The Rates of Ca2+ Dissociation and Cross-bridge Detachment from Ventricular Myofibrils as Reported by a Fluorescent Cardiac Troponin C*

Background: The rate-limiting step of cardiac muscle relaxation is not completely understood. Results: We were able to measure two proposed rate-limiting steps of relaxation in ventricular myofibrils. Conclusion: The rate of Ca2+ dissociation from troponin C may be rate-limiting during myofilament inactivation under physiological conditions. Significance: Strategies that target both troponin C and myosin will be needed to treat diastolic dysfunction. The rate-limiting step of cardiac muscle relaxation has been proposed to reside in the myofilament. Both the rates of cross-bridge detachment and Ca2+ dissociation from troponin C (TnC) have been hypothesized to rate-limit myofilament inactivation. In this study we used a fluorescent TnC to measure both the rate of Ca2+ dissociation from TnC and the rate of cross-bridge detachment from several different species of ventricular myofibrils. The fluorescently labeled TnC was sensitive to both Ca2+ dissociation and cross-bridge detachment at low Ca2+ (presence of EGTA), allowing for a direct comparison between the two proposed rates of myofilament inactivation. Unlike Ca2+ dissociation from TnC, cross-bridge detachment varied in myofibrils from different species and was rate-limited by ADP release. At subphysiological temperatures (<20 °C), the rate of Ca2+ dissociation from TnC was faster than the rate of cross-bridge detachment in the presence of ADP. These results support the hypothesis that cross-bridge detachment rate-limits relaxation. However, Ca2+ dissociation from TnC was not as temperature-sensitive as cross-bridge detachment. At a near physiological temperature (35 °C) and ADP, the rate of cross-bridge detachment may actually be faster than the rate of Ca2+ dissociation. This provides evidence that there may not be a simple, single rate-limiting step of myofilament inactivation.

Cardiac function is dynamically regulated to meet the demands of the human body. However, in many cardiovascular diseases, relaxation becomes impaired, which can lead to diastolic dysfunction of the heart (1,2). A more complete understanding of the molecular mechanisms that govern cardiac muscle relaxation will help to develop better treatment strategies for diastolic dysfunction. Biochemically, there are two main factors thought to control the rate of relaxation; the decline of the intracellular Ca 2ϩ transient and the inactivation of the myofilament (3,4). Although the removal of intracellular Ca 2ϩ is a vital component to relaxation, it would appear that myofilament inactivation is equally important if not rate-limiting (4,5). For the myofilament to relax, Ca 2ϩ must dissociate from troponin C (TnC) 2 to inactivate the thin filament and cross-bridges must detach from actin to alleviate the force (6). It is generally thought that the rate of cross-bridge detachment is substantially slower than the rate of Ca 2ϩ dissociation from TnC and thus rate-limits myofilament inactivation (7,8).
ATP binding to the actomyosin complex detaches myosin from actin. However, this ATP cannot complete the crossbridge cycle and detach myosin until the previous hydrolysis products of ATP (phosphate (P i ) and ADP) are released (9). It is believed that P i release is associated with either the transition of myosin from a weakly bound state to a strongly bound state or with the force-producing power stroke (10 -12). In cardiac muscle, ADP release is thought to follow the power stroke and must dissociate before another ATP can bind and detach myosin (13,14). The rate of ADP dissociation from myosin has been proposed to rate-limit cross-bridge detachment and potentially relaxation (15,16). In support of this idea, increasing concentrations of ADP slowed the rate of cross-bridge detachment from actomyosin (15) and slowed the rate of relaxation of skeletal myofibrils and skinned cardiac muscle (17,18). Furthermore, the rate of ADP dissociation from actomyosin and the rate of cardiac muscle relaxation were substantially slower when each system utilized ␤-myosin as compared with ␣-myosin (19). However, ADP dissociation has only been measured from unregulated actomyosin and was quantitatively one order of magnitude faster than the rate of cardiac muscle relaxation (15,18,20,21). Either the rate of ADP dissociation does not rate-limit cardiac muscle relaxation or the actual rate of ADP dissociation is slower when measured from within the confines of the sarcomere.
The other component of myofilament relaxation is the inactivation of the thin filament, which is controlled by Ca 2ϩ dissociation from TnC. Ca 2ϩ dissociates from isolated TnC at least 2 orders of magnitude faster than the rate of myocardial relaxation (22)(23)(24). Because of this fact, Ca 2ϩ dissociation from TnC was deemed an insignificant regulator of relaxation. However, TnC does not function in isolation but as an integral component of the thin filament system. Incorporation of TnC into the thin filament slowed the rate of Ca 2ϩ dissociation by at least one order of magnitude (25). The binding of myosin-S1 to the thin filament further slowed the rate of Ca 2ϩ dissociation from TnC one additional order of magnitude (25). These biochemical studies suggest that the rate of Ca 2ϩ dissociation can be slowed to a rate that could influence the rate of relaxation. A recent study increased the complexity of the biochemical system even further by using ventricular myofibrils (21). This study concluded that the rate of Ca 2ϩ dissociation from TnC in isolated cardiac myofibrils was still too rapid to rate-limit myofibril relaxation at 10°C. However, the rate of Ca 2ϩ dissociation from cardiac TnC is little affected by temperature (26), whereas cross-bridge function is highly temperature-sensitive (20). The actual rate of Ca 2ϩ dissociation from TnC and the rate of crossbridge detachment in ventricular myofibrils at physiological temperature are unknown.
