Cross-Bridge versus Thin Filament Contributions to the Level and Rate of Force Development in Cardiac Muscle

Abstract

In striated muscle thin filament activation is initiated by Ca²⁺ binding to troponin C and augmented by strong myosin binding to actin (cross-bridge formation). Several lines of evidence have led us to hypothesize that thin filament properties may limit the level and rate of force development in cardiac muscle at all levels of Ca²⁺ activation. As a test of this hypothesis we varied the cross-bridge contribution to thin filament activation by substituting 2 deoxy-ATP (dATP; a strong cross-bridge augmenter) for ATP as the contractile substrate and compared steady-state force and stiffness, and the rate of force redevelopment (k_tr) in demembranated rat cardiac trabeculae as [Ca²⁺] was varied. We also tested whether thin filament dynamics limits force development kinetics during maximal Ca²⁺ activation by comparing the rate of force development (k_Ca) after a step increase in [Ca²⁺] with photorelease of Ca²⁺ from NP-EGTA to maximal k_tr, where Ca²⁺ binding to thin filaments should be in (near) equilibrium during force redevelopment. dATP enhanced steady-state force and stiffness at all levels of Ca²⁺ activation. At similar submaximal levels of steady-state force there was no increase in k_tr with dATP, but k_tr was enhanced at higher Ca²⁺ concentrations, resulting in an extension (not elevation) of the k_tr-force relationship. Interestingly, we found that maximal k_tr was faster than k_Ca, and that dATP increased both by a similar amount. Our data suggest the dynamics of Ca²⁺-mediated thin filament activation limits the rate that force develops in rat cardiac muscle, even at saturating levels of Ca²⁺.

INTRODUCTION

Striated muscle contraction results from cyclic interactions between myosin and actin, which are regulated by the troponin-tropomyosin complex of the thin filament. Ca²⁺ binding to the thin filament regulatory protein troponin C (TnC) initiates contraction (Gordon et al., 2000) and the level of force developed is determined by the availability of myosin binding sites on thin filaments, which is ultimately controlled by the distribution of tropomyosin (Tm) between positions on the surface of thin filaments that either block or uncover binding sites on actin for myosin cross-bridge attachment. Transduction of the Ca²⁺ signal is through the troponin complex, resulting in increased mobility of Tm over the surface of actin and exposure of strong binding sites for myosin. This shifts the thin filament from a “blocked” to a “closed” state (McKillop and Geeves, 1993), thereby allowing further myosin cross-bridge formation. Strong cross-bridge binding is then thought to stabilize Tm in an “open” state and the thin filament is considered “activated.” In effect, the level of contractile activation appears to result from a balance of the effects of Ca²⁺ binding to TnC and cross-bridges on Tm state. The growing amount of evidence suggests that cross-bridge displacement of Tm into an activating position may play a more prominent role in the regulation of cardiac contraction (Fuchs and Smith, 2001).

Like force, the rate of force development is Ca²⁺ dependent, but the mechanisms that determine the thin filament activation dependence of force kinetics are less well established. In demembranated skeletal muscle fibers the relationship between steady-state force and the rate that force redevelops (k_tr) after a release-restretch cycle is Ca²⁺ sensitive, being slow (1–2 s⁻¹ at 10–15°C) at low levels of force and increasing 10–15-fold as force rises to maximum at high levels of Ca²⁺ activation (Brenner, 1986, 1988). Based on these and other findings it was suggested that Ca²⁺ controls the rate-limiting process in force development rather than simply controlling the level of cross-bridge recruitment (Brenner, 1986, 1988). However, we and others have shown that in skeletal muscle when the Ca²⁺-activated force is high the kinetics of actin-myosin interaction are likely limiting to the rate of force development, whereas at submaximal force the rate-limiting process may be the kinetics of thin filament activation (Chase et al., 1994; Regnier et al., 1996, 1998; Fitzsimmons et al., 2001a,b). This limiting process is greatest at low levels of activation, as most of the increase in k_tr comes at Ca²⁺ concentrations that produce >50% of maximal Ca²⁺-activated steady-state force.

Similar to reports for fast skeletal muscle, k_tr is Ca²⁺ dependent in demembranated cardiac muscle. However, most studies have reported smaller magnitude (one- to fivefold) change in k_tr (compared with skeletal muscle) as the level of Ca²⁺ activation is varied (Wolff et al., 1995; Palmer and Kentish, 1998; Patel et al., 2001), though one recent report suggest a >8-fold change (Fitzsimons et al., 2001). These studies mostly show that at similar temperatures k_tr is similar or faster in cardiac versus skeletal muscle at low Ca²⁺ concentrations, and slower at high levels of Ca²⁺ activation. This suggests the lower Ca²⁺ dependence of k_tr results from different actin-myosin cycling kinetics (the main determinant of maximal k_tr in skeletal muscle) and/or the relative kinetics of thin filament processes versus actin-myosin cycling kinetics is different than in skeletal muscle (Hancock et al., 1997). Coupled with evidence that thin filament activation level is more dependent on strong cross-bridge binding in cardiac (versus skeletal) muscle, it also suggests that Ca²⁺ binding to Tn may result in less complete activation of cardiac thin filaments (Regnier et al., 2000b).

