasian	CVA/ICH	435	X	X
958987	40	Male	Caucasian	Drug overdose/cardiac arrest	615	X
481043	65	Female	Caucasian	CVA/ICH	451		X
380071	43	Female	Caucasian	Anoxia, respiratory arrest	603		X
476074	29	Female	Caucasian	Anoxia, cardiac arrest	271		X
481041	41	Male	African American	CVA/ICH	455		X
168021	62	Female	Caucasian	CVA/ICH	896		X*
768159	44	Male	African American	Cardiac arrest	279		X
600245	51	Female	Caucasian	CVA/ICH	507		X
685884	36	Male	Caucasian	Anoxia, drug intoxication, cardiac arrest	415		X
435578	20	Male	Caucasian	Drug overdose, anoxia	324		X

Six-digit heart numbers are codes, not medical record numbers. “X” indicates yes. CVA, cerebrovascular accident; ICB, intracerebral bleeding; ICH, intracerebral hemorrhage; K_tr, tension redevelopment; MVA, motor vehicle accident; NA, not assessed/not available.

^*Hearts in which first derivative of force (dF/dt), first derivative of force normalized to developed force (dF/dt/F), negative dF/dt, and negative dF/dt/F were not measured because of poor curve fit.

Table 3.

Characteristics of failing hearts

Heart	Age	Sex	Race	Etiology	Heart Weight, g	LVAD?	Ktr?	Twitch Kinetics?
369452	61	Male	African American	NICM	540		X
611422	68	Male	Caucasian	ICM	576		X	X
631231	49	Female	Asian	NICM	306
679533	47	Male	Caucasian	NICM	930	X		X
963542	60	Female	Caucasian	NICM	329
497522	62	Female	Caucasian	NICM	649	X	X	X
994744	53	Male	Caucasian	NICM	393	X	X	X
335581	40	Female	Caucasian	NICM	715		X	X
554147	56	Male	Caucasian	NICM	474	X	X	X
834671	51	Female	Caucasian	NICM	572			X
214010	64	Male	Caucasian	ICM	495			X
323104	63	Male	Caucasian	NICM	543			X
328163	63	Male	Caucasian	ICM	506			X*
963542	60	Female	Caucasian	NICM	329			X
569897	48	Male	Caucasian	NICM	401	X		X

Six-digit heart numbers are codes, not medical record numbers. All failing hearts had EF <25%. “X” indicates yes. ICM, ischemic cardiomyopathy; K_tr, tension redevelopment; LVAD, left ventricular assist device; NICM, nonischemic cardiomyopathy.

^*Hearts in which first derivative of force (dF/dt), first derivative of force normalized to developed force (dF/dt/F), negative dF/dt, and negative dF/dt/F were not measured because of poor curve fit.

Resting tension in nonfailing and failing trabeculae.

Resting tension in nonfailing and failing trabeculae did not change significantly as frequency was increased. Failing trabeculae had greater resting tension overall, possibly because of increased fibrosis or myofilament posttranslational modification, but not at particular frequencies (ANOVA, P < 0.05) (Fig. 1C).

Contraction kinetics in nonfailing and failing trabeculae.

In both nonfailing and failing trabeculae, time-to-peak (TTP) decreased as stimulation frequency was increased (Fig. 2A). However, there were no significant differences in TTP between nonfailing and failing trabeculae at any of the frequencies we tested. Nonfailing trabeculae exhibited no significant changes in dF/dt, the maximal rate of force development, as frequency was increased from 0.5 to 3 Hz (Fig. 2B). Failing trabeculae also displayed the same pattern. However, when nonfailing and failing trabeculae were compared, failing trabeculae had significantly slower dF/dt compared with nonfailing trabeculae at 2 and 3 Hz (P = 0.004 and 0.011, respectively). We divided dF/dt by the developed tension to calculate a true kinetic parameter in the unit s⁻¹, dF/dt/F. Nonfailing and failing trabeculae exhibited increased dF/dt/F as frequency was increased from 0.5 to 3 Hz (Fig. 2C). However, there were no significant differences in dF/dt/F between nonfailing and failing trabeculae at the frequencies we tested.

Fig. 2.