To further investigate the rate-limiting steps of myofilament inactivation, we utilized a fluorescent TnC that was able to report the rate of Ca 2ϩ dissociation from TnC and cross-bridge detachment in ventricular myofibrils over a wide range of temperatures. Myofibrils provided a unique biochemical system in which both Ca 2ϩ dissociation and cross-bridge detachment could be measured from within the confines of the regulated sarcomere, which contained the myofilament proteins at a physiologically relevant geometry and stoichiometry.
Cross-bridge detachment in the absence of ADP was always faster than the rate of Ca 2ϩ dissociation from TnC. However, the addition of ADP substantially slowed the rate of crossbridge detachment across all temperatures. This would suggest that cross-bridge detachment is rate-limited by ADP dissociation. At temperatures below 20°C, the rate of cross-bridge detachment in the presence of ADP was slower than the rate of Ca 2ϩ dissociation. This would suggest that cross-bridge detachment rate-limits myofilament inactivation and potentially relaxation at cold temperatures. However, the rate of cross-bridge detachment was more sensitive to changes in temperature than was the rate of Ca 2ϩ dissociation. At a more physiological temperature (35°C) and ADP, the rate of cross-bridge detachment may actually be faster than the rate of Ca 2ϩ dissociation. This provides evidence that there may not be a simple, single rate-limiting step of myofilament inactivation. Thus, Ca 2ϩ dissociation and cross-bridge detachment may both influence the rate of relaxation and may be potential therapeutic targets for improving relaxation across a broad range of cardiomyopathies.
Protein Purification, Labeling, and Complex Formation-The human cardiac TnC with the C35S, T53C, and C84S mutations (herein designated as TnC T53C ) was expressed and purified as previously described (25). The D65A mutation was inserted into the TnC T53C expression vector using techniques previously described (25). The mutations were confirmed by DNA sequence analysis. Expression and purification of D65A TnC T53C was conducted as previously described (27). Human cardiac TnI and TnT (isoform 3) were expressed and purified as previously described (25). TnC T53C was labeled with the environmentally sensitive thiol-reactive fluorescent probe IANBD (ϳ80% labeling efficiency) following a protocol previously described for IAANS labeling (28). The Tn complexes were reconstituted as previously described (27).
Myofibril Preparation-Ventricular cardiac muscle was obtained from male New Zealand White rabbits (2-3 months old, ϳ2 kg in weight (29)), male LBN-F1 rats (175-225 g), heartworm free mongrel dogs (weighing 19.0 Ϯ 0.4 kg; 2-3 years old (29)), and de-identified failing human tissue from ongoing studies with collaborators. All ventricular muscle was dissected in a Krebs-Henseleit solution containing 137 mM NaCl, 5 mM KCl, 1.2 mM MgSO 4 , 1.2 mM NaH 2 PO 4 , 20 mM NaHCO 3 , 10 mM glucose, 0.25 mM CaCl 2 , and 20 mM 2,3-butanedione monoxime to prevent contraction and damage to the cut tissue. The ventricular myofibrils were prepared as previously described (30). Briefly, the isolated ventricular tissue was minced with scissors in Krebs-Henseleit solution and then homogenized with 10-s bursts of a Polytron homogenizer. The suspension was filtered through cheesecloth to remove connective tissue and then further Dounce-homogenized. The myofibrils were collected by centrifugation and washed in rigor buffer before resuspension in glycerol rigor buffer. The myofibril stocks were stored at Ϫ20°C. All animals and tissues were handled in accordance with the National Institutes of Health Guidelines and approved by the Institutional Laboratory Animal Care and Use Committee at The Ohio State University. All human tissue was obtained in accordance with a tissue acquisition protocol approved by the Biomedical Sciences Institutional Review Board of The Ohio State University.
Exchange of Labeled Tn into Myofibrils-0.5 ml of myofibril stock (ϳ20 mg/ml) was resuspended in 1 ml of Buffer A (10 mM MOPS, 150 mM KCl, 3 mM MgCl 2 , 1 mM DTT, and 0.02% Tween 20 (pH 7.0)). The myofibrils were washed 3 times in Buffer A to remove the cryogenic glycerol solution. Briefly, the myofibrils were pelleted at 2000 ϫ g for 1 min, and the supernatant was removed and resuspended in 1 ml of Buffer A by gently pipetting. After the final centrifugation and removal of the supernatant, the myofibril pellet was resuspended in 1 ml of IANBD-labeled Tn (ϳ6 M stock in Buffer A) and stored at 4°C overnight. After the overnight Tn exchange, the myofibrils were washed three times with Buffer A to remove the un-exchanged Tn. After the last centrifugation, the myofibrils were resuspended in a total volume of 5.0 ml of Buffer A and filtered through 100-m nylon mesh (Sefar Inc.) to remove any large pieces of myofibrils that could potentially clog the stopped-flow apparatus. For certain experiments the myofibrils were crosslinked for 90 min as previously described (31). Briefly, myofibrils in Buffer A without DTT were cross-linked with 2 mM N-(3-dimethylaminopropyl)-N-ethyl-carbodiimide hydrochloride and 5 mM N-hydroxysulfosuccinimide sodium salt at 4°C for 90 min, and the reaction was quenched with the addition of 25 mM glycine and 10 mM DTT. The myosin concentration of the myofibrils was determined by weight per volume. An aliquot of myofibrils from the glycerol stock was washed in water three times. The myofibrils were then dried by vacuum centrifugation for 2 h at 3000 ϫ g and weighed. Assuming that the myosin concentration is ϳ43% that of the total myofibril protein by weight (32), the myosin head concentration of the stock was determined to be ϳ18 M, with a final concentration after mixing in the stopped-flow of ϳ0.8 M.