We have been studying the relative contribution of Ca²⁺ binding to TnC and cross-bridge binding on both the level of thin filament activation and the rate that force develops in cardiac and skeletal muscle. A particularly successful approach has been to alter cross-bridge binding and cycling rates by using nucleotide analogs of ATP (NTPs) as the substrate for contraction. We have shown that one of these analogs, 2 deoxy-ATP (dATP), has an affinity for skeletal myosin similar to that of ATP and increases both solution NTPase and unloaded shortening velocity of rabbit skeletal and rat cardiac muscle (Regnier and Homsher, 1998; Regnier et al., 1998, 2000b). Interestingly, dATP enhances both maximal Ca²⁺-activated steady-state force (F_max) and k_tr in cardiac trabeculae from both untreated rats, containing mostly α-MHC (fast cardiac myosin), and PTU-treated rats, containing β-MHC (slow cardiac myosin) (Regnier et al., 2000b). This contrasts with our findings in skeletal muscle where dATP has no effect on F_max, but increases maximal k_tr in rabbit psoas fibers (fast skeletal myosin) and rat soleus fibers (slow skeletal myosin) (Regnier et al., 1998). These results indicate that cross-bridge cycling rate per se cannot explain these cardiac versus skeletal muscle differences, and suggests that the properties of cardiac thin filament regulatory proteins may limit cross-bridge cycling. Therefore we have hypothesized that (compared to ATP) dATP augments the level of thin filament activation and/or strong cross-bridge binding at saturating Ca²⁺ concentration in demembranated cardiac muscle, resulting in an increase in F_max and maximal k_tr over that attained with ATP as the contractile substrate. This hypothesis is supported by our finding that dATP also enhances both force and k_tr at subsaturating levels of Ca²⁺ activation in skeletal muscle (Regnier et al., 1998).

In this study we hypothesized that the relatively low activation dependence of k_tr in cardiac muscle (compared with skeletal muscle) reflects a relative limitation of Ca²⁺ binding to TnC to activate thin filaments at all levels of Ca²⁺. Consequently, this would give cross-bridge binding a more prominent role in the activation process. To test this hypothesis we compared steady-state force and stiffness, and k_tr in rat cardiac trabeculae as Ca²⁺ concentration was varied with either 5 mM ATP or 5 mM dATP as the contractile substrate. In addition, we tested the hypothesis that thin filament kinetics can limit the rate of force development during maximal Ca²⁺ activation. To do this we compared the rate force develops after a step increase in Ca²⁺ concentration with photorelease of Ca²⁺ from NP-EGTA (k_Ca) (Ellis-Davies and Kaplan, 1994) to maximal k_tr, where Ca²⁺ binding to thin filaments should be in (near) equilibrium during force development. We found that dATP enhanced steady-state force and stiffness at all levels of Ca²⁺ activation. We also found that at similar submaximal levels of steady-state force there was no increase in k_tr with dATP, but dATP enhanced k_tr at higher Ca²⁺ concentrations, resulting in an extension (not elevation) of the curvilinear force-k_tr relationship. Similar results were obtained for rat trabeculae with either α-MHC or β-MHC isoforms, indicating this effect was not due to the absolute rate of cross-bridge cycling. Significantly, we found that k_tr is faster than k_Ca with maximal Ca²⁺ activation, and that both are increased a similar amount by dATP. This finding suggests the kinetics and/or level of Ca²⁺-mediated thin filament activation may limit the rate that force develops in rat cardiac muscle, even at saturating levels of Ca²⁺. In a subsequent article, (Adhikari et al., 2004), we show that this property of cardiac thin filaments may at least partially explain the greater sarcomere length-dependence of force found in cardiac (versus skeletal) muscle. Preliminary reports of this work have been published previously (Regnier et al., 2000a; Martin et al., 2002).

METHODS

Male Sprague-Dawley rats (200–250 grams), either treated with propylthiouracil (PTU; 0.5% added to drinking water for 4–8 weeks) or untreated (as age-matched animals) were euthanized with sodium pentabarbitol (50 mg/kg). Hearts were excised and the interior wall of the right ventricle exposed to relaxing solution (see below) containing glycerol (50% v/v) and Triton X-100 (1%) overnight at 5°C. Individual trabeculae were dissected and stored at 5°C for up to 4 days.

Solutions

Experimental solutions contained (in mM): 15 phosphocreatine, 15 EGTA, at least 40 3-(N-morpholino)propanesulfonic acid (MOPS), 1 free Mg²⁺, 135 Na⁺ + K⁺, 1 dithiothreitol (DTT), 250 units ml⁻¹ creatine kinase (CK, Sigma, St. Louis, MO), and either 5 ATP or 5 mM 2 deoxy-ATP (Sigma, St. Louis, MO) at pH 7.0 and 15 ± 1°C. Ionic strength was 0.2 M. Affinities of dATP and ATP for Mg²⁺ were assumed to be the same (Regnier and Homsher, 1998). For activation solutions, the Ca²⁺ level (expressed as pCa = −log [Ca²⁺]) was set by adjusting Ca(propionate)₂.

