Danicamtiv increases cardiac mechanical efficiency

Abstract

Abstract

Danicamtiv is a recently developed cardiac‐specific myosin activator that directly increases actomyosin interaction and, thereby, increases force. It is currently studied pre‐clinically and has entered clinical trials. In this study, we provide the first assessment of its effects on cardiac energetics. Ventricular trabeculae from rats were isolated and studied in vitro. Each trabecula was stimulated to produce twitch force over ranges of muscle lengths and afterloads, with muscle heat output simultaneously measured. Each trabecula was superfused with and without danicamtiv (5 µM). In both interventions, muscles were required to undergo both isometric and work‐loop contraction protocols to quantify the components of heat output – that from muscle contraction, that from intracellular Ca²⁺ cycling and that from actomyosin cross‐bridge cycling. Danicamtiv was found to increase twitch force, muscle heat and cross‐bridge heat, with no effect on the energy expenditure associated with intracellular Ca²⁺ cycling. Danicamtiv slowed isometric twitch force kinetics and slowed shortening kinetics. Danicamtiv increased the extent of muscle shortening, increased muscle work output and increased mechanical efficiency, while preserving cross‐bridge efficiency. In conclusion, danicamtiv increases mechanical efficiency, the mechanism of which arises from the greater increase of force and muscle shortening that increase mechanical work than the increase of heat. The relative smaller increase of heat is due to a null effect of danicamtiv on the energy for intracellular Ca²⁺ cycling. Our results provide perspective for the utility of danicamtiv as an inotropic agent not only for mechanical enhancement but also for energetic enhancement in increasing mechanical efficiency.

Key points

• Clinical trials are presently exploring the potential of a new myotrope, danicamtiv, to improve cardiac mechanics by increasing muscle force output.

• A range of active pre‐clinical research is likewise uncovering the biophysical mechanism of how danicamtiv affects cardiac mechanics, increases cardiac force and slows contractile kinetics.

• This study provides the first energetic assessment of danicamtiv by measuring muscle heat output, and shows that danicamtiv increases cardiac energy efficiency.

• The mechanism of the increased efficiency resides in the increase of muscle force–length work greater than the increase of muscle heat.

• This study demonstrates that danicamtiv is a cardiac myotrope that not only enhances mechanics, but also exerts its enhancement effect on energetics.

Abstract figure legend We studied intact cardiac muscles, contracting within our experimental device. We subjected each muscle in the device to perform force–length work‐loop contractions – contraction patterns that replicate the pressure–volume loop of the intact heart. From these contractions, we assessed muscle mechanoenergetic performance including twitch force, extent of shortening (the width of the loop), kinetics of contraction and of shortening, heat output, work production (the area of the loop), change of enthalpy (work + heat), and mechanical efficiency (the ratio of work to change of enthalpy). Each muscle was tested in both the absence and the presence of a new myotrope, danicamtiv. We found that danicamtiv enhanced all mechanoenergetic metrics except for contraction and shortening kinetics, which it reduced; notably, it increased mechanical efficiency of contraction.

Introduction

The contractile force of cardiac muscle can be augmented by inotropic agents including calcitropes and/or myotropes. Calcitropes increase intracellular Ca²⁺ availability to the contractile myofilaments. Myotropes directly activate the contractile actomyosin, particularly myosin and, hence, are often referred to as ‘myosin activators’. Conventional inotropic agents using calcitropes are reported to elicit adverse effects, including initiation of arrhythmias, induction of cell‐death pathways through Ca²⁺‐mediated mechanisms, and increased myocardial energy demand, particularly for intracellular Ca²⁺ cycling (Maack et al., 2018; Psotka et al., 2019). The contemporary strategy is then to directly target the contractile actomyosin using myotropes, which have no effect on Ca²⁺ handling (Bernier & Buckley, 2021; Ráduly et al., 2022) and, thus, can overcome the adverse effects of calcitropes. Specifically, myotropes can avoid the energetic consequences of calcitropes and, thereby, ‘increase force economically’ (Hasenfuss, 2020).

Omecamtiv mecarbil is the first‐in‐class myotrope. Mechanistic studies identified its mechanistic action as prolonging actomyosin attachment of a small population of non‐force‐generating myosin while activating the thin filament to recruit more drug‐free myosin for crossbridge cycling – the net effect is an increased number of myosin in the force‐generating state (Lehman et al., 2022).

Danicamtiv is the second, recently developed, myotrope. It, like omecamtiv mecarbil, also targets cardiac myosin and increases actomyosin activity. Since its development about half a decade ago, its mechanistic action has been investigated, and its clinical potential is presently being explored in phase II trials. Voors et al. (2020) studied dogs and patients with heart failure and demonstrated the effects of danicamtiv on increasing left‐ventricular systolic function and stroke volume, while prolonging systolic ejection time. These in vivo functional results were later supported in a study on healthy rats (Ráduly et al., 2022). Pharmacokinetic properties of danicamtiv have been characterized and the results support its advancement into clinical investigations (Grillo et al., 2021). In a study using X‐ray diffraction and mechanical measurements on porcine tissues, the biophysical mechanisms of how danicamtiv increases cardiac force and slows contractile kinetics were revealed to arise, respectively, via increasing myosin recruitment and decreasing ADP release rate resulting in prolonged actomyosin interaction (Kooiker et al., 2023). That is, danicamtiv slows cross‐bridge kinetics, both the rates of detachment and recruitment, in human myocardia (Choi et al., 2023, 2025). The inotropic effects and the slowing of cross‐bridge kinetics effects of danicamtiv were likewise obtained in engineered human myocardial tissues (Shen et al., 2021) and found to be sex‐independent in muscle strips from healthy rats (Awinda et al., 2024).

In this study, we present the first evaluation of the effects of danicamtiv on cardiac muscle energetics. Specifically, we aimed to test two existing assumptions, which required measurements of energy output: (1) that danicamtiv, like other myotropes, has a null effect on the energy demand for intracellular Ca²⁺ cycling, and (2) that danicamtiv increases force economically. More importantly, we aimed to characterize the energetic consequences of danicamtiv on the energy expenditure and efficiency of muscle, as well as of actomyosin cross‐bridges. We achieved this by conducting experiments measuring simultaneously the force, length and heat output of ventricular muscles isolated from healthy rats. Our experiments span ranges of muscle lengths and afterloads, enabling a thorough assessment of key relations between mechanical performance and energetic parameters.

Methods

Ethical approval

Animal handling and euthanasia were approved by the University of Auckland Animal Ethics Committee (Ethics 26534). Rats were purchased from the Vernon Jansen Unit – the laboratory animal facility of the University of Auckland – where rats were housed in a 12/12 h light–dark cycle at room temperature (22°C) with ad libitum access to standard chow and water.

Muscle preparation

Ventricular muscle tissues (trabeculae) from male rats were used in this study. We have recently shown that ventricular trabeculae isolated from male and female rats of the age range studied produced no sex‐dependent effects in muscle mechanoenergetic performance (Rahmani et al., 2024). The experiments were conducted under various conditions encompassing ranges of afterloads, muscle lengths and stimulus frequencies. We found that none of the indices of mechanoenergetic functional performance of trabeculae was different between sexes – that is, twitch kinetics, heat production, mechanical work output and mechanical efficiency are sex‐independent, at the trabecula level and within the rat age range studied (Rahmani et al., 2024).

Wistar rats (male, 10–13 weeks of age, 300–500 g of body mass) were anaesthetized with isoflurane (5% in O₂) and injected with heparin (1000 IU kg⁻¹) before euthanasia via cervical dislocation. The heart was then immediately excised (within 30 s) following a thoracotomy and plunged into chilled Tyrode solution. The aorta was promptly cannulated (within 30 s) to perfuse the heart with 100% oxygenated Tyrode solution (containing 130 mM NaCl, 6 mM KCl, 0.3 mM MgCl₂, 0.5 mM NaH₂PO₄, 10 mM HEPES, 10 mM glucose and 20 mM 2,3‐butanedione monoxime; pH adjusted to 7.4 using Tris) at room temperature.