Determination of Ca 2ϩ Dissociation and Cross-bridge Detachment Kinetics from Myofibrils-The kinetic values were measured using an Applied Photophysics Ltd. (Leatherhead, UK) model SX.18 MV stopped-flow instrument with a dead time of 1.4 ms at 15°C. IANBD fluorescence was excited at 470 nm and monitored using a 500-nm long-pass interference filter from Newport (Irvine, CA). The experimental conditions for the Ca 2ϩ dissociation and cross-bridge detachment kinetics are described in the figure legends. Both ATP and ADP contained an equimolar concentration of Mg 2ϩ so as to maintain a constant concentration of free Mg 2ϩ . The data were fit using a program (by P. J. King, Applied Photophysics) that utilizes the nonlinear Levenberg-Marquardt algorithm. Each Ca 2ϩ dissociation and cross-bridge detachment event represents an average of at least three separate experiments, each averaging at least five traces fit with either a single-or double-exponential equation. The total increase in fluorescence upon Ca 2ϩ dissociation or the total decrease in fluorescence upon cross-bridge detachment was an ϳ1.2-fold change in fluorescence intensity.
Determination of the S1 Detachment Rate from Reconstituted Thin Filaments-Regulated thin filaments were prepared and reconstituted as previously described using TnC IAANS T53C Tn (27). IAANS fluorescence was excited at 330 nm and monitored using a 510-nm broad spectrum filter from Newport (Irvine, CA). Varying concentrations of rabbit skeletal myosin S1 were added to the thin filament and then rapidly mixed with 2 mM ATP to detach the S1 from the regulated actin. The experimental conditions are described in the figure legends.
Statistical Analysis-All data are expressed as the mean (ϮS.E.). Statistical significance of the data was determined by using one-way analysis of variance followed by a post-hoc least significance difference test using GraphPad Prism 4 (La Jolla, CA). Statistical significance was defined as p Ͻ 0.05.

RESULTS
Measuring Ca 2ϩ Dissociation-We have previously shown that TnC T53C labeled with IAANS reported the Ca 2ϩ binding properties of reconstituted Tn and thin filaments with high fidelity (25). Although TnC T53C IAANS reported similar Ca 2ϩ dissociation and cross-bridge detachment kinetics from myofibrils (data not shown), the signal-to-noise ratio was at least half that reported by TnC T53C labeled with IANBD (TnC IANBD  T53C ). Fig. 1A shows the apparent rate of Ca 2ϩ dissociation from TnC IANBD T53C reconstituted into the Tn complex (40.8 Ϯ 0.5/s) at 15°C, which was nearly identical to that previously reported by the IAANS probe (25). Unlike the fluorescence signal from Tn labeled with IANBD at Cys-84 in TnC (21), the complete fluorescence change was observed between the Ca 2ϩ -saturated and Ca 2ϩ -free base lines of TnC IANBD T53C Tn. Similarly, upon exchange of the Tn complex containing TnC IANBD T53C into rabbit ventricular myofibrils in rigor, the complete fluorescence change also occurred between the Ca 2ϩ -saturated and Ca 2ϩ -free base lines, with an apparent rate of Ca 2ϩ dissociation at 25.3 Ϯ 0.7/s at 15°C (Fig. 1B). The inset of Fig. 1B shows that TnC IANBD T53C Tn incorporated into the thin filaments of the myofibrils and replaced ϳ60% of the endogenous Tn as determined by the Western blot ratio of endogenous to exogenous TnI (data not shown). A similar rate of Ca 2ϩ dissociation was observed from the myofibrils when endogenous TnC was extracted and reconstituted with TnC IANBD T53C (21.1 Ϯ 0.9/s; data not shown); however, the signal-to-noise ratio was approximately half that observed with Tn exchanged myofibrils. Thus, TnC IANBD T53C was capable of measuring the apparent rate of Ca 2ϩ dissociation from the Tn complex, which slowed ϳ2-fold upon incorporation into the rigor cardiac myofibrils.
To verify that the observed rate of Ca 2ϩ dissociation occurred from the regulatory domain of TnC, Ca 2ϩ binding to the N-terminal domain of TnC T53C was abolished by the D65A mutation (33). The fluorescence intensity of reconstituted D65A TnC IANBD T53C Tn and Tn-exchanged myofibrils was insensitive to changes in Ca 2ϩ (data not shown). Thus, the observed kinetics of TnC IANBD T53C Tn and TnC IANBD T53C myofibrils can be attributed to the apparent rate of Ca 2ϩ dissociation from the N-terminal, regulatory domain of TnC.