Individual chemically demembranated (“skinned”) trabeculae were wrapped in aluminum foil T-clips for attachment to a force transducer (either Model 400A, 2.2 kHz-resonant frequency (Cambridge Technology, Watertown, MA) or Model AE801, ≥5 kHz resonant frequency (SensoNor, Horton, Norway)) and a servo-motor (model 300, Cambridge Technology, Watertown, MA) tuned for a 300 μs-step response. For most experiments trabecular sarcomere length (L_s) was measured with helium-neon laser diffraction (Chase et al., 1993) and set to 2.25 μm at pCa 9.2. At this L_s trabeculae lengths ranged from 1.2–2.3 mm and diameters averaged 140 ± 16 μm, with no significant difference between trabeculae from normal versus PTU-treated rats. Stiffness was determined by small amplitude (0.05% L_s) sinusoidal length oscillations (1 kHz). Steady-state isometric force is reported as kN/M² in Results, assuming circular geometry and the rate of isometric tension redevelopment (k_tr) was determined from the halftime of force recovery after a rapid release-restretch transient, as previously described (Regnier et al., 1998). For comparisons, force and k_tr are reported normalized to measurements obtained during maximal Ca²⁺ activation with ATP as the contractile substrate.

Comparisons of maximal k_tr with the rate of force development (k_Ca) after photorelease of Ca²⁺ from NP-EGTA were made during the same activation done in a separate set of experiments under identical conditions, except temperature was 17°C instead of 15°C. k_Ca was measured as described in Martin et al. (2004). Briefly, after initial Ca²⁺ activations in either ATP or dATP, during which control measurements of k_tr were performed at maximal activation (k_{tr control}), the preparation was transferred through a series of NTP containing solutions to a final photolysis solution containing 5 mM NTP, 2mM NP-EGTA, and ∼1.8 mM added calcium. The output from a frequency doubled ruby laser producing ∼150 mJ in a 50 ns pulse at a wavelength of 347 nm (Laser Applications, Winter Park, FL) was focused onto the trabeculae. Incident energy was measured by deflecting 10% of the beam energy onto a Scientech Energy Meter (365, Boulder, CO); ∼2 s after peak force was attained, k_tr was measured (k_{tr photolysis}) and the tissue relaxed. For these experiments, all elicited transients were collected for subsequent analysis but only those resulting in maximal activations within 5% of maximal Ca²⁺-activated force at pCa 4.0 (F_max) have been included in the data.

Data analysis

Force-pCa and k_tr-pCa relationships are fit with the Hill equation:

(1)

From each Hill fit, the pCa at half-maximal force or k_tr (pCa₅₀) and the slope (n_H) are determined, and reported pCa₅₀ and n_H values represent the means of the values from the individual fits, ± SE. The time course of force development from k_tr and k_Ca protocols was determined by exponential fits or from t_1/2 values, as previously described in (Chase et al., 1994), both yielding similar results. Summarized data are reported as mean ± SE and means were compared using Student's t-test, with significance at the 95% confidence level.

RESULTS

Steady-state force

In chemically demembranated (skinned) trabeculae, substitution of 5 mM dATP for 5 mM ATP increased steady-state force at all levels of Ca²⁺ activation. The steady-state force data for all the experiments is summarized in Fig. 1. Trabeculae from untreated and PTU-treated rats were used to determine if force enhancement by dATP depends on the myosin isoforms present. Force values were normalized to F_max with ATP for each trabecula and the Ca²⁺-dependence of force was characterized from fitting the Hill equation to the data (Eq. 1 in Methods) to obtain 1), the [Ca²⁺] required to elicit half of maximal Ca²⁺-activated force (i.e., pCa₅₀ or Ca²⁺ sensitivity); and 2), the slope (n_H or Hill coefficient) of the steep portion of the force-pCa relationship. For trabeculae from untreated rats F_max was 20.0 ± 4.2 kN/m² with ATP (n = 8) and was increased by 31 ± 3% with dATP (Fig. 1 a). In trabeculae from PTU-treated rats F_max was 17.6 ± 4.2 kN/m² with ATP (n = 10) and was increased by 39 ± 4% with dATP (Fig. 1 b). These increases in F_max with dATP are similar to what we have previously reported (Regnier et al., 2000b). dATP increased the Ca²⁺-sensitivity of force (pCa₅₀) of trabeculae from both untreated and PTU-treated rats, but had no significant effect on n_H. For trabeculae from untreated rats, dATP increased the Ca²⁺-sensitivity of force (P < 0.05) by 0.13 pCa units from 5.41 ± 0.01 (ATP) to 5.54 ± 0.03 (dATP), whereas corresponding n_H values were 5.4 ± 0.6 (ATP) and 4.6 ± 1.2 (dATP). For trabeculae from PTU-treated rats, pCa₅₀ with ATP was similar to untreated rats, whereas the increase of Ca²⁺ sensitivity of force with dATP was greater than found for untreated rats, increasing 0.26 pCa units from 5.43 ± 0.01 (ATP) to 5.69 ± 0.03 (dATP). n_H was unaffected by dATP in PTU treated trabeculae (3.2 ± 0.2 (ATP) vs. 3.5 ± 0.6 (dATP)). The Hill fit parameters (pCa₅₀ and n_H) for the force-pCa relations for trabeculae from untreated and PTU-treated rats included in Table 1.