Contraction kinetics in nonfailing and failing trabeculae. A: time-to-peak (TTP) becomes shorter in both groups as frequency is increased. B: first derivative of force (dF/dt) is slower in failing trabeculae at 2 and 3 Hz. C: first derivative of force normalized to developed force (dF/dt/F) accelerates in both groups as frequency is increased. $P < 0.05 vs. corresponding 0.5-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test. ΦP < 0.05 vs. corresponding 1-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test. n = 11 nonfailing, and n = 12 failing for TTP (1 trabecula per heart). n = 9 nonfailing, and n = 10 failing for dF/dt and dF/dt/F (1 trabecula per heart). #P < 0.05 between interaction of frequency and etiology.

Relaxation kinetics in nonfailing and failing trabeculae.

Time to 50% relaxation (RT₅₀) decreased as frequency was increased in both nonfailing and failing trabeculae and were not significantly different at any of the frequencies we tested (Fig. 3A). Across the frequencies we tested, failing trabeculae had slower RT₅₀ compared with nonfailing trabeculae (ANOVA, P = 0.014). Time to 90% relaxation (RT₉₀) decreased as stimulation frequency was increased in both nonfailing and failing trabeculae (Fig. 3B). Across the frequencies we tested, failing trabeculae had slower RT₉₀ compared with nonfailing trabeculae (ANOVA, P = 0.026). However, there was no significant difference between RT₉₀ in nonfailing and failing trabeculae at specific frequencies. In both nonfailing and failing trabeculae, negative dF/dt, the maximal rate of force decay, did not change significantly as frequency was increased (Fig. 3C). Overall, failing trabeculae had slower negative dF/dt (ANOVA, P < 0.0001), especially at 2 and 3 Hz (P = 0.005 and 0.01, respectively). Negative dF/dt/F accelerated as stimulation frequency increased in nonfailing trabeculae and in failing trabeculae (Fig. 3D). Failing trabeculae had slower negative dF/dt/F (ANOVA, P = 0.001), especially at 3 Hz (P = 0.006).

Fig. 3.

Relaxation kinetics in nonfailing and failing trabeculae. A: time from peak force to 50% relaxation (RT₅₀) becomes shorter as frequency is increased in both groups. B: time from peak force to 90% relaxation (RT₉₀) becomes shorter as frequency is increased in both groups. C: negative first derivative of force (dF/dt) is slower in failing trabeculae at 2 and 3 Hz. D: negative first derivative of force normalized to developed force (dF/dt/F) accelerates as frequency is increased in both groups. At 3 Hz, negative dF/dt/F is slower in failing trabeculae. *P < 0.05 nonfailing vs. failing after two-way ANOVA with Bonferroni post hoc test between nonfailing and failing trabeculae. #P < 0.05 nonfailing vs. failing after two-way ANOVA between nonfailing and failing trabeculae. $P < 0.05 vs. corresponding 0.5-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test. ΦP < 0.05 vs. corresponding 1-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test. δP < 0.05 vs. corresponding 2-Hz parameter in the same group after two-way ANOVA with Bonferroni post hoc test. n = 11 nonfailing, and n = 12 failing for RT₅₀ and RT₉₀ (1 trabecula per heart). n = 9 nonfailing, and n = 10 failing for negative dF/dt/F (1 trabecula per heart).

Cross-bridge cycling kinetics in nonfailing and failing trabeculae.

Nonfailing and failing muscles were stabilized at each stimulation frequency for 3 min and activated via K⁺ contracture. At maximal activation, K_tr was measured. In nonfailing and failing LV trabeculae, there were no significant changes in K_tr as frequencies were changed (Fig. 4A). Overall, failing trabeculae had slower K_tr (ANOVA, P = 0.01), especially at 3 Hz (P = 0.008) (Fig. 5). Overall, failing trabeculae had lower maximal forces during K⁺ contracture compared with nonfailing trabeculae, and these differences were statistically significant at 3 Hz (P = 0.036) and potentially at 2 Hz (P = 0.05) (Fig. 4B). There were no significant changes in maximal forces during K⁺ contracture as stimulation frequencies were changed in nonfailing and failing trabeculae.

Fig. 4.

Cross-bridge cycling kinetics in nonfailing and failing trabeculae. A: tension redevelopment (K_tr) does not change as frequency is increased in both groups. At 3 Hz, K_tr is slower in failing trabeculae. B: maximal K⁺ contracture tension does not depend on frequency, and it is smaller in failing trabeculae at 3 Hz. *P < 0.05 nonfailing vs. failing after two-way ANOVA with Bonferroni post hoc test between nonfailing and failing trabeculae. #P < 0.05 nonfailing vs. failing after two-way ANOVA between nonfailing and failing trabeculae. n = 7 nonfailing, and n = 6 failing (1 trabecula per heart).