Effect of Myofibril Species on the Rigor Ca 2ϩ Dissociation Rate-It is well established that isoforms of the thick and thin filament proteins vary across large and small species of mammals (9,19,34,35). It is currently unknown whether these different protein isoforms can affect the rate of Ca 2ϩ dissociation from TnC. Fig. 2 shows the apparent rates of Ca 2ϩ dissociation from ventricular myofibrils in rigor from small, medium, and large sized mammals at 15°C. Compared with the Ca 2ϩ dissociation rate from rabbit myofibrils, the rates from rigor rat (16.7 Ϯ 0.5/s) and dog (21.6 Ϯ 0.7/s) were moderately, but statistically slower, whereas the rate from failing human ventricular myofibrils (22.9 Ϯ 0.9/s) was not significantly different.
Thus, the rigor Ca 2ϩ dissociation rate from human cardiac TnC IANBD T53C in different species of myofibrils was similar. Effect of ATP on the Observed Fluorescence Changes of TnC-In a normal relaxing heart, Ca 2ϩ dissociates from TnC, whereas cross-bridges actively cycle rather than in rigor. Unfortunately, the fluorescence intensity remained at the Ca 2ϩ -saturated base line and did not change when Ca 2ϩ was removed from the rabbit myofibrils by EGTA that were also simultaneously mixed with 2 mM ATP to allow cross-bridge cycling (Fig.  3A, Ca 2ϩ -saturated control versus EGTA ϩ ATP trace). This phenomenon was also observed when ATP was added to the Ca 2ϩ -saturated myofibrils before mixing with EGTA. However, when D65A TnC IANBD T53C myofibrils were rapidly mixed with 2 mM ATP, a rapid decrease in fluorescence intensity was observed at a rate of 203 Ϯ 11/s at 15°C (Fig. 3A, D65A versus EGTA ϩ ATP trace). This ATP-dependent fluorescence decrease occurred from D65A TnC IANBD T53C myofibrils in the absence or presence of Ca 2ϩ and originated from the rigor baseline fluorescence level (Fig. 3A, D65A Rigor trace). These data suggest that there are at least two processes influencing the fluorescence level of TnC IANBD T53C myofibrils during Ca 2ϩ dissociation in the presence of cycling cross-bridges.
Sensing Cross-bridge Detachment through TnC-Unlike D65A TnC IANBD T53C myofibrils, the fluorescence intensity did not change when Ca 2ϩ -saturated TnC IANBD T53C myofibrils were mixed with 2 mM ATP (Fig. 3B, Ca 2ϩ Saturated versus Ca 2ϩ ϩ ATP). However, the addition of 2 mM ATP to Ca 2ϩ -free TnC IANBD T53C myofibrils decreased the fluorescence at a rate of 163 Ϯ 8/s at 15°C (Ca 2ϩ Free Control versus ATP trace). Similar to D65A TnC IANBD T53C myofibrils, the ATP-dependent fluorescence decrease originated from the Ca 2ϩ -free rigor baseline (Ca 2ϩ Free Control trace). If this ATP-dependent decrease in fluorescence was related to cross-bridge detachment, then the addition of ADP to the myofibrils would be expected to slow the rate (15). As expected, the addition of 2 mM ADP to the Ca 2ϩ -free TnC IANBD T53C myofibrils substantially slowed the rate of the ATPdependent fluorescence decrease to 6.9 Ϯ 0.6/s (Ca 2ϩ Free Control ϩ ADP versus ATP trace). 2 mM ADP also slowed the ATP-dependent fluorescence decrease from D65A TnC IANBD T53C myofibrils to 6.4 Ϯ 0.2/s (data not shown). Thus, TnC IANBD T53C fluorescence was sensitive to both Ca 2ϩ binding and crossbridge dissociation, each of which decreased IANBD fluorescence.
Although changes in Trp fluorescence, light scatter, and various fluorescent nucleotides have been utilized to more directly observe cross-bridge detachment in other systems (15,20,36), none of these approaches were suitable with the ventricular myofibrils (data not shown). To further test the idea that the fluorescent TnC was sensitive to cross-bridge detachment, reconstituted thin filaments bound by myosin S1 were utilized. A fluorescent skeletal Tn was previously shown to be sensitive to myosin S1 binding and dissociation in reconstituted thin filaments (37). Similarly, Fig. 3C shows a decrease in TnC IAANS T53C fluorescence as increasing amounts of myosin S1 were detached from the reconstituted thin filaments by 2 mM ATP. At a myosin S1 to actin subunit ratio of 1:7, the fluorescence decreased at a rate of ϳ300/s. This rate, although substantially slower than the true rate of myosin S1 detachment from nucle- otide free reconstituted thin filaments, was similar to that observed by a fluorescent skeletal Tn (37) and that observed in the ventricular myofibrils of this study. Increasing the myosin S1-to-actin ratio to 2:7 and then 4:7 had little effect on the apparent detachment rates (268 Ϯ 11/s and 265 Ϯ 10/s, respectively) but linearly increased the amplitude of the signals. Similar to the ventricular myofibrils, in the presence of Ca 2ϩ , the fluorescence intensity of TnC IAANS T53C on the reconstituted thin filament was not sensitive to myosin S1 detachment (data not shown). Thus, it would appear that the fluorescent TnC in the ventricular myofibrils and reconstituted thin filament reported similar ATP-dependent decreases in fluorescence that are related to myosin detachment. Unfortunately, the addition of ADP to the reconstituted thin filaments caused myosin S1 detachment, complicating any further studies with ATP.