The perfused heart was then cut open, with incisions on either side of the septum to open both the left and right ventricles. Trabeculae of appropriate size and uniformity were dissected. Each trabecula was transported in a 1 mL syringe to the mounting chamber of the work‐loop calorimeter (Taberner et al., 2011). Prior to muscle mounting, the calorimeter had been warmed to body temperature (37°C) using a heater and control system. Oxygenated Tyrode superfusate (the same ingredients as described above except for 1.5 mM CaCl₂ and without 2,3‐butanedione monoxime; pH pre‐adjusted to 7.4 for body temperature) flowed into the measurement chamber and exited past the mounting chamber of the calorimeter.

In the mounting chamber of the calorimeter, the trabecula was attached to two platinum hooks (the upstream one was connected to a linear length‐motor; the downstream one was connected to a force transducer; Han et al., 2009; Taberner et al., 2011). The mounted trabecula was then translated into the measurement chamber and positioned between the two thermopiles upstream and downstream. The muscle was superfused with the oxygenated Tyrode solution at an optimal rate of 0.6 µL s⁻¹ (Johnston et al., 2015; Taberner et al., 2018) while avoiding tissue hypoxia (Han et al., 2011). The trabecula was stimulated at 2 Hz, and its optimal length (L _o) was found by progressively stretching the muscle via the linear length‐motor. L _o was determined as the muscle length where further stretching resulted in no further increase in active force but a great elevation of passive force. At L _o, active force is maximal, and mechanical efficiency is also optimal (Han et al., 2022). Once the trabecula acclimatised as evident from reaching its steady‐state force, stimulus frequency was increased to 3 Hz. The entire calorimeter was covered to render it light‐proof and thermally insulated, with the interior temperature maintained at 37°C.

Danicamtiv dose‐response curve

The danicamtiv dose–response relationship in trabeculae was examined at 37°C to determine the optimal dosage for experimentation in the work‐loop calorimeter. These experiments were conducted using n = 3 trabeculae. Each trabecula was subjected to isometric contractions until steady state of twitch force was reached in Tyrode superfusate containing either danicamtiv dissolved in dimethyl sulfoxide (DMSO) or simply DMSO with 0 µM danicamtiv. Tyrode superfusate was switched to different danicamtiv (and DMSO) concentrations: 0 µM (0% DMSO), 1 µM (0.004% DMSO), 3 µM (0.012% DMSO), 5 µM (0.02% DMSO), 7 µM (0.028% DMSO) and 10 µM (0.04% DMSO). In between each danicamtive dose, the muscle was superfused with the corresponding vehicle DMSO concentration without danicamtiv. Thus, changes in twitch force were expressed relative to the corresponding vehicle control. Based on muscle stress and twitch duration (Fig. 1), 5 µM danicamtiv was selected as the optimal dose for all experimental procedures.

Figure 1. Danicamtiv dose‐dependent effects on isometric active stress and twitch duration.

A, overlaid steady‐state active stress twitches, each labelled by danicamtiv dose, from a representative trabecula. Active stress twitches from danicamtiv were scaled relative to corresponding vehicle control. The time axis ends at the terminal of the duration of 3 Hz twitches (i.e. 333 ms). B, average data points (n = 3 trabeculae per danicamtiv dose) fitted with sigmoidal regression (maximum relative active stress = 1.84, minimum relative active stress = 1.01, EC₅₀ = 4.14 µM and Hill coefficient = 5.62; R ² = 0.9975). C, average data points (same trabeculae as in panel B) fitted with sigmoidal regression (maximum relative twitch duration = 2.96, minimum relative twitch duration = 1.01, EC₅₀ = 9.03 µM and Hill coefficient = 1.76; R ² = 0.9992). Twitch durations (quantified at 5% of peak active stress) of danicamtiv dose are expressed relative to vehicle control. In panels B and C, the plots are presented as means ± SD, and sigmoidal regression fitted to the mean values. Large circles (black) denote mean values; small circles (grey) denote individual data points from n = 3 trabeculae × 6 danicamtiv doses. See Table 1 for statistical comparisons including individual P‐values.

Experimental protocols

In all mechanoenergetics experiments, 5 µM danicamtiv was used. The Tyrode superfusate was supplemented with 0.02% DMSO either with or without 5 µM danicamtiv. Each trabecula was exposed to both Tyrode solutions, the order of presentation of which was random. In total, n = 10 trabeculae received danicamtiv‐free control Tyrode first, followed by danicamtiv, while in the other n = 10 trabeculae the order was the converse. In each Tyrode solution, the trabeculae were subjected to both an isometric contraction protocol and a work‐loop contraction protocol, as described below.

Each trabecula first underwent a force–length work‐loop contraction protocol. The muscle performed work‐loops at six varying afterloads interchanged with isometric contractions (baseline measurements) to obtain steady‐state readings of muscle force and heat. A work‐loop has four phases: isometric contraction, isotonic shortening, isometric relaxation and re‐lengthening, where afterload is defined as the force where the muscle transitions from isometric contraction to isotonic shortening. The afterload studied ranged from minimum (in the vicinity of passive force) to maximum (approaching the isometric force). From this protocol, mechanoenergetics variables including force–length work output, shortening kinetics, shortening heat output (Tran et al., 2017) and efficiency as functions of afterload were examined.

In the second protocol, muscle performed isometric contractions at six different lengths ranging from the optimal length (L_o ) to the minimal length (where active force was minimal). Stimulation was halted between each of the six length steps to establish baselines of force and heat. From this protocol, the isometric force–length relation, isometric twitch force kinetics, and the heat–force relation were quantified, and the activation heat (Han et al., 2023; Pham et al., 2017) and crossbridge heat (Han, Tran et al., 2019) were examined.

At the completion of the two contraction protocols, heat measurements from the work‐loop calorimeter were corrected by subtracting two sources of artefacts. First, with the muscle quiescent by halting the stimulation, the length‐motor was oscillated at 3 Hz between muscle optimal length and the peak work‐loop shortening length to measure the heat artefact generated by the motor movement. Second, to measure the stimulus heat, the muscle was transferred downstream of the measurement chamber into the mounting chamber. In the absence of the muscle in the measurement chamber, the stimulator was turned on and stimulus heat was measured. On average, the motor‐movement heat signal was 0.038 ± 0.049 (mean ± SD; quantified at the maximal shortening of the muscle), and the stimulus heat signal was 0.049 ± 0.047 (mean ± SD), of the total heat signal measured when the muscle was contracting isometrically.

Muscle dimensions

Muscle dimensions at the optimal length (L _o) were measured with a microscope. A lens graticule was used to measure the length and diameter of the muscle from a top‐down view, and the side diameter of the muscle was observed with a 45° angle mirror embedded in the measurement chamber. Due to the elliptical shape of the muscle, the top and side diameters were both used to calculate the cross‐sectional area. In total, the 20 trabeculae had an average (mean ± SD) top diameter of 322 ± 115 µm, side diameter of 267 ± 72 µm, length at L _o of 4.57 ± 0.86 mm. Thus, they had a mean ± SD cross‐sectional area of 0.07 ± 0.04 mm², and volume of 0.33 ± 0.20 mm³.