Fig. 5.

Sample 3-Hz tension redevelopment (K_tr) tracings of nonfailing and failing trabeculae. Failing trabeculae have slower K_tr at 3 Hz compared with nonfailing trabeculae.

DISCUSSION

We have found that in intact human LV trabeculae: 1) failing trabeculae exhibit a negative FFR; 2) at optimal length, failing trabeculae have increased resting tension compared with nonfailing trabeculae, and resting tension does not change as frequency is increased; 3) kinetics of contraction (TTP and dF/dt/F) accelerate as frequency is increased, and they were not different between nonfailing and failing trabeculae; 4) kinetics of relaxation (RT₅₀, RT₉₀, and negative dF/dt/F) accelerate as frequency increases, and these parameters are slower overall in failing trabeculae; failing trabeculae have slower negative dF/dt/F at 3 Hz; 5) frequency does not affect cross-bridge cycling kinetics in nonfailing and failing trabeculae at maximal activation, and failing trabeculae have slower cross-bridge kinetics, especially at 3 Hz; and 6) failing trabeculae have reduced capacity to generate force upon K⁺ contracture at high frequencies.

Cardiac muscle utilizes three primary mechanisms to modulate cardiac output and kinetics of contraction and relaxation: 1) length-dependent activation, 2) β-adrenergic activation, and 3) frequency-dependent activation (1). This study focuses on the effect of frequency on cardiac muscle cross-bridge cycling kinetics, which is deeply implicated in contraction and relaxation kinetics of the muscle. Most previous studies on cross-bridge cycling kinetics have utilized permeabilized cardiac muscle often at room temperature, which does not permit investigation on the effect of stimulation frequency, as permeabilized muscles lack membrane and voltage-gated ion channels. We have performed all of our experiments in intact LV trabeculae at physiological temperature in an attempt to best replicate in vivo conditions in our ex vivo experiments.

Our laboratory has previously reported negative FFR in failing right ventricular trabeculae (3, 17). In this study, we have confirmed this finding in failing LV trabeculae. Myocardium relaxation is not a passive process (1), and resting tension is, specifically at high frequencies, a parameter that reflects the relaxation capacity in trabeculae (24). We did not observe significant changes in resting tension as frequency was increased. Failing trabeculae exhibited increased resting tension across the tested frequencies but not at specific frequencies.

We found that contraction kinetics (TTP and dF/dt/F) accelerate in both nonfailing and failing trabeculae when frequency was increased, as cardiac trabeculae must develop tension more quickly at faster heart rates. We did observe that dF/dt is slower in failing trabeculae. However, dF/dt depends on developed force, and the differences between nonfailing and failing were no longer observed once dF/dt was divided by developed force. Overall, we did not observe differences between nonfailing and failing trabeculae in their ability to accelerate contraction kinetics. These results suggest that acceleration of contraction kinetics is not necessarily impaired in heart failure.

As the heart beats faster, not only does it have to contract faster but it also has to relax faster. Our kinetics data support this concept, as relaxation kinetics parameters (RT₅₀, RT₉₀, and negative dF/dt/F) accelerate as frequency is increased independent of β-adrenergic effect, which speeds up the sinoatrial firing rate to increase the heart rate in vivo. Nonfailing and failing trabeculae were both able to accelerate their relaxation kinetics to a certain extent, but failing trabeculae had a limited capacity to accelerate their relaxation kinetics compared with nonfailing trabeculae, especially negative dF/dt/F at 3 Hz.

We assessed cross-bridge cycling and twitch kinetics in nonfailing and failing trabeculae stimulated at various stimulation frequencies and found that K_tr does not depend on frequency in both nonfailing and failing trabeculae at maximal activation. To our knowledge, there is no other published work on frequency-dependent cross-bridge kinetics, and it is possible that cross bridges are already cycling at maximal speed when the trabeculae are maximally activated by K⁺ contractures. If this is true, cross-bridge cycling kinetics would not depend on frequency. We suspect that cross-bridge cycling kinetics may depend on frequency if the muscle is submaximally activated, as posttranslational modifications of various proteins in the myofilament could occur. Our findings from this study suggest that cross-bridge cycling kinetics do not play a significant role in negative FFR observed in failing myocardium, at least at 2 Hz.