To further characterize the influence of ATP on the rabbit ventricular myofibrils, Fig. 4 shows the dependence of the concentration of ATP and ADP on the apparent rate of crossbridge detachment from the Ca 2ϩ -free TnC IANBD T53C myofibrils at 15°C. When the myofibrils were mixed with 5, 10, 50, and 100 M ATP (concentration before mixing), the apparent rate of cross-bridge detachment occurred at 0.51 Ϯ 0.02/s, 1.1 Ϯ 0.2/s, 5.5 Ϯ 0.6/s, and 16 Ϯ 2/s, respectively (Fig. 4A). Over a longer time, there was an additional slow decrease in fluorescence that occurred at ϳ0.02/s for all concentrations of ATP (Fig. 4B). Presumably as the ATP was depleted, the fluorescence slowly increased at ϳ0.03/s back toward the Ca 2ϩ -free rigor base line as the cross-bridges went back into rigor (Fig. 4B). Increasing amounts of ATP delayed the reverse in fluorescence and rebinding of rigor cross-bridges (Fig. 4B). Fig. 4C shows the ATP dependence on the rate of the initial fast phase of crossbridge detachment from Ca 2ϩ -free TnC IANBD T53C myofibrils. Previously, it was shown that increasing concentrations of ATP linearly accelerated the rate of rabbit cardiac myosin-S1 detachment from unregulated actin with an apparent second order rate constant of 2 ϫ 10 6 M Ϫ1 s Ϫ1 at 15°C (20). The apparent rate of cross-bridge detachment in Ca 2ϩ -free TnC IANBD T53C myofibrils increased linearly at an apparent second order rate constant of 2.8 ϫ 10 5 Ϯ 0.01 M Ϫ1 s Ϫ1 for concentrations of ATP between 0 and 250 M (Fig. 4C). At higher concentrations of ATP, the cross-bridge detachment rate hyperbolically plateaued at ϳ330/s (Fig. 4C). To show that these ATP-dependent fluorescent changes were not due to myofibril shortening, the insets of Fig. 4, C and D, show the transmitted light microscopy images of Ca 2ϩ -free rabbit myofibrils taken from the stoppedflow after mixing with 50 M ATP or 2 mM ATP, respectively. Thus, similar to skeletal myofibrils (38), the addition of ATP to   Fig. 1B was included as a reference point for the fluorescence decay induced by ATP. Panel C shows the time course of IAANS fluorescence decay from Ca 2ϩ -free reconstituted thin filaments containing TnC IAANS T53C Tn upon dissociation of myosin-S1. TnC IAANS T53C reconstituted thin filaments with a myosin S1 to actin subunit ratio of 1:7, 2:7 and 4:7 in Buffer A (without Tween 20) ϩ 5 mM EGTA were rapidly mixed with equal volumes of the Buffer A (without Tween 20) ϩ 5 mM EGTA ϩ 2 mM ATP. The rates of S1 detachment reconstituted at the ratios of 1:7, 2:7, and 4:7 concentrations were ϳ300/s, 268 Ϯ 11/s, and 265 Ϯ 10/s, respectively. All data were collected at 15°C. Ca 2ϩ -free ventricular myofibrils did not cause shortening of the myofibrils.
In muscle, ATP cannot dissociate myosin from actin until ADP dissociates from the actomyosin complex (for review, see Refs. 14 and 13). As previously shown, the ADP dissociation rate limits rabbit cardiac myosin-S1 detachment from actin at ϳ115/s at 15°C (20). Increasing concentrations of ADP in Ca 2ϩ -free TnC IANBD T53C myofibrils hyperbolically decreased the apparent rate of cross-bridge detachment when the myofibrils were mixed with 2 mM ATP (Fig. 4D). The rate of cross-bridge detachment began to plateau at ϳ25 M ADP, signifying an ADP dissociation rate of 13 Ϯ 1/s with an apparent affinity of 1.3 Ϯ 0.1 M. Further increases in ADP resulted in a mild linear decrease in the apparent rate of cross-bridge dissociation. At these high ADP concentrations, increasing the ATP concentration did not accelerate the apparent rate of cross-bridge detachment beyond the plateau rate of ϳ13/s (data not shown). Although increasing the concentration of P i can increase the rate of relaxation (17), increasing the P i concentration up to 20 mM in the myofibrils had no effect on the rate of rigor Ca 2ϩ dissociation or the rate of cross-bridge detachment with or without ADP (data not shown).