Normalisation of data

Data were acquired using LabVIEW software (National Instruments, Austin, TX, USA) and analysed using a custom‐written MATLAB program (MathWorks Inc., Natick, MA, USA). Muscle stress was calculated by dividing muscle force by cross‐sectional area. Twitch duration for isometric twitches was quantified as the duration of time where stress rises above 5% of peak active stress until it falls back to 5%. Muscle length was expressed relative to the optimal length (L/L _o). Work‐loop afterload was expressed relative to peak isometric stress. Extent of shortening during work‐loops was estimated as the difference between the end‐systolic length and end‐diastolic length. Velocity of shortening was estimated from the length–time trace during the initial 10–20 ms isotonic shortening phase of each work‐loop. The product of velocity of shortening and afterload provides a measure of power of shortening. Work is the area of the work‐loop and was calculated by integrating stress as a function of muscle length throughout the entire twitch. Change of enthalpy is the total energy used to perform work with the remaining unused energy liberated as heat. It was thus taken as the sum of work and heat. Heat per twitch was calculated by dividing the measured rate of heat production by the stimulation frequency (3 Hz) and normalized to muscle volume. Mechanical efficiency is the ratio of muscle work output to muscle change of enthalpy. Crossbridge efficiency has a denominator term of crossbridge change of enthalpy, calculated by subtracting activation heat from muscle change of enthalpy.

Statistical analysis

Data from individual trabeculae were fitted using polynomial regressions (up to the third order), and the regression lines were averaged within groups using the ‘Random Coefficient Model’ within the PROC MIXED procedure (implemented in the SAS software package; SAS Institute, Cary, NC, USA). Statistical comparisons of the regression lines between the two Groups (with or without danicamtiv) were made with the ‘MODEL’ statement consisting of the Regression Equation, the ‘RANDOM’ statement consisting of Group, and the ‘REPEATED’ statement consists of Muscle. In contrast to comparison of regression lines, Student's t test (two‐tailed, paired) was used to compare values of mechanoenergetics variables of interest (either interpolated from the regression lines or measured at L _o) between the two groups (when the assumption of normality was met, as tested using Shapiro–Wilk test). One‐way ANOVA repeated measure with Tukey's post hoc test was used to compare between the two groups the data (expressed relative to vehicle control) obtained from the dose–response experiment. Values were expressed as means ± SD. Statistical significance was when P < 0.05.

Results

Danicamtiv increased active stress and twitch duration in a dose‐dependent manner

Application of danicamtiv on isometrically contracting trabeculae increased muscle active stress and twitch duration in a dose‐dependent manner (Fig. 1). Kinetics of the twitches as functions of danicamtiv dose are shown in Table 1. Significant differences in the twitch kinetics between with and without danicamtiv (with matched DMSO concentration) were detected at danicamtiv doses of 5 µM and greater. From this pilot experiment, 5 µM danicamtiv was selected for mechanoenergetics experiments. A 50–100 ms time period was required for the completion of work‐loop diastolic re‐stretch of muscle at 3 Hz stimulation frequency, and hence danicamtiv concentrations above 5 µM were inappropriate due to the prolonged twitch duration.

Table 1.

Kinetics of isometric twitches as functions of danicamtiv dose

	Danicamtiv dose (µM)
	1	3	5	7	10
Value relative to vehicle control
Active stress	1.01 ± 0.02	1.12 ± 0.09	1.64 ± 0.07*	1.77 ± 0.20*	1.86 ± 0.33*
Twitch duration 5% total	1.07 ± 0.08	1.25 ± 0.12	1.52 ± 0.16*	1.78 ± 0.22*	2.07 ± 0.17*
Twitch duration 5% rise	0.99 ± 0.02	1.17 ± 0.05	1.49 ± 0.03*	1.63 ± 0.20*	1.94 ± 0.26*
Twitch duration 5% fall	1.11 ± 0.12	1.30 ± 0.18	1.54 ± 0.23*	1.88 ± 0.26*	2.17 ± 0.31*
Twitch duration 50% total	1.00 ± 0.01	1.25 ± 0.08	1.62 ± 0.04*	1.91 ± 0.21*	2.44 ± 0.11*
Twitch duration 50% rise	1.00 ± 0.02	1.20 ± 0.05	1.58 ± 0.04*	1.75 ± 0.25*	2.20 ± 0.34*
Twitch duration 50% fall	1.00 ± 0.02	1.28 ± 0.11	1.65 ± 0.06*	2.04 ± 0.21*	2.65 ± 0.20*
dS/dt rise	1.04 ± 0.06	1.01 ± 0.06	1.20 ± 0.06*	1.26 ± 0.03*	1.19 ± 0.09*
dS/dt fall	0.98 ± 0.04	0.94 ± 0.10	1.14 ± 0.10	0.98 ± 0.11	0.81 ± 0.18
P‐value vs. vehicle control
Active stress	1.0000	0.9300	0.0049	0.0013	0.0005
Twitch duration 5% total	0.9829	0.2611	0.0067	0.0003	< 0.0001
Twitch duration 5% rise	1.0000	0.6738	0.0144	0.0027	0.0001
Twitch duration 5% fall	0.9697	0.3764	0.0386	0.0014	0.0001
Twitch duration 50% total	1.0000	0.1324	0.0003	< 0.0001	< 0.0001
Twitch duration 50% rise	1.0000	0.7354	0.0236	0.0044	0.0001
Twitch duration 50% fall	1.0000	0.1022	0.0004	< 0.0001	< 0.0001
dS/dt rise	0.9456	1.0000	0.0118	0.0017	0.0166
dS/dt fall	0.9999	0.9304	0.4197	0.9995	0.1418

Upper part: values are expressed relative to the corresponding DMSO concentration‐matched vehicle (i.e. in the absence of danicamtiv). Data are means ± SD from n = 3 trabeculae. Twitch durations (‘total’) are partitioned into ‘rise’ and ‘fall’ components, and are assessed at 5% and 50% of peak active stress. *P < 0.05 vs. corresponding DMSO concentration‐matched vehicle control; one‐way ANOVA repeated measure with Tukey's post hoc test. Lower part: P‐values vs. corresponding DMSO concentration‐matched vehicle control for each of the five danicamtiv doses.

Work‐loop and isometric contractions

Measurements of trabecula length, stress and heat rate were simultaneous during both the work‐loop contractions protocol and the isometric length‐change protocol (Fig. 2A ). In the work‐loop contractions protocol, steady‐state stress and heat rate were highest during isometric contraction (indicated by the label ‘a’) and reduced with decreasing afterloads (‘b’ and ‘c’). Muscle extent of shortening increased with decreasing afterloads. In the isometric length‐change protocol, muscle stress and heat rate decreased with decreasing muscle length (‘a’, ‘d’ and ‘e’). Parametric plots of muscle stress and length at steady‐state produced stress–length work‐loops (Fig. 2B ), where end‐systolic and end‐diastolic stress–length relations were obtained (Fig. 2B and  C ; red dotted lines, respectively, for work‐loop and isometric contractions).

Figure 2. Experimental records of a representative trabecula undergoing work‐loop and isometric contractions at 3 Hz.

A, trabecula length, stress and heat rate were measured simultaneously. In the work‐loop protocol (left‐hand side), upon reaching steady‐states of stress and heat rate, the contraction mode was switched from isometric contraction (‘a’) to work‐loop contractions (‘b’ and ‘c’, as examples). The trabecula performed six afterloaded work‐loop contractions. Initial length was at L _o. In the isometric length‐change protocol (right‐hand side), successive length reductions were applied (‘a’, ‘d’ and ‘e’, as examples). Insets show length change (top) and twitch stress (bottom) for one representative twitch at steady states, with ‘b’ and ‘c’ insets illustrating twitches divided into four work‐loop phases: 1, isometric contraction; 2, isotonic shortening; 3, isometric relaxation; and 4, diastolic re‐lengthening. Arrows on the heat rate record indicate the ‘on’ (upward) and ‘off’ (downward) of electrical stimulation. B, stress–length work‐loops (six different afterloads) overlaid with the isometric twitch (‘a’). Data from steady‐state twitches from panel A. The inset shows length change and individual twitch stress as functions of time. C, overlaid isometric stress at optimal length, L _o = 1 (labelled ‘a’) with those at reduced lengths (‘d’ and ‘e’). The inset shows length and individual twitch stress as functions of time. In both panels B and C, the top red dotted line represents cubic regression fitting to end‐systolic points (total stress); the bottom red dotted line represents cubic regression fitting to end‐diastolic points (passive stress).