FIGURE 1.

dATP (○) enhances the Ca²⁺-sensitivity of steady-state force in skinned trabeculae. Force was normalized to maximal force at pCa 4.5 with ATP (•) for trabeculae from normal (a) and PTU-treated (b) rats for comparisons. Data are reported as the mean ± SE for each pCa. Data were fit to Eq. 1 (Methods) and values for pCa₅₀ and n_H are reported in the text of Results. Fits for the data were r² > 0.997.

TABLE 1.

Comparison of Hill fit parameters for force-pCa and k_tr-pCa relations in untreated (control) and PTU-treated rats

	NTP	Fmax (normalized)	pCa50 (force)	nH (force)	Max. ktr (normalized)	pCa50 (ktr)
Untreated	ATP (8)	1.0	5.41 ± 0.01	5.4 ± 0.6	1	5.28 ± 0.01
	dATP (8)	1.31 ± 0.03	5.54 ± 0.03	4.6 ± 1.2	1.27 ± 0.05	5.38 ± 0.01
PTU	ATP (10)	1.0	5.43 ± 0.01	3.2 ± 0.2	1	5.42 ± 0.01
	dATP (10)	1.39 ± 0.04	5.69 ± 0.03	3.5 ± 0.6	1.47 ± 0.04	5.47 ± 0.01

The number of trabeculae used for each NTP condition is given in parentheses. F_max and maximal k_tr values are normalized to force obtained with 5 mM ATP at pCa 4.5.

In a subgroup of trabeculae from PTU-treated rats, steady-state stiffness was measured with small amplitude oscillations (see Methods). Linear regression analysis of the stiffness versus force relationship (as Ca²⁺ was varied) for activations in ATP versus dATP demonstrated a proportional increase in stiffness versus force for both nucleotides (Fig. 2). This suggests the increased force with dATP at each pCa results from enhanced strong cross-bridge binding.

FIGURE 2.

Relationship between steady-state force and trabeculae stiffness with varying [Ca²⁺] for 3 trabeculae from PTU-treated rats. Force and stiffness values are normalized to the level measured at pCa 4.0 with ATP as the contractile substrate. Linear regression fits to the ATP (▪) and dATP (□) data yielded identical slopes of 0.96 with r² values >0.97. The values for dATP >1.0 demonstrate that trabeculae stiffness is elevated proportionally with force over values for ATP, suggesting an increase in the number of strongly bound, force-producing cross-bridges at high levels of Ca²⁺ activation.

Rate of force redevelopment (k_tr)

Similar to the results with steady-state force, dATP enhanced k_tr at varying levels of Ca²⁺ activation. The force traces in Fig. 3 show this enhancement for example trabeculae from an untreated (Fig. 3 a) and a PTU-treated (Fig. 3 b) rat during activations at pCa 5.5 and pCa 4.0. In both of these examples trabeculae dATP elevated steady-state force at pCa 5.5 and pCa 4.0 by similar amounts versus ATP, and at pCa 5.5, k_tr was approximately doubled whereas maximal k_tr (pCa 4.0) was increased by 27 ± 5% (untreated) and 47 ± 4% (PTU-treated). Fig. 4 summarizes k_tr measurements at varying levels of activating Ca²⁺, normalized to pCa 4.0 and ATP. At 15°C maximal k_tr with ATP was 10.6 ± 0.8 s⁻¹ (n = 8) for trabecula from untreated rats and 3.0 ± 0.2 s⁻¹ (n = 10) for trabeculae from PTU-treated rats. The data show that dATP increased k_tr at higher levels of Ca²⁺ activation, but had little or no effect on k_tr at the lowest levels of activation. Although the starting level of force after restretch of muscle fibers varied somewhat between preparations and between measurements with ATP versus dATP within the same preparation, there was no indication these variable starting-force levels affected the magnitude differences in k_tr between ATP and dATP. Hill fits to the data indicate an increase in the Ca²⁺ sensitivity (pCa₅₀) of k_tr with dATP. In trabeculae from untreated rats (Fig. 4 a) pCa₅₀ was increased by 0.10 units (P < 0.05) from 5.28 ± 0.01 (ATP) to 5.38 ± 0.01 (dATP). This increase in the Ca²⁺ sensitivity of k_tr was similar to the increase seen for steady-state force (Fig. 1 a) and stiffness (Fig. 2). For trabeculae from PTU-treated rats (Fig. 4 b), the pCa₅₀ of k_tr was increased by 0.05 units (P < 0.05) from 5.42 ± 0.01 (ATP) to 5.47 ± 0.01 (dATP), an increase not as profound as that seen for steady-state force (Fig. 1 b). These data are included in Table 1 for comparison to the corresponding steady-state force values.