Species Dependence on the Cross-bridge Detachment Rate-It is well established that larger mammals have a higher percentage of the slow ␤-myosin, whereas smaller mammals have a greater percentage of the fast ␣-myosin (39,40). Thus, the rate of cross-bridge detachment in myofibrils from large and small mammals would be expected to differ (15,20). In the absence of ADP, only the failing human myofibrils (41 Ϯ 4/s) had a rate of cross-bridge detachment (induced by 2 mM ATP) that was significantly slower than that of the rat (159 Ϯ 6/s), dog (125 Ϯ 17/s), and rabbit myofibrils (163 Ϯ 8/s, Fig. 5A). However, compared with the rabbit (6.9 Ϯ 0.6/s), the rate of cross-bridge detachment in the presence of 2 mM ADP was significantly faster for rat myofibrils (42 Ϯ 3/s) and slower for the failing human myofibrils (2.3 Ϯ 0.3/s) but similar to dog myofibrils (6.3 Ϯ 0.6/s) (Fig. 5B). This data support the hypothesis that TnC IANBD T53C fluorescence is sensitive to cross-bridge detachment in ventricular myofibrils and that different myosin isoforms have different rates of cross-bridge detachment that appear rate-limited by ADP (41).
Recovery of the Ca 2ϩ Dissociation Signal with Slowed Crossbridge Detachment-In the presence of rapidly detaching cross-bridges, the Ca 2ϩ dissociation event from TnC IANBD T53C myofibrils could not be observed (Fig. 3A, Control versus EGTA ϩ ATP trace). However, when Ca 2ϩ was rapidly chelated from TnC IANBD T53C rabbit myofibrils with slowly detaching crossbridges due to 2 mM ADP, a biphasic fluorescence signal was observed (Fig. 6A, myofibrils ϩ ADP versus EGTA ϩ ATP trace). The signal originated from the Ca 2ϩ -saturated base line and then decayed back to the Ca 2ϩ -saturated base line. Under these conditions, the Ca 2ϩ dissociation event (increase in fluorescence) and cross-bridge detachment event (decrease in fluorescence) were both observed. A monophasic increase in fluorescence (Fig. 6B, Subtraction trace) resulted from subtracting the cross-bridge detachment event (Fig. 6A, Ca 2ϩ -free Myofibrils ϩ ADP versus ATP trace) from the biphasic fluorescence signal. The rate of this fluorescence increase (ϳ25/s) was nearly identical to that of rigor Ca 2ϩ dissociation. A biphasic fluorescence signal was also observed when Ca 2ϩ was dissociated from TnC IANBD T53C rabbit myofibrils, whereas cross-bridge detachment was slowed by decreasing the temperature (5°C, data not shown). Similarly, mixing the Ca 2ϩ -saturated myofibrils with EGTA and a reduced amount of ATP (200 M, before mixing) at 15°C also caused a biphasic signal (data not shown).
Additional studies were performed to examine the effect of ADP and cross-linking on the apparent rates of Ca 2ϩ dissociation and cross-bridge detachment. The addition of 2 mM ADP to Ca 2ϩ -saturated TnC IANBD T53C rigor rabbit myofibrils did not alter the apparent rate of Ca 2ϩ dissociation (24 Ϯ 1/s; Fig. 6B, Myofibril ϩ ADP versus EGTA) compared with that of nucleotide free rigor myofibrils. Although cross-linked myofibrils were no longer sensitive to ATP-dependent cross-bridge detachment events (no change in fluorescence observed upon mixing with ATP (data not shown)), the rate of Ca 2ϩ dissociation from cross-linked myofibrils in the absence (21.3 Ϯ 0.2/s; data not shown) or presence of ATP (22 Ϯ 1/s; Fig. 6B, Crosslinked Myofibril versus EGTA ϩ ATP) was also similar to that of rigor myofibrils. Thus, myofibrils with slowed or non-dissociating cross-bridges had similar apparent rates of Ca 2ϩ dissociation. Furthermore, cross-bridge detachment could be prevented by cross-linking the myofibrils.
Physiological Significance of Ca 2ϩ Dissociation and Cross-Bridge Detachment-Controversy abounds in the literature regarding the rate-limiting step of cardiac muscle relaxation. It is commonly thought that the rate of Ca 2ϩ dissociation from TnC is too rapid, and therefore, cross-bridge detachment would rate-limit relaxation (6,16,21,42). Consistent with this hypothesis, Fig. 7A shows that the apparent rate of cross-bridge detachment in the presence of 2 mM ADP was significantly slower than the rate of Ca 2ϩ dissociation from TnC IANBD T53C rabbit myofibrils at temperatures less than 25°C. However, at a more physiological temperature (35°C), the two rates were nearly identical.