Danicamtiv elevated twitch stress and heat

The average steady‐state end‐systolic stress–length relation and end‐diastolic (passive) stress–length relation (n = 20 trabeculae) from both work‐loop contractions (Fig. 3A ) and isometric contractions (Fig. 3B ) were all significantly higher in trabeculae when exposed to danicamtiv. Peak active twitch stress, obtained at L _o, was greater under danicamtiv (from 36.4 ± 21.5 to 46.0 ± 22.6 kPa; P = 0.0130). Peak passive stress, obtained at L _o, was greater under danicamtiv (from 11.4 ± 5.1 to 14.7 ± 7.0 kPa; P = 0.0170). Likewise, the average steady‐state heat–length relation from work‐loop contractions (Fig. 3C ) and that from isometric contractions (Fig. 3D ) were both significantly elevated following danicamtiv exposure (both P < 0.0001). Peak active heat, obtained at L _o, was greater under danicamtiv (from 8.3 ± 5.1 to 12.1 ± 5.7 kJ m⁻³; P = 0.0001).

Figure 3. Steady‐state stress–length and heat–length relations from work‐loop (WL) contractions (A and C) and isometric (Isom) contractions (B and D).

A, danicamtiv exposure significantly elevated work‐loop end‐systolic (P = 0.0011) and end‐diastolic (P = 0.0022) stress–length relations (top lines and bottom lines, respectively). Average regression lines (quadratic fitting; n = 20 trabeculae × 7 afterloads per trabecula; continuous line for danicamtiv and dashed line for control) are overlaid with 95% confidence intervals (grey, control; blue, danicamtiv); *P < 0.05. The inset displays data from a representative trabecula, and cubic fitting to the end‐systolic points (top) and end‐diastolic points (bottom), the same as in Fig. 2B red dotted lines. B, danicamtiv exposure significantly elevated total stress–length (P < 0.0001) and passive stress–length (P = 0.0046) relations (top lines and bottom lines, respectively). Average regression lines (cubic fitting; n = 20 trabeculae × 6 length‐steps per trabecula) are overlaid with 95% confidence intervals (grey, control; blue, danicamtiv). The inset shows data from the same representative trabecula, and cubic fitting to the total stress data points (top) and passive stress data points (bottom). C, the average work‐loop heat–length relation was elevated (n = 20 trabeculae × 7 afterloads per trabecula) when trabeculae were exposed to danicamtiv (P < 0.0001). The inset shows individual data points from the same representative trabecula, and linear fitting to the data. D, the average isometric heat–length relation was elevated (n = 20 trabeculae × 6 length‐steps per trabecula) when trabeculae were exposed to danicamtiv (P < 0.0001). The inset shows individual data points from the same representative trabecula, and data fitted with cubic regression. Statistical comparisons of the regression lines between the two Groups (Control and Danicamtiv) were made with the ‘MODEL’ statement consisting of the Regression Equation, the ‘RANDOM’ statement consisting of Group, and the ‘REPEATED’ statement consists of Muscle.

Danicamtiv slowed isometric twitch kinetics

Isometric twitch kinetics were assessed as twitch duration, stress–time integral, and rates of rise and fall of the twitch, plotted in Fig. 4A–D , respectively. These kinetics indices were fitted with linear functions of active stress. They were all significantly lower in trabeculae when exposed to danicamtiv (all P < 0.0001).

Figure 4. Isometric twitch kinetics.

Average regression lines (linear fitting; n = 20 trabeculae × 6 length‐steps per trabecula; continuous line for danicamtiv and dashed line for control) are overlaid with 95% confidence intervals (grey, control; blue, danicamtiv). Statistical comparisons of the regression lines between the two Groups (Control and Danicamtiv) were made with the ‘MODEL’ statement consisting of the Regression Equation, the ‘RANDOM’ statement consisting of Group, and the ‘REPEATED’ statement consists of Muscle. A, danicamtiv exposure significantly elevated the relation between twitch duration (quantified at 5% of peak active stress) and active stress (P < 0.0001). B, danicamtiv exposure significantly elevated the relation between stress–time integral (STI) and active stress (P < 0.0001). C, danicamtiv exposure significantly lowered the relation between rate of rise of twitch (dS/dt rise) and active stress (P < 0.0001). D, danicamtiv exposure significantly elevated the relation between rate of fall of twitch (dS/dt fall) and active stress (P < 0.0001), that is slowed kinetics. Insets show individual data points from a representative trabecula, and the data fitted with linear regression. The 95% confidence intervals in panels B–D are small.

Danicamtiv slowed shortening kinetics

From work‐loop contractions at various afterloads, muscle shortening (phase 2 of work‐loop; Fig. 2A inset) was assessed from the duration, the extent, the velocity and the power, plotted as functions of relative active afterload in Fig. 5A–D , respectively. The linear relation between the duration of shortening and relative active afterload was significantly elevated (P < 0.0001) following danicamtiv administration (Fig. 5A ). The extent of shortening had a quadratic relation with relative active afterload, and the relation was elevated (P < 0.0001) following danicamtiv administration (Fig. 5B ). The velocity of shortening was fitted with a cubic function of relative active afterload, and the relation was lowered (P < 0.0001) following danicamtiv administration (Fig. 5C ). The power of shortening showed no significant effect of danicamtiv (P = 0.8699) when assessed as a cubic function of relative active afterload (Fig. 5D ).

Figure 5. Work‐loop shortening kinetics.

Average regression lines (linear fitting; n = 20 trabeculae × 6 afterloads per trabecula; continuous line for danicamtiv and dashed line for control) are overlaid with 95% confidence intervals (grey, control; blue, danicamtiv). Statistical comparisons of the regression lines between the two Groups (Control and Danicamtiv) were made with the ‘MODEL’ statement consisting of the Regression Equation, the ‘RANDOM’ statement consisting of Group, and the ‘REPEATED’ statement consists of Muscle. A, danicamtiv exposure significantly elevated the linear relation between the duration of shortening (phase 2 of work‐loop; Fig. 2A inset) and relative active afterload (P < 0.0001). The inset shows individual data points from a representative trabecula, with the data point at the isometric contraction (i.e. at relative active afterload of 1, where the duration of shortening was 0) omitted for linear fitting. B, danicamtiv exposure significantly elevated the quadratic relation between the extent of shortening and relative active afterload (P < 0.0001). C, danicamtiv exposure significantly lowered the velocity of shortening as a cubic function of relative active afterload (P < 0.0001). D, the power of shortening as a cubic function of relative active afterload was not significantly affected by danicamtiv (P = 0.8699).

Danicamtiv increased cross‐bridge heat

Danicamtiv significantly increased the heat output of trabeculae as a function of active stress during both work‐loop (Fig. 6A ) and isometric (Fig. 6B ) contractions. These two contraction modes allowed partitioning of muscle heat into the heat associated with intracellular Ca²⁺ cycling (activation heat) and that with cross‐bridge cycling. Activation heat was not affected by danicamtiv (Fig. 6C ). The effects of danicamtiv were more noticeable on the cross‐bridges, where it increased the peak cross‐bridge heat (Fig. 6D ), reduced cross‐bridge economy (Fig. 6E ) and increased the peak shortening heat (Fig. 6F ).

Figure 6. Muscle heat, crossbridge heat and activation heat.