FIGURE 3.

Example force records comparing k_tr with ATP versus dATP at pCa 5.5 and 4.0 for trabeculae from an untreated (a) and a PTU-treated (b) rat. The pCa and rate of force redevelopment (k_tr) for each trace are given next to the respective force records. These examples demonstrate elevation of both steady-state force and k_tr at submaximal and maximal levels of Ca²⁺ activation.

FIGURE 4.

The Ca²⁺ dependence of k_tr with ATP (•) versus dATP (○) for trabeculae from untreated (a) and PTU-treated (b) rats. Measurements with ATP and dATP are made in the same preparations and values are normalized to maximal k_tr at pCa 4.0 with ATP. Values are mean ± SE. Data were fit with Eq. 1 (Methods) to demonstrate an increase in the Ca²⁺ sensitivity (pCa₅₀) of k_tr, as well as maximal k_tr, with dATP in both (a) and (b).

The leftward shift observed for both the force versus pCa (Fig. 1) and k_tr versus pCa (Fig. 4) relationships suggest that increases in k_tr with dATP may have resulted from an increased number of cross-bridges available to participate in the force-generating process. To test this idea, the k_tr data in Fig. 4 were replotted as k_tr versus force (both normalized to pCa 4.0 with ATP) at each level of Ca²⁺ in activation solutions (Fig. 5). This replotting of the data clearly demonstrates that k_tr was similar for dATP during Ca²⁺-activated contractions producing similar levels of steady-state force and was elevated by dATP only at activation levels resulting in greater steady-state force than that obtained with ATP during maximal Ca²⁺ activation. When coupled with the stiffness data, which rises proportionally with force (Fig. 2), it also suggests that k_tr with dATP is increased because of a greater recruitment of force generating cross-bridges at each level of activating Ca²⁺, including during maximal Ca²⁺ activation.

FIGURE 5.

The relationship between steady-state force and k_tr as pCa is varied with ATP (▪) versus dATP (□) for trabeculae from untreated (a) and PTU-treated (b) rats. Values are mean ± SE. In both a and b, k_tr is not different for ATP versus dATP at similar levels of steady-state force, but dATP extends the curve beyond maximal values of force and k_tr for ATP (demonstrated by dotted lines) at high levels of Ca²⁺ activation.

Rate of force development (k_Ca)

To investigate how Ca²⁺ activation processes might limit the rate of force development in cardiac muscle we compared the rate of force development under conditions where the [Ca²⁺] is kept constant (k_tr) versus when activation is initiated by a step increase in [Ca²⁺] by laser photolysis of NP-EGTA (k_Ca). Fig. 6 (upper panel) shows an example of the protocol used for these comparisons. This trabecula was placed in a control activation solution (pCa 4.5, 5 mM ATP) to determine F_max and maximal k_tr (first arrow), then was placed in relaxing solution (pCa 8.0). Next the trabecula was transferred into the NP-EGTA activation solution to allow the caged-Ca²⁺ to diffuse into the contractile lattice (2–3 min.). At this point the trabecula was activated by photolysis of the NP-EGTA (second arrow) and steady-state force developed, ∼2 s later a k_tr measurement was made (third arrow) and the trabecula was relaxed. In several preparations the k_tr measurement was repeated in control activation solution (not shown) to confirm that photolysis produced maximal Ca²⁺-activation conditions. The k_Ca → k_tr procedure was then repeated for activations with dATP as the contractile substrate (right side of Fig. 6, upper panel). The order of NTP testing was varied between preparations. Comparisons of k_tr and k_Ca were made only for maximal Ca²⁺ activation and then only if the photolysis of NP-EGTA produced Ca²⁺ activation within 5% of the force level in control activations. For this example trabecula it is obvious this is the case and that F_max was increased by dATP in both control and NP-EGTA activations.

FIGURE 6.

Procedure for sequential measurements of k_Ca and k_tr. (Upper panel) An example triton skinned rat trabeculae was transferred through a series of solutions containing 5 mM ATP (left) or dATP (right) with the indicated pCa (17°C). Where NP-EGTA is indicated the trabecula was transferred to a solution containing 2 mM NP-EGTA with 1.8 mM added Ca²⁺, then activated by a pulse of laser light (arrow, hν). Measurements of k_tr were made at pCa 4.5 in control solution (k_{tr control}) and during maximal activation induced by photolysis of NP-EGTA (k_{tr photolysis}). k_Ca was determined from the time course of the transient after NP-EGTA photolysis. (Lower panel) Rapid time base recordings from upper panel. (a) k_{tr photolysis} for ATP determined immediately after activation by photolysis of NP-EGTA. (b) k_Ca induced by photolysis of NP-EGTA with ATP. (c) k_{tr control} for ATP. Rates are determined from the t_1/2 of the respective transients. (d) Force records from a and b superimposed to show relative rates.