Physiologically, the concentration of cytosolic ADP in the cardiac myocyte is thought to be ϳ30 M and can increase to mM levels during ischemia (43). Thus, ADP dependence on the apparent rate of cross-bridge detachment in TnC IANBD T53C myofibrils was determined at 35°C. Similar to the results at 15°C, at 35°C (in Fig. 7B) ADP hyperbolically decreased the apparent rate of cross-bridge detachment when the myofibrils were mixed with 2 mM ATP. The rate of cross-bridge detachment began to plateau at ϳ100 M ADP, signifying an ADP dissociation rate of 104 Ϯ 5/s with an apparent affinity of 4 Ϯ 1 M. At concentrations of ADP that were supraphysiological, there was a more pronounced linear drop-off in the rate of cross-bridge detachment at 35°C (Fig. 7B) than was observed at 15°C (Fig.  4D). At these supraphysiological ADP concentrations, increasing the ATP concentration increased the apparent rate of crossbridge detachment to that comparable to the plateau rate of ϳ100/s (data not shown). We suggest that this plateau rate is the rate of ADP dissociation (Fig. 7A, Maximal Proposed Rate of ADP Dissociation), which rate-limits cross-bridge detachment. This rate was measured for 15 and 35°C and calculated from the 2 mM ADP curve for 5 and 25°C. The rate previous measured in the calculated plateau rate for 5 and 25°C was based on the 1.7-fold increase of the detachment rate (as compared with the detachment rate at 2 mM ADP) that was observed for both 15 and 35°C. Thus, the apparent rate of cross-bridge detachment may actually be slightly faster than the rate of Ca 2ϩ dissociation at 35°C in the presence of ADP and saturating ATP but not at temperatures below ϳ20°C (Fig. 7A). The data suggest that there may not be a simple, single rate-limiting step for myofilament inactivation. Furthermore, both the rates of Ca 2ϩ dissociation and cross-bridge detachment may be slow enough to influence ventricular muscle relaxation.

DISCUSSION
The fluorescent TnC IANBD T53C was able to report Ca 2ϩ dissociation as well as ATP-dependent events. We hypothesized that the fluorescent TnC was capable of measuring the rate of crossbridge detachment. This hypothesis was supported by several lines of evidence in regard to the measured rate of cross-bridge detachment in the presence of EGTA (low Ca 2ϩ ): 1) increasing ATP hyperbolically increased the apparent rate (Fig. 4C); 2) upon ATP depletion the cross-bridges rebound and reversed the fluorescence signal, which took longer for increased concentrations of ATP; 3) in the absence of ADP (Fig. 4C), the maximal rate was similar to the rate of rigor cardiac muscle relaxation induced by ATP (44); 4) increasing concentrations of ADP hyperbolically decreased the apparent rate (Fig. 4D); 5) in the presence of ADP, the rate varied in a manner expected for different species of myofibrils (i.e. fastest in rat and slowest with the failing human; Fig. 5B); 6) TnC fluorescence was sensitive to S1 detachment from the reconstituted thin filaments; 7) the addition of ADP slowed the cross-bridge detachment rate in ventricular myofibrils to that comparable to the rate of relaxation for different cardiac muscles at similar temperatures (44 -46); 8) cross-linking the myofibrils abolished the rate.
In the absence of ADP, ATP-induced detachment of myosin from reconstituted skeletal and cardiac actomyosin is extremely fast (Ͼ500/s) (15,47). This may suggest that crossbridge detachment observed by TnC IANBD T53C is sensed and limited by movement of another myofilament protein (tropomyosin (Tm), TnI, or TnT) or that the rate of cross-bridge detachment from the nucleotide free myofibrils is slower than that measured from reconstituted actomyosin. The idea that Tm movement may limit the observed rate of cross-bridge detachment is consistent with skeletal thin filament studies where myosin detachment was Ͼ500/s as measured by light scatter but sensed at ϳ200/s by a fluorescent Tm (47). However, by following a change in Trp fluorescence, ATP also hyperbolically accelerated the rate of cross-bridge detachment in skeletal myofibrils (in the absence of ADP) (36). The maximal rate observed in skeletal myofibrils was similar to the maximal rate of cross-bridge detachment observed by TnC IANBD T53C in cardiac myofibrils. Furthermore, the rate of rigor relaxation induced by caged ATP (in the absence of ADP (44)) was also similar to the maximal rate of cross-bridge detachment sensed by TnC IANBD T53C . Thus, cross-bridge detachment in the sarcomere may be slower than that from a reconstituted filament. In any regard, it would appear that the other myofilament proteins (Tm, TnI, and TnT) can move on the thin filament at a rate that is significantly faster than either Ca 2ϩ dissociation from TnC or cross-bridge detachment in the presence of ADP.
The actual mechanism behind the ability of TnC IANBD T53C to sense cross-bridge detachment at low Ca 2ϩ (presence of EGTA) is unknown. However, there is ample evidence that the binding of myosin to actin in reconstituted thin filaments, myofibrils, and muscle can influence the fluorescence of multiple TnC constructs (21, 48 -50). It has been shown that crossbridges can alter the orientation of various helices within TnC exchanged into cardiac trabeculae (51). These changes in TnC structure are not thought to be caused by a direct interaction of myosin with TnC but through myosin ability to alter Tm position on actin (14). Thus, the structure of TnC could be directly altered by the movement of Tm or indirectly through the movement modifying TnC interactions with TnI and TnT. Regardless, TnC IANBD T53C was only able to report cross-bridge detachment in the absence of Ca 2ϩ . One explanation for this phenomenon is that Ca 2ϩ -bound TnC exists in a structural conformation that makes the fluorescent probe insensitive to cross-bridge detachment. However, in the presence of Ca 2ϩ and ATP, the myofibrils rapidly and irreversibly shortened, which may have compromised the fluorescence change.