A, average regression lines relating muscle heat and active stress (linear fitting; n = 20 trabeculae × 7 afterloads per trabecula; continuous line for danicamtiv and dashed line for control) are overlaid with 95% confidence intervals (grey, control; blue, danicamtiv). Danicamtiv significantly elevated the relation (P < 0.0001). The inset shows individual data points from a representative trabecula, and the data fitted with linear regression. Data obtained from work‐loop contractions (WL) over a range of afterloads. B, danicamtiv significantly elevated the average relation (linear fitting; n = 20 trabeculae × 6 length‐steps per trabecula; continuous line for danicamtiv and dashed line for control) describing muscle active heat (Q _active) and active stress (S _active) (P < 0.0001), obtained under isometric (Isom) contractions over a range of muscle lengths. The y‐intercept of this relation (i.e. activation heat) was not affected (P = 0.5495), but the slope of this relation (the reciprocal is an index of cross‐bridge economy) was significantly elevated (P < 0.0001), by danicamtiv. For the Control group, the heat–stress relation is Q _active = (0.20 ± 0.04). S _active + (0.95 ± 0.40); for the Danicamtiv group, the heat–stress relation is Q _active = (0.24 ± 0.04). S _active + (1.03 ± 0.45). C, the y‐intercept of the muscle isometric heat–stress relation in panel B quantifies the activation heat (heat associated with intracellular Ca²⁺ cycling), and was unaffected by danicamtiv (P = 0.5495). D, peak cross‐bridge heat was calculated as peak isometric heat (obtained at L _o) minus activation heat (i.e. from the relation in panel B by subtracting the y‐intercept from the peak heat value). Danicamtiv increased it (P = 0.0002). E, the reciprocal of the slope of the muscle isometric heat–stress relation in panel B quantifies cross‐bridge economy (force produced per unit of cross‐bridge heat), and was significantly reduced by danicamtiv (P < 0.0001). F, peak shortening heat was calculated as the difference between the y‐intercepts of the heat–stress relations in panels A and B (i.e. heat at peak shortening minus activation heat) – danicamtiv increased peak shortening heat (P = 0.0197). In panels A and B, statistical comparisons of the regression lines between the two Groups (Control and Danicamtiv) were made with the ‘MODEL’ statement consisting of the Regression Equation, the ‘RANDOM’ statement consisting of Group, and the ‘REPEATED’ statement consists of Muscle. In panels C–F, t test (two‐tailed, paired) was used (the condition of normality met, as tested using Shapiro–Wilk test); bars represent means ± SD; n = 20 trabeculae.

Danicamtiv increased mechanical efficiency

As danicamtiv increased both stress (Fig. 3) and the extent of shortening (Fig. 5B ), the work performed thus significantly increased, as a function of relative afterload (Fig. 7A ; P < 0.0001). The peak value of work was greater under danicamtiv (P = 0.0065). Change of enthalpy (the sum of work and heat) as a function of relative afterload increased (P < 0.0001) under danicamtiv (Fig. 7B ). The peak value of work was greater under danicamtiv (P = 0.0002). Mechanical efficiency (the ratio of work to change of enthalpy) significantly increased under danicamtiv as a function of relative afterload (Fig. 7C ; P = 0.0084). Peak mechanical efficiency was significantly elevated from 11.9 ± 4.3% to 14.0 ± 4.7% (P = 0.0196). The work and change of enthalpy at the relative afterload at the point where mechanical efficiency peaked were both greater under danicamtiv (P = 0.0051 and P = 0.0039, respectively).

Figure 7. Muscle work, change of enthalpy and mechanical efficiency.

Average regression lines (n = 20 trabeculae × 7 afterloads per trabecula; continuous line for danicamtiv and dashed line for control) are overlaid with 95% confidence intervals (grey, control; blue, danicamtiv). Statistical comparisons of the regression lines between the two Groups (Control and Danicamtiv) were made with the ‘MODEL’ statement consisting of the Regression Equation, the ‘RANDOM’ statement consisting of Group, and the ‘REPEATED’ statement consists of Muscle. Insets show individual data points from a representative trabecula and fitted with regression lines. A, danicamtiv significantly elevated the cubic relation describing work (stress–length work‐loop area; Fig. 2B ) and relative afterload (P < 0.0001). B, danicamtiv significantly elevated the quadratic relation describing change of enthalpy (work plus heat) and relative afterload (P < 0.0001). C, danicamtiv significantly elevated the cubic relation describing mechanical efficiency (the ratio of work to change in enthalpy) and relative afterload (P = 0.0084). Regression lines for work (panel A) and mechanical efficiency (panel C) were both constrained to the origin. All regression lines are plotted to commence in the vicinity of the passive stress relative to the isometric stress (Fig. 3B ), that is at around 0.2 relative afterload.

Danicamtiv preserved cross‐bridge efficiency

Danicamtiv elevated cross‐bridge heat (Fig. 8A ; P < 0.0001) and preserved cross‐bridge efficiency (the ratio of work to the sum of cross‐bridge heat and work; Fig. 8B ; P = 0.2799), as functions of relative afterload. The peak value of cross‐bridge efficiency was preserved in trabeculae despite being presented with danicamtiv (15.2 ± 4.8% and 16.7 ± 5.4%; P = 0.1687).

Figure 8. Cross‐bridge energetics.

Average regression lines (n = 20 trabeculae; continuous line for danicamtiv and dashed line for control) are overlaid with 95% confidence intervals (grey, control; blue, danicamtiv). Statistical comparisons of the regression lines between the two Groups (Control and Danicamtiv) were made with the ‘MODEL’ statement consisting of the Regression Equation, the ‘RANDOM’ statement consisting of Group, and the ‘REPEATED’ statement consists of Muscle. Insets show individual data points from a representative trabecula and fitted with regression lines. A, danicamtiv elevated cross‐bridge heat production as a linear function of relative afterload (P < 0.0001). Peak values of cross‐bridge heat are plotted in Fig. 6D . B, danicamtiv preserved cross‐bridge efficiency (the ratio of work to work plus cross‐bridge heat) as a cubic function of relative afterload (P = 0.2799). Regression lines for cross‐bridge efficiency were constrained to the origin. All regression lines are plotted to commence in the vicinity of the passive stress relative to the isometric stress (Fig. 3B ), that is at around 0.2 relative afterload.

Danicamtiv increased peak mechanical efficiency

Danicamtiv elevated work output (Fig. 9B ), change of enthalpy (Fig. 9C ) and peak mechanical efficiency (Fig. 9D ) as functions of muscle length and preload. Work and change of enthalpy were estimated at the optimal afterload that yielded peak mechanical efficiency (refer to Fig. 7, where the optimal afterload at L _o is 0.44). All these three indices were greater in magnitude in danicamtiv compared with control over quite broad ranges of muscle lengths and preloads.

Figure 9. Simulation based on experimental data.

A, data from the danicamtiv group are displayed to illustrate simulations of work‐loops using the end‐systolic framework (Han, Pham et al., 2019; Tran et al., 2020). On the stress–length plane (left panel), the end‐systolic zone (the area coloured in grey) is bounded by three relations based on experimental data: (i) the isometric end‐systolic stress–length relation (data in Fig. 3B ; top black line), (ii) the work‐loop end‐systolic stress–length relation at L _o (data in Fig. 3A ; blue line), and (iii) the isometric passive stress–length relation (data in Fig. 3B ), the preload point that yields the maximum extent of shortening at the lowest afterload of work‐loop (lower boundary). On the heat–stress plane (right panel), the heat–stress zone (the area coloured in grey) is bounded by two relations: (i) the isometric heat–stress relation (data in Fig. 6B ; black line) and (ii) the work‐loop heat–stress relation at L _o (data in Fig. 6A ; blue line). These two zones contain the end‐systolic points and the heat‐equivalent of the end‐systolic points. Ten work‐loops, ranging from initial length of 0.85 to 1, are simulated. Filled circles mark the work‐loop end‐systolic points. Each end‐systolic point is defined by an initial length (preload) and an afterload. For each initial length (preload), the afterload was chosen such that the mechanical efficiency was maximal. For example, at an initial length of L _o, the peak mechanical efficiency is obtained at a relative afterload of 0.44 (see Fig. 7). B, at these optimal relative afterloads over ranges of initial length and preload, simulated work output is greater in the danicamtiv group (filled circles) than in the control group (open circles). C, the same is obtained for simulated change of enthalpy, which is greater in the danicamtiv group than in the control group. D, peak mechanical efficiency is also greater in the danicamtiv group than in the control group. Note that at L _o (which yields the highest preload), the values for work, change of enthalpy and peak mechanical efficiency are consistent with those presented in Fig. 7 at the optimal relative afterload of 0.44.