Fig. 6, a–d, shows the k_tr and k_Ca measurements for ATP at greater resolution and demonstrates two consistent findings: 1), steady-state force and k_tr after activation by photolysis of NP-EGTA was similar to control activations; and 2), k_Ca was slower than k_tr. Fig. 7 shows higher time resolution examples of k_Ca and the subsequent k_tr force traces comparing activations with ATP versus dATP, demonstrating that both measurements are faster when dATP is the substrate for contraction. The data for all experiments is summarized in Table 2. The k_tr averaged 75% faster than k_Ca for ATP and 65% faster for dATP, suggesting that the relative difference between the two measurements is independent of cross-bridge cycling rates. The data also show that dATP increased k_tr by 50% and k_Ca by 60% over measurements with ATP, suggesting faster cross-bridge cycling (dATP) enhances both rates by a similar mechanism.

FIGURE 7.

Comparison of k_Ca and k_{tr photolysis} for ATP versus dATP from an untreated rat trabecula. The nucleotide and rate determined from t_1/2 are shown next to each trace. (a) k_Ca. (b) k_{tr photolysis}.

TABLE 2.

Comparison of k_Ca and k_tr in the same maximal Ca²⁺ activation with ATP versus dATP as the contractile substrate

	ATP	dATP	dATP/ATP
kCa	4.3 ± 0.7 (8)	6.9 ± 0.4 (10)	1.61
ktr	7.6 ± 0.6	11.4 ± 1.4	1.49

The number of trabeculae used for each NTP condition is given in parentheses. For most trabeculae, k_Ca and k_tr measurements were made for both ATP and dATP.

DISCUSSION

There are three major findings from this study: 1), altering cross-bridge cycling by substitution of dATP for ATP as the contractile substrate increases the Ca²⁺ sensitivity of steady-state force (Fig. 1) and k_tr (Fig. 4) in rat cardiac muscle at 15°C, independent of myosin isoform; 2), both maximal Ca²⁺-activated steady-state force and the rate of force development (k_tr, k_Ca) can be significantly increased by increased cross-bridge binding and/or cycling rate (Fig. 5); and 3), the maximal rate that force develops by a step increase in Ca²⁺ (k_Ca) is significantly slower than the maximal rate of force redevelopment (k_tr) during activations with saturating levels of Ca²⁺. The maximal steady-state force and k_tr data agree with previous work in cardiac muscle where we demonstrated that dATP elevates maximal Ca²⁺ activated force, k_tr, unloaded shortening velocity and filament sliding velocity in vitro motility assays (Regnier et al., 2000b). These studies provide information about how altering cross-bridge cycling influences cross-bridge binding and thin filament activation at submaximal levels of Ca²⁺ in cardiac muscle. The results also allow us to compare results in cardiac muscle with previous work in fast and slow skeletal muscle, where we have extensively studied the effects of cross-bridge binding and cycling rates versus thin filament dynamics on the Ca²⁺ activation dependence of steady-state force and force development kinetics (Regnier et al., 1996, 1998, 1999, 2000a,b). Those studies provided strong evidence that thin filament dynamics are limiting to the level and rate of force development in skeletal muscle during submaximal Ca²⁺ activation.

The existing structural sequences and conservation information for myosin heavy chains offer no explanation about how dATP might increase F_max in cardiac muscle, but not in skeletal muscle (Regnier et al., 2000b). Indeed the increase in F_max with dATP may be peculiar to mammalian cardiac muscle in general, as we have also found similar results in chemically demembranated trabeculae from mouse and dog heart (unpublished observations). In previous studies we suggested that dATP might increase the number of force-generating cross-bridges during submaximal Ca²⁺ activation in skeletal muscle (Regnier et al., 1998) and during maximal Ca²⁺ activation in cardiac muscle (Regnier et al., 2000b) by increasing one or more transition rates associated with cross-bridge attachment and force generation while increasing the rate of cross-bridge detachment to a lesser extent. In this article, we now propose this enhanced cross-bridge attachment leads to increased activation, which is apparent at submaximal Ca²⁺ in both muscle types and in maximal force in cardiac muscle.