Perhaps cross-bridge cycling does play a role in negative FFR when stimulated at 3 Hz, as failing trabeculae had slower cross-bridge cycling kinetics and lower K⁺ contracture force at 3 Hz. However, this study suggests that cross-bridge cycling is not the primary mechanism of the FFR, at least when cardiac muscle is maximally activated. Other factors such as calcium mishandling by L-type calcium channels or SERCA or sodium mishandling may play more important roles in the dysregulated FFR in heart failure. Since we used intact multicellular trabeculae, cytosolic enzymes such as Ca²⁺/calmodulin-dependent protein kinase II (CaMKII), which phosphorylates myofilament proteins such as myosin-binding protein-C and myosin light chain-2, may play an important role in the regulation of cross-bridge cycling at various frequencies (4, 25, 27, 28, 30). CaMKII is a possible mediator of frequency-dependent acceleration of relaxation through a frequency-dependent increase in cytosolic Ca²⁺ that results in CaMKII activation (5). The rate of cytosolic Ca²⁺ reuptake into the sarcoplasmic reticulum, which involves phospholamban, could affect the level of CaMKII activation as well (16). Phospholamban Thr-17 is a phosphorylation target of CaMKII, and this process may be involved in the cross talk between CaMKII and sarcoplasmic reticulum in the frequency-dependent regulation of cytosolic Ca²⁺ levels and CaMKII activation (10).

In this study, we have found that K_tr is slightly slower in failing trabeculae, especially at 3 Hz. Interestingly, we also found slower negative dF/dt/F in failing trabeculae at 3 Hz. Cross-bridge cycling is one of three key mechanisms of cardiac muscle relaxation in addition to myofilament deactivation and intracellular decline (1). Although in this study we did not specifically test for the effect of cross-bridge cycling kinetics on relaxation kinetics, our findings suggest that the slower cross-bridge cycling may play an important role in impaired relaxation kinetics at high heart rates. However, other parameters of relaxation kinetics such as RT₅₀ and RT₉₀ were not significantly slower in failing trabeculae at 3 Hz. Nonfailing trabeculae had lower developed force at 0.5 Hz compared with failing trabeculae, but upon K⁺ contracture, nonfailing and failing trabeculae did not have different tension at 0.5 Hz (P = 0.569). Overall, failing trabeculae had lower K⁺ contracture force compared with nonfailing trabeculae. Actually, at 3 Hz, failing trabeculae had significantly lower K⁺ contracture tension compared with nonfailing trabeculae (P = 0.036).

One of the limitations of this study is that in the heart in vivo, increased frequency is coupled with increased β-adrenergic stimulation, and one must take both variables into account to make more physiologically relevant observations. However, in an attempt to study one variable at a time, this study focuses on frequency instead of looking at two variables at the same time. In addition, heart failure is a disease with multiple etiologies such as dilated cardiomyopathy, ischemic cardiomyopathy, and hypertrophic cardiomyopathy. Many of our failing trabeculae came from hearts broadly categorized as nonischemic cardiomyopathy, which itself has multiple etiologies. It is possible that there are small but potentially still relevant differences in cross-bridge cycling and twitch kinetics for different etiologies of heart failure. In addition, some of our failing hearts had an LV assist device placed in them before transplantation, which may have affected the function of the left ventricle, as it may undergo atrophy when an LV assist device replaces its contractile function.

GRANTS

This study was supported by National Institutes of Health Grants RC1-HL-099538, R01-HL-113084 (to P. M. L. Janssen), and R01-HL-132213 (to J. P. Davis) and American Heart Association Grant 16PRE33410549 (to J.-H. Chung).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

J.-H.C., J.P.D., and P.M.L.J. conceived and designed research; J.-H.C., N.M.-N., and B.A.W. performed experiments; J.-H.C., N.M.-N., and P.M.L.J. analyzed data; J.-H.C., N.M.-N., N.W., P.J.M., and P.M.L.J. interpreted results of experiments; J.-H.C. prepared figures; J.-H.C. drafted manuscript; J.-H.C. and P.M.L.J. edited and revised manuscript; J.-H.C., N.M.-N., J.P.D., N.W., B.A.W., P.J.M., and P.M.L.J. approved final version of manuscript.