We describe our simulation here in the Results section to provide direct reference to the relevant data and figures (Figs 3 and 6) supporting its development and validation, and to facilitate clarity and interpretation. The simulation was performed by simulating work‐loops, at the optimal afterload, at given initial muscle lengths (shown in Fig. 9A ). Details of our simulation are available elsewhere (Han et al., 2023, 2022). Briefly, the simulation was based on our end‐systolic zone framework (Han, Pham et al., 2019; Tran et al., 2020), which requires experimental data for establishing on the stress–length plane (i) the isometric end‐systolic stress–length relation (from Fig. 3B ), (ii) the work‐loop end‐systolic stress–length relation at L _o (from Fig. 3A ), and (iii) the isometric passive stress–length relation (from Fig. 3B ), that yields the maximum extent of shortening under work‐loop contraction at the lowest afterload at each initial length. On the heat–stress plane, it requires (iv) the isometric heat–stress relation (from Fig. 6B ) and (v) the work‐loop heat–stress relation at L _o (from Fig. 6A ). These relations serve as boundaries for the end‐systolic zone and the heat–stress zone, shaded in Fig. 9A left and right panels, respectively. The end‐systolic zone contains end‐systolic points of work‐loops unique to any combinations of initial length ≤ L _o (or preload) and afterload. The heat–stress zone contains the heat‐equivalent of end‐systolic points.

Discussion

This is the first study, to our knowledge, that has assessed the energetic effects of danicamtiv on contracting trabeculae, specifically determining, as functions of muscle length or afterload, the force–length work output, active heat output and efficiency of contraction. The two latter parameters were evaluated at the muscle level and at the cross‐bridge level. We found that danicamtiv, as a myotrope, increased muscle contractile force and slowed contractile kinetics – results that are consistent with the literature. Our novel finding is that the increased force was accompanied by an increase in the heat output at the cross‐bridge level. At the muscle level, the increase in the force–length work output was greater than the increase in the change of enthalpy, such that mechanical efficiency (the ratio of work output to muscle change of enthalpy) was increased, in the presence of danicamtiv. However, because both extent of increase in work output and extent of increase in cross‐bridge change of enthalpy were comparable, we found that danicamtiv preserved cross‐bridge efficiency (the ratio of work output to crossbridge change of enthalpy).

In testing the first assumption, we found that danicamtiv, like other myotropes such as omecamtiv mecarbil (Bernier & Buckley, 2021; Lehman et al., 2022), does not increase the energy demand for intracellular Ca²⁺ cycling. We measured this energetic cost by quantifying the heat‐intercept of the isometric heat–stress relation (Fig. 6B ), which provides a measure of cardiac activation heat (Pham et al., 2017) – the energy expenditure associated with Ca²⁺ pumping by the sarcoplasmic reticular Ca²⁺‐ATPase and Ca²⁺ translocation by the Na⁺/K⁺‐ATPase in concert with the Na⁺/Ca²⁺ exchanger. Danicamtiv preserves activation heat (Fig. 6C ), consistent with the intended action of a myotrope, which, unlike a calcitrope, does not affect intracellular Ca²⁺ handling. Measurements by Ráduly et al. (2022) in canine cardiomyocytes of intracellular Ca²⁺ transients show non‐significant effects of danicamtiv on the transient amplitude and duration. However, their data do suggest a steady decrease in the Ca²⁺ transient amplitude with increasing danicamtiv doses from 0.01 to 2 µM, which may have been significant at the 5 µM dose used in our study. Their experiments were conducted at room temperature with fura‐2 dye‐loaded cardiomyocytes contracting at 0.5 Hz. Under similar experimental conditions, Horváth et al. (2017) demonstrated in 1‐Hz contracting canine myocytes the lack of effect of omecamtiv mecarbil on parameters of the Ca²⁺ transient at 1 and 10 µM doses. We do not have data to show the direct linkage between the magnitude of intracellular Ca²⁺ transient and activation heat. Our measurement of activation heat quantifies the energy expenditure associated with intracellular Ca²⁺ cycling and, thus, can be used to quantify the number of Ca²⁺ ions cycled per twitch, which is estimated to be 4.1 × 10¹² and 4.5 × 10¹² in trabeculae without and with danicamtiv, respectively. Our measurements show that danicamtiv does not affect the energy expenditure associated with intracellular Ca²⁺ cycling and the number of Ca²⁺ ions cycled per twitch. We note that Bakkehaug et al. (2015), studied the ex vivo mouse heart, reported no effects of omecamtiv mecarbil on the energy associated with excitation–contraction coupling (which directly quantifies the energy expenditure associated with intracellular Ca²⁺ cycling; Han et al., 2023), despite using a dose higher than those used clinically (Teerlink et al., 2015). Likewise, Obata et al. (2019) showed no effects of omecamtiv mecarbil on the energy associated with excitation–contraction coupling in control isolated rat hearts, although it could be argued that that was due to a lower dose (200–500 ng/mL, ∼0.5–1.25 µM) – no effects were observed in the left‐ventricular end‐systolic and end‐diastolic pressure–volume relations either.

The second assumption that danicamtiv increases force economically is not upheld. Our results show that cross‐bridge economy (force per heat under isometric contractions) is decreased by danicamtiv (Fig. 6E ) – the slope of the heat–stress relation is elevated by danicamtiv (Fig. 6B ). Similar to skinned muscle preparations where their economy, directly the crossbridge, is quantified by the ratio of force to ATP consumption, crossbridge economy in intact muscles is assessed by the inverse slope of the isometric heat–stress relation (Kiriazis & Gibbs, 2000; Loiselle, 1979) – that is, the ratio of isometric force to crossbridge heat (Johnston et al., 2016; Pham et al., 2019). Our finding shows that danicamtiv decreases the force development per unit energetic cost of cross‐bridges – that is, for any given unit of force, heat output from cross‐bridges is greater in danicamtiv, obtained under isometric contractions (Fig. 6B ). Under work‐loop contraction, heat is also greater per afterload in the presence of danicamtiv (Fig. 8A ). We postulate that the greater heat output from cross‐bridges for a given increase of force in danicamtiv compared with control arises from ‘greater recruitment of myosin heads’ from the low‐energy, reserved super‐relaxed (SRX) state pool to the disordered‐relaxed (DRX) state. The myosins in the DRX state are available for binding actin to generate force (transitioning into the actin‐bound active‐cycling state), whereas those in the SRX are unavailable for actin binding. Both SRX and DRX actin‐free myosins hydrolyse ATP and, thereby, generate heat. The ATP turnover time for the actin‐bound myosins, DRX myosins, and SRX myosins is reported to be 3 s, 0.03 s, and 0.003 s, respectively (Nag & Trivedi, 2021). Our steady‐state measurement of the rate of muscle heat output in the isometric length‐change protocol is about 3 min (Fig. 2A ). Thus, our heat measurement would have captured the thermal activity of the myosins in the DRX–SRX states. It is shown that stretching muscle can release myosin from the SRX state, moving it away from the thick filament backbone and into the disordered (DRX) state, where they are available to form crossbridges (Ma et al., 2021). This transition would be the same in both control and danicamtiv cases because in our experiments, muscles were stretched in both cases, that is the linear heat–stress relations described in Fig. 6B were obtained by changing muscle force via changing muscle length in isometric contractions. We emphasise ‘greater recruitment of myosin heads’ (see above) as we postulate that danicamtiv induces an increase of the myosin pool in the DRX state over the SRX state, because we measured a greater amount of heat per unit force in danicamtiv compared with control (Fig. 6B ). Thus, we believe that the mechanism behind danicamtiv‐induced reduction of crossbridge economy is the increase of myosin populations in the DRX state, released from the more energetically economical SRX state. Our conjecture is supported by the X‐ray diffraction data of Kooiker et al. (2023) that show danicamtiv‐induced increase of the equatorial intensity ratio, which reflects increased proximity of myosin heads to the thin filament, suggesting that myosin heads move away from the thick filament backbone and close to the thin filament, that is DRX state.