Steady-state force

Proportional increases in steady-state force and stiffness were found at all levels of Ca²⁺ activation with dATP. This apparent increase of thin filament activation could result from recruitment of strongly bound force-generating cross-bridges, from an increase in force per cross-bridge, or both. Since dATP increases cross-bridge cycling rate, it suggests the cross-bridge detachment rate (g_app) is increased, which further implies the rate of strong cross-bridge attachment (f_app) must increase to a greater extent to provide a higher level of steady-state force and stiffness (Regnier et al., 2000b). This effect of dATP appears to be partially or completely independent of the overall cross-bridge cycling rate, as there were increases in both maximal force and Ca²⁺ sensitivity of force for trabeculae from untreated versus PTU-treated rats (Fig. 1; Table 1). This suggests the effects of dATP are largely independent of myosin kinetics (Figs. 3 and 4), implying that enhanced strong cross-bridge binding is the important contribution of dATP to increased activation. Whatever the detailed mechanism of force enhancement by dATP, the net result is an apparent elevation of thin filament activation at all levels of Ca²⁺ in cardiac muscle with either α- or β-MHC. Interestingly, dATP increased the Ca²⁺ sensitivity of force (pCa₅₀) for trabeculae with β-MHC twice as much as for trabeculae with α-MHC (0.26 vs. 0.13 pCa units, respectively; Fig. 1 and Table 1). This suggests enhanced cross-bridge binding and/or cycling rate may improve activation in cardiac muscle with slow myosin. Our steady-state force-pCa results (Fig. 1) contrast with those reported by Metzger et al. (1999) in their study of myocytes from untreated versus PTU-treated rats. Although we report no difference in the Ca²⁺ sensitivity of force (pCa₅₀) for trabeculae from untreated versus PTU-treated rats, they observed a decrease in pCa₅₀ by 0.17 units with no change in the slope (Metzger et al., 1999). It is unclear why our results differ, but the focus and novel discovery of our study is that increased cross-bridge binding (with dATP) increases pCa₅₀ of force for both untreated and PTU-treated trabeculae without changing the apparent cooperativity (n_H) of the force-pCa relationship (Fig. 1).

In cardiac muscle changes in cross-bridge binding have been shown to increase Ca²⁺ binding to troponin C (Hofmann and Fuchs, 1987; Wang and Fuchs, 1994; Martyn and Gordon, 2001). Our experiments do not allow us to determine if enhanced Ca²⁺ sensitivity of force with dATP is correlated with increased Ca²⁺ binding to troponin C or via Ca²⁺-independent increases in strong cross-bridge binding (or both). The increased force at saturating [Ca²⁺], however, is most likely explained by increased cross-bridge binding alone because of proportional increases in trabeculae stiffness. To determine if force enhancement by dATP directly results in a shifting of thin filament toward the “on” state will require a more direct measurement of the state of tropomyosin on thin filaments.

Rate of force development (k_tr, k_Ca)

In this study the Ca²⁺ sensitivity of k_tr was increased with dATP (Fig. 4; Table 1), similar to steady-state force. When k_tr was plotted versus the steady-state force produced at varying [Ca²⁺], it was apparent that k_tr was similar at force and stiffness levels common to both substrates. However, with dATP both k_tr and force rose substantially above maximal levels for ATP for trabeculae from both untreated (Fig. 5 a) and PTU-treated (Fig. 5 b) rats, indicating the effect was independent of cardiac myosin isoforms (Regnier et al., 2000b). Changes in the force-k_tr relationship at high levels of Ca²⁺ with dATP could result from greater activation of thin filaments by cross-bridges, as suggested by the increased stiffness (Fig. 2). Evidence for elevation of the rate force develops during submaximal Ca²⁺ activation by enhanced cross-bridge binding was provided by Fitzsimons et al. (2001), who found that NEM S-1 binding to thin filaments at submaximal forces raised k_tr to levels measured during maximal Ca²⁺ activations. Since NEM S-1 cross-bridges are strong binding and slowly dissociating, they suggested that thin filaments with bound NEM S-1 were at a much higher activation level than in the absence of NEM S-1 at the same level of submaximal Ca²⁺, thus allowing a faster rate of force development. An additional effect of dATP may be directly attributed to a faster rate of cross-bridge cycling with dATP since maximal k_tr is the sum of f_app + g_app, both of which could be increased by dATP (Regnier et al., 2000b). However, the similar effect of dATP on k_tr-force relations (Fig. 5; Table 1) for trabeculae from untreated (mostly α-MHC) and PTU-treated (β-MHC) rats indicates that cross-bridge cycling rate cannot be the only factor limiting the level and rate of force development in cardiac muscle. If k_tr at submaximal levels of activation is limited by the intrinsic acto-myosin NTPase rate of cross-bridges, we should have observed an elevation of k_tr at all levels of force. In contrast, Fig. 5 shows similar values of k_tr for ATP and dATP at the same level of force and stiffness, imply that a noncross-bridge cycling mechanism is limiting the rate that force can develop. An alternative interpretation of our results would be that the amount of force generated per cross-bridge is enhanced with dATP (compared with ATP). However, our previous observation that maximal Ca²⁺-activated force is not elevated by dATP in either fast or slow soleus skeletal muscle fibers argues against this. Slow skeletal fibers contain the same myosin isoform (β-MHC) present in trabeculae from PTU-treated rats, indicating dATP-induced elevation of force does not result from an increased level of force per cross-bridge. This suggests that the elevation of force and k_tr by dATP result from cross-bridge dependent enhancement of thin filament activation.