Our data are consistent with the hypothesis that the rate of cross-bridge detachment is limited by the rate of ADP dissoci-ation. Previously, the rates of ADP dissociation from different cardiac myosins were inferred from the study of unregulated actomyosin in solution (20), laser trap experiments (44), and sinusoidal analysis on skinned cardiac muscle (52)(53)(54). The apparent rate of ADP dissociation measured by our fluorescent TnC is nearly one order of magnitude slower than that measured from actomyosin in solution for the different fast and slow myosins (20). This result may not be surprising considering that previous studies have shown that myosin S1 and singleheaded myosin have different properties than even heavy meromyosin (55). In agreement with our results, the apparent rates of ADP dissociation measured from the rabbit, rat, and human myofibrils were similar to the calculated rates of ADP dissociation measured by the laser trap assay (rabbit and rat) (41,44) and from the sinusoidal length perturbation analysis of crossbridge detachment (human, rabbit, and rat) (52)(53)(54).
There is evidence that the rate of ADP dissociation from various non-muscle and smooth muscle myosins is load-dependent (56 -59). This phenomenon has recently been suggested to be true for cardiac myosin (60). It may be that the cross-bridges in the myofibrils are under some sort of load that would affect the rate of ADP dissociation. Consistent with this idea, it is striking that the apparent rate of ADP dissociation from the myofibrils from different species in this study was similar to the reported rates of cardiac muscle relaxation from the respective species. Again, these data would suggest that in more complex physiological systems the rate of ADP dissociation is slowed and that cross-bridge detachment rate-limits relaxation at subphysiological temperatures.
Previous data from skinned skeletal muscle and cardiac myofibrils suggest that the rate of Ca 2ϩ dissociation from TnC can influence the duration and rate of relaxation (61,62). This was primarily determined by utilizing TnC mutants that possessed slower rates of Ca 2ϩ dissociation. To date, the rate of Ca 2ϩ dissociation from TnC in intact muscle has not been measured. Current evidence suggests that the rate of Ca 2ϩ dissociation from TnC may not be constant and is dependent upon strongly bound cross-bridges. For instance, the rate of Ca 2ϩ dissociation is slowed nearly one order of magnitude upon myosin-S1 binding to the thin filament (25). Consistent with these studies, extra bursts of Ca 2ϩ were released from the myofilament when myosin was forced to rapidly detach from skinned and intact cardiac muscle (63,64). It is believed that the rapid detachment of myosin from actin weakened the affinity of TnC for Ca 2ϩ and accelerated the rate of Ca 2ϩ dissociation. Interestingly, the rate of Ca 2ϩ dissociation from TnC in guinea pig ventricular myofibrils with detached cross-bridges was nearly one order of magnitude faster than the rate of relaxation (21). This study concluded that the rate of Ca 2ϩ dissociation from TnC could not rate-limit relaxation as it was substantially faster than that of relaxation. However, the initiation of relaxation occurs in the presence of strongly bound cross-bridges. Thus, we set out to measure the rate of Ca 2ϩ dissociation from TnC in Tn-exchanged ventricular myofibrils under conditions of strongly bound cross-bridges.
In rigor, the rate of Ca 2ϩ dissociation from TnC IANBD T53C myofibrils was slightly influenced by the species of myofibrils (Fig.  2). The addition of ADP to the rigor myofibrils had little effect on the rate of Ca 2ϩ dissociation from TnC (Fig. 6B). In the presence of ATP, the Ca 2ϩ dissociation signal could not be observed unless ADP was present, the temperature was decreased, a minimal amount of ATP was used, or the myofibrils were cross-linked. This data suggest that the rates of Ca 2ϩ dissociation from TnC in myofibrils with different strongly bound cross-bridge states are similar. However, even with a slower rate of cross-bridge detachment, the rate of Ca 2ϩ dissociation appears too fast to rate-limit myofilament inactivation and relaxation at subphysiological temperatures.
It has been suggested that the myofilament rather than Ca 2ϩ dynamics rate-limits relaxation at physiological temperature (4). It is clear that cross-bridge dynamics are highly temperature-sensitive with a Q 10 as high as 5 (65). On the other hand, the rate of Ca 2ϩ dissociation from cardiac TnC has been shown to have negligible temperature sensitivity (26). Consistent with these studies, the rate of cross-bridge detachment was substantially more affected by temperature than the rate of Ca 2ϩ dissociation from TnC in the cardiac myofibrils. Interestingly, at physiological temperature, the rate of Ca 2ϩ dissociation and cross-bridge detachment in the ventricular myofibrils were equivalent at 1 mM ADP (after mixing). However, at more physiological concentrations of ADP, the rate of cross-bridge detachment (ϳ100/s) was actually faster than the rate of Ca 2ϩ dissociation (ϳ60/s). Thus, under physiological conditions the rate of Ca 2ϩ dissociation may actually rate-limit myofilament inactivation. The maximal proposed rate of ADP dissociation (Fig. 7A) may be an overestimation of the rate as the myofibrils may be an unloaded system. Thus, in cardiac muscle the two proposed rate-limiting mechanisms for myofilament inactivation may actually be kinetically tuned with one another. This would suggest that slowing one of these two rates would slow relaxation, but to accelerate relaxation both rates would have to be increased. Therefore, therapeutic strategies designed to increase the rate of relaxation may need to target both the thin and thick filament.