Danicamtiv increases muscle force, albeit not economically. We find that danicamtiv, on intact muscles, elevates the total (end‐systolic) stress–length relation, under both isometric contraction and work‐loop contractions (Fig. 3). Such an inotropic effect of danicamtiv has been shown in skinned muscles at sub‐maximal Ca²⁺‐activated force (Awinda et al., 2024; Choi et al., 2023; Kooiker et al., 2023; Voors et al., 2020) and in twitching engineered myocardia (Shen et al., 2021). We also find that danicamtiv elevates the passive (end‐diastolic) stress–length relation, replicating the results of increased passive force obtained in twitching engineered myocardia (Shen et al., 2021) and the echocardiographic results of decreased E/A ratio indicative of elevated left‐ventricular filling pressures (Ráduly et al., 2022). Thus, it has been claimed that danicamtiv enhances systolic function with minimal diastolic cost (Shen et al., 2021); however, the increase in the passive (end‐diastolic) force can potentially limit systolic function at higher rates of contraction.

The lusitropic effect of danicamtiv is manifested in the twitch kinetics, in terms of the rates of rise and fall of the twitch, which increases the force–time integral and prolongs twitch duration (Fig. 4). Under isometric contractions, twitch prolongation is consistent with the results obtained in mouse trabeculae and papillary muscles (Kooiker et al., 2023) and in engineered human myocardia (Shen et al., 2021). We show that slowed twitch kinetics are likewise reflected in work‐loop contractions, where danicamtiv‐treated trabeculae spend more time shortening, with reduced velocity of shortening, but shorten to greater extents (Fig. 5). The latter result reveals a greater shortening fraction, a muscle equivalence of whole‐heart ejection fraction. Our data from isolated trabeculae are consistent with in vivo echocardiographic measurements of increased ejection fraction and fractional shortening in healthy mice (Kooiker et al., 2023) and rats (Ráduly et al., 2022) administered with danicamtiv. We also find that the danicamtiv‐induced increase in muscle shortening is associated with an increase in the shortening heat (Fig. 6F ).

At the stimulation frequency of 3 Hz, we were limited in conducting our experiments using danicamtiv doses higher than 5 µM (Fig. 1) given the dose‐dependent increase of twitch duration, which decreases the diastolic period following twitch relaxation. For future experiments using lower stimulation frequencies, which deviate further below the physiological heart rate of the rat, higher doses of danicamtiv could be considered. The constraint imposed by both the diastolic cost and the lusitropic cost of danicamtiv renders experiments challenging at higher stimulation frequencies or heart rates. In a phase 2a clinical trial, danicamtiv was shown to improve cardiac function in patients with heart failure with reduced ejection fraction (Voors et al., 2020), supported by pre‐clinical findings in transgenic mice with desensitised myofilament to Ca²⁺ (Kooiker et al., 2023) and in cardiac tissue samples from patients with end‐stage heart failure (Choi et al., 2025). Danicamtiv, with its diastolic and lusitropic constraints, may not be useful in treating heart failure where disease‐induced twitch prolongation and slowed kinetics are significant including heart failure caused by diabetes (Han, Tran, Nielsen et al., 2014), systemic (Han, Tran, Johnston et al., 2014) and pulmonary (Pham et al., 2018) hypertension, and normotensive ventricular hypertrophy (Han, Tran et al., 2021). If the treatment is aimed solely at enhancing systolic function by increasing inotropy, we predict that cross‐bridge energy expenditure in diseased muscles would increase to sustain the increased force. Their mechanical efficiency would increase given the disproportional increase of work and of heat, as in healthy muscles. We further predict that the extent of increase of mechanical efficiency in diseased muscles would be greater than that in healthy muscles. This prediction arises from findings in the literature. From skinned muscle experiments, the extent of the danicamtiv‐induced increase of Ca²⁺‐sensitivity is greater in failing myocardium than in donor myocardium (Choi et al., 2025). Thus, for a given intracellular Ca²⁺, the extent of inotropy is greater in failing myocardium when treated with danicamtiv than that in donor myocardium. Likewise, from intact muscle experiments, the danicamtiv‐induced increase of twitch force is greater in the diseased muscles than in the control muscles (Kooiker et al., 2023).

The inotropic effect of danicamtiv in increasing twitch force is accompanied by an increase in heat output both at the muscle level (Fig. 3) and at the cross‐bridge level (Fig. 8A ). The latter increase may arise from increased myosin ATPase rate (Voors et al., 2020) as danicamtiv increases the number of myosin heads in the ‘On’ state (Kooiker et al., 2023). Note that, in Fig. 3, at a given muscle length, the end‐systolic stress is greater under the isometric contraction than under the work‐loop contraction, which involves muscle shortening. This exhibits the contraction‐mode dependence of the cardiac force–length relation (Han, Pham et al., 2019; Han, Tran et al., 2019; Han, Loiselle et al., 2021), and is reflected also in the heat–force relation, where at a given force, heat is greater under the work‐loop contraction than under the isometric contraction (Tran et al., 2020), reflecting shortening heat of cardiac muscle (Tran et al., 2017). Danicamtiv preserves cardiac contraction‐mode dependency, both on the force–length plane (mechanics) and on the heat–force plane (energetics).

In cardiac muscle contraction, the major part of the energy or change of enthalpy is liberated as heat (Loiselle et al., 2016). The energy converted to perform mechanical work is only 0.10–0.15 of the total energy – about 0.85–0.90 is wasted as heat, giving a mechanical efficiency that peaks at 10–15%, and is load‐dependent (Han, Taberner, Tran, Goo et al., 2012). The afterload relative to the isometric force at which mechanical efficiency peaks is around 0.4 (Fig. 7). We find that danicamtiv increases mechanical efficiency as a function of relative afterload – the peak value is increased by a factor of 1.17. Such an increase arises from the greater extent of increase in the work output (by 1.47) than the increase in the muscle heat output (by 1.27) and in the change of enthalpy (by 1.29). The increased work arises from increases in both the force and the shortening. We also measured cross‐bridge heat, by subtracting muscle heat from the activation heat. The increase by danicamtiv of the cross‐bridge heat (by 1.47) parallels the increase of work (by 1.47). Thus, at the cross‐bridge level, danicamtiv preserves cross‐bridge efficiency (Fig. 8).

We compare our findings of danicamtiv increasing cardiac mechanical efficiency with those obtained from use of omecamtiv mecarbil. Shen et al. (2010) studied dogs with systolic heart failure infused with 0.25 mg/kg omecamtiv mecarbil (∼9 µM) and showed increased stroke volume and cardiac output with change in myocardial oxygen consumption, suggesting improved cardiac efficiency. Utilising the pressure–volume area concept, Bakkehaug et al. (2015) reported decreased ‘contractile efficiency’ post‐treatment of omecamtiv mecarbil (500–1000 ng/mL, ∼1.2–2.5 µM) in both healthy pigs and pigs with post‐ischaemic left‐ventricular dysfunction. Obata et al. (2019) also utilised the pressure–volume area concept and studied isovolumically contracting rat hearts (both healthy and failing) infused with 200–500 ng/mL omecamtiv mecarbil (∼0.5–1.25 µM), but showed improved ‘contractile efficiency’. We have found the pressure–volume area concept to be thermodynamically flawed (Han, Taberner, Tran, Goo et al., 2012; Han, Taberner, Tran, Nickerson et al., 2012) as the estimation of the ‘potential energy’ term is ambiguous and subjective (Han et al., 2024a, 2024b). We contend that the inverse slope of the relation between myocardial oxygen consumption and the so‐called ‘pressure–volume area’ does not measure cardiac efficiency. In the present study, our measurement of cardiac efficiency is consistent with the conventional measure of efficiency where the numerator consists of only the actual, measurable, mechanical work, and that cardiac efficiency is afterload dependent. We show that danicamtiv increases cardiac mechanical efficiency as a function of afterload.