In cardiac muscle the relatively low Ca²⁺-activation dependence of k_tr (a two- to fivefold change compared with a 10–15-fold change for skeletal muscle) could result from an intrinsically lower acto-myosin ATPase activity at maximal activation or a relative limitation of the level of thin filament activation, even at saturating [Ca²⁺]. The last possibility is supported by the observations that 1), cardiac TnC (cTnC) has a single regulatory Ca²⁺-binding site (Johnson et al., 1980) compared with the two regulatory Ca²⁺-binding site for fast skeletal muscle; 2), cTnC undergoes less and slower conformational change upon Ca²⁺-binding than skeletal TnC (Sia et al., 1997; Hazard et al., 1998); 3), the Ca²⁺ induced conformational change associated with TnI interaction is slow for cardiac TnC (perhaps as low as 4–6 s⁻¹) (Dong et al., 1996, 1997); and 4), when cardiac TnC is fully substituted for endogenous TnC in skeletal fibers maximum force is significantly lower (Chase et al., 1994; Morris et al., 2001). An additional consideration is that the signal transmission of Ca²⁺ binding to TnC to tropomyosin could be slow relative to cross-bridge cycling rates, especially at low Ca²⁺ concentrations. If the level or rate of thin filament activation is limited in cardiac muscle, interventions that elevate thin filament activation might be expected to enhance both maximum force and k_tr, as we observe with dATP. We have previously demonstrated that thin filament activation dynamics are rate limiting to force development at low levels of Ca²⁺ in skeletal fibers (Chase et al., 1994; Regnier et al., 1998). We now suggest this mechanism may also apply to cardiac muscle.

This hypothesis is further supported by comparisons of k_tr and k_Ca during the same activation of trabeculae at pCa 4.5. For both ATP and dATP k_Ca was consistently only about half as fast as k_tr (Table 2, Figs. 6 and 7), indicating the dynamics of thin filament activation by Ca²⁺ limit the rate of force development. However, because force develops from different levels during k_tr and k_Ca measurements, differences in time courses of force rise could be influenced by the distribution of compliance between cross-bridges and noncross-bridge components (Huxley et al., 1994; Luo et al., 1994; Higuchi et al., 1995; Martyn et al., 2002). Comparison of the force traces in Fig. 7 indicates that the k_tr force redevelops from a nonzero value, whereas it rises from relaxed levels after the Ca²⁺ transient. As a result, in k_tr measurements the compliance of noncross-bridge structures would be high relative to cross-bridge compliance at the beginning of force development, resulting in an apparent slowing of force kinetics (Luo et al., 1994; Martyn et al., 2002) and thus would underestimate the difference between k_tr and k_Ca. In comparison, during k_Ca measurements noncross-bridge compliance would be lowest at the beginning of force rise, thus the initial rate of force development would be faster, slowing as cross-bridge compliance decreases with increasing force. However, there was no indication of an initial rapid rise of force that approached k_tr during k_Ca measurements. Thus differences in distribution of compliance between cross-bridge and noncross-bridge components do not explain the different time course of k_tr and k_Ca.

If cross-bridge cycling kinetics were limiting the rate of force development, we would expect k_Ca and k_tr to be similar. The thin filament activation state should be in near equilibrium during tension redevelopment with the k_tr protocol. Changes in myoplasmic [Ca²⁺] during tension transients (Vandenboom et al., 1998; Wang and Kerrick, 2002) or Ca²⁺ binding when force is inhibited in cardiac muscle (Wang and Fuchs, 1994; Martyn and Gordon, 2001) indicate that some Ca²⁺ dissociates from thin filaments during the release-restretch transients of k_tr measurements. However, this amount is likely to be a small fraction of Ca²⁺ bound to troponin C at pCa 4.5 (Martyn and Gordon, 2001). Furthermore, any thin filament deactivation during the release phase of the release-restretch transient would lead to a slower value of maximal k_tr, and thus underestimate differences between k_tr and k_Ca. In contrast, force development during the k_Ca measurement begins with near zero levels of Ca²⁺ bound to thin filaments. Thus our observation that k_Ca is always slower than k_tr supports a hypothesis that thin filament activation by Ca²⁺ is rate limiting for force development even at saturating [Ca²⁺]. Previously Palmer and Kentish (1998) reported no difference between k_act (reported here as k_Ca) and k_tr in rat and guinea pig trabeculae at 22°C. However these measurements were not obtained during the same activation and may have been from separate trabeculae. Additionally, their k_act values in rat trabeculae varied almost fourfold with some activations producing more force than the control activations, and maximum k_tr varied by more than twofold. This suggests some of their measurements may have been made in submaximally activating conditions. The argument that thin filament regulatory protein dynamics limit the rate that force develops in cardiac muscle is further supported by the observation that both rates were elevated ∼60% by dATP, whereas k_Ca remained slower than k_tr (Table 1). Thus, even though cross-bridge cycling kinetics can be increased by dATP, the kinetics of thin filament activation by Ca²⁺ may limit the rate force develops in rat cardiac muscle. If so, mechanisms that enhance cross-bridge binding at a given submaximal level of intracellular Ca²⁺, such as phosphorylation of myosin-binding protein C could be important in increasing the power output of the heart during sympathetic stimulation. This mechanism would not be available to skeletal muscle, where Ca²⁺ binding to TnC appears to be much more successful in fully activating thin filaments and where contraction usually results from action potential trains that allow muscles to contract at the upper end of the force-pCa relationship.