Cardiac efficiency is likewise preload dependent. In Fig. 9, our simulation shows that danicamtiv increases cardiac mechanical efficiency as a function of preload (and muscle length). The simulation, described in the Results, is conducted using our experimental data (Fig. 3) covering wide ranges of muscle lengths and afterloads, and from using two modes of contraction (work‐loop and isometric). In our experiments, the initial muscle length for work‐loop contractions was fixed at L _o in both control and danicamtiv groups. Under a condition where the initial muscle length (or preload) is lowered in the danicamtiv group, then mechanical efficiency could be preserved, that is at L _o for control, but at 0.92 L _o for danicamtiv. Such a condition may replicate the results of Voors et al. (2020) in their multi‐dose trial reporting a reduction in end‐diastolic volume following danicamtiv treatment in patients only at high dose (≥3500 ng/mL), and those of Ráduly et al. (2022) reporting a reduction in diastolic sarcomere length in isolated canine cardiomyocytes, following danicamtiv treatment. Similarly, a reduction in end‐diastolic volume in vivo following treatment of omecamtiv mecarbil in pigs has been reported (Bakkehaug et al., 2015). However, whether a myotrope has an effect on end‐diastolic volume and on diastolic sarcomere and muscle lengths remains inconclusive. Fülöp et al. (2021) showed that omecamtiv mecarbil did not change the left‐ventricular end‐diastolic volume in the rat and was accompanied by an increased end‐diastolic pressure. Voors et al. (2020) in their single‐ascending dose trial showed no reduction in end‐diastolic volume post treatment of danicamtiv. They also observed negligible effects of danicamtiv on left‐ventricular end‐diastolic dimensions in patients and in dogs. Similarly, while Ráduly et al. (2022) reported a reduction in diastolic sarcomere length in isolated canine cardiomyocytes, they observed no change in the left‐ventricular (LV) end‐diastolic diameter (EDD) in rats under danicamtiv, leading them to propose that the reduction in diastolic sarcomere length (SL) is associated with increase in cardiomyocyte passive force. They further proposed that ‘the apparent discrepancy between the danicamtiv‐evoked decrease in the diastolic SL of unloaded cardiomyocytes and the unchanged EDDs is the consequence of danicamtiv‐aided atrial compensation during LV filling’. A new study by Holmes & Stelzer (2025) on in situ pressure–volume measurements in the mouse shows a reduction of end‐diastolic volume at the higher dose of danicamtiv (6 mg/kg/min infusion) but not at the lower dose (3 mg/kg/min). Our data‐driven simulation demonstrates that the increase in mechanical efficiency with danicamtiv persists over quite broad ranges of muscle lengths and preloads.

Perspectives

Integrating our findings with the current literature on cardiac myotrope, we present three perspectives that may be useful for future investigations. First, the dose‐dependent effect of danicamtiv could be compared between intact muscles and permeabilised muscles. We studied intact muscles and found that low doses (<5 µM) do not significantly increase muscle force. Low doses (1–3 µM) have been typically used in studying permeabilised muscles (Awinda et al., 2024; Choi et al., 2023; Choi et al., 2025; Kooiker et al., 2023; Voors et al., 2020). Mechanical response of muscle to danicamtiv may be different between these two types of muscle preparations as the process of chemical skinning can induce loss of endogenous Ca²⁺ sensitizers and inadvertent proteolysis, altering myosin light chain phosphorylation and crossbridge kinetics, leading to a right‐shift of the myofilament force–Ca²⁺ relation (Gao et al., 1994). Second, when studying intact preparations, experimental conditions including temperature and stimulus frequency can be factors affecting the dose‐dependent effect of danicamtiv. We note that the dose‐effects of a myotrope, that is omecamtiv mecarbil as reported by Horváth et al. (2017), can be stimulus‐frequency dependent. Cardiac muscle force is certainly temperature and stimulus‐frequency dependent; mechanical efficiency is likewise (Loiselle, 1979; Pham et al., 2025). Lower stimulus frequencies and/or temperatures may reduce the EC₅₀ of the force dose–response of danicamtiv. For example, Shen et al. (2021) studied 1‐Hz twitching human engineered tissues, presumably at 37°C (Schwan et al., 2016), and obtained an EC₅₀ of 1.0 µM; we obtained 4.14 µM. For these cultured tissues, muscle force and its relation with muscle length progressively increases with culture days (Schwan et al., 2016), and tissue response to danicamtiv could be time‐dependent. Kooiker et al. (2023) studied intact mouse muscles twitching at 1 Hz and at 33°C and obtained inotropic effects of danicamtiv at low doses (1–3 µM). Ráduly et al. (2022) stimulated intact myocytes at 0.5 Hz and at room temperature, at low danicamtiv doses. Third, when studying intact muscles, if the intention is to investigate time‐varying systemic loads on muscle work‐loops in danicamtiv, then the work‐loop contraction protocol can be expanded to Windkessel loading that is informed by vascular biophysical properties including arterial compliance and impedance, and peripheral resistance (Garrett et al., 2019). A set of Windkessel parameters (Garrett et al., 2022) would have to be measured in the in vivo setting during infusion of danicamtiv. Subsets of Windkessel parameters such as systemic vascular resistance (Bakkehaug et al., 2015) and total peripheral resistance (Shen et al., 2010) can be measured in vivo and have been shown to be reduced following administration of omecamtiv mecarbil.

In conclusion, we measured the energetic responses of cardiac muscles exposed to danicamtiv. We used both isometric contraction and work‐loop contraction protocols to tease apart the heat output arising from intracellular Ca²⁺ cycling from that arising from cross‐bridge cycling. We show that danicamtiv is a myotrope that exerts its inotropic effects on muscle force production, muscle shortening, work, muscle heat and cross‐bridge heat. Danicamtiv increases cardiac mechanical efficiency at the muscle level, given its null effects on the energy expended by intracellular Ca²⁺ cycling and on the cross‐bridge efficiency.

Competing interests

The authors declare they have no competing interests.

Author contributions

T.M. and J.‐C.H. conceived and designed the study, performed experiments, prepared figures and drafted the manuscript. T.P., K.T. and A.T. contributed to the conception of the research and interpretation of the data. All authors edited and revised the manuscript. All authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

This work was supported by funding received from the Heart Foundation of New Zealand through Project Grant (1929, awarded to J.‐C.H.) and Senior Research Fellowship Grant (2027, awarded to J.‐C.H.); the Health Research Council of New Zealand through Sir Charles Hercus Health Research Fellowship Grants (21/116; awarded to K.T.), and Emerging Researcher First Grant (21/653, awarded to T.P.); and the Royal Society of New Zealand through Marsden Project Grant (MFP‐UOA2206, awarded to J.‐C.H.).

Supporting information

Peer Review History

Biography

Trent Meehan completed the work for this manuscript during a 9‐month research assistant placement in the Cardiac Energetics lab at the Auckland Bioengineering Institute, where he was trained and supervised by Dr June‐Chiew Han. This was following the completion of a Master of Biomedical Sciences from the University of Auckland in 2024. He now works as a research associate at a biotechnology company focusing on cell line development.

Data availability statement

The data supporting this study's findings are available upon request from the corresponding author.