INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ACTIVATION OF CARDIAC MUSCLE is a complex process involving multiple regulatory mechanisms. Control of myofilament activation is mediated primarily through the thin filament and thin filament-associated proteins, as well as intracellular signaling molecules that modify these myofibrillar components. The level of cardiac myofilament protein phosphorylation is an important determinant of physiological activity and has been implicated in the contractile dysfunction of several myocardial pathologies (1, 19, 20). Phosphorylation of cardiac troponin I (cTnI), a thin filament regulatory protein, appears to be of particular importance in modulating myofilament activity. cTnI functions in the troponin complex with cardiac troponin C (cTnC), a Ca²⁺ receptor protein, and cardiac troponin T (cTnT), a tropomyosin binding protein. Troponin and tropomyosin not only confer Ca²⁺ sensitivity to the actin-myosin interaction but also modulate myofilament activity through covalent and noncovalent modifications that alter Ca²⁺ sensitivity and maximum activity. For example, -adrenergic receptor-dependent phosphorylation of cTnI at serines 23 and 24 (S23/S24) promotes Ca²⁺ dissociation from cTnC (39) and induces reduced myofilament Ca²⁺ sensitivity (36). In association with down regulation of -adrenergic receptors in human heart failure (4), it has been inferred that levels of phosphorylation of cTnI at its protein kinase A (PKA) sites are depressed. In addition, protein kinase C (PKC)-dependent cTnI phosphorylation has been reported to affect maximum actomyosin MgATPase rate by controlling the number of actin-myosin interactions (15, 29-32, 44). Compared with controls, activation of PKC (3) and levels of cTnI phosphorylation (1) are increased in samples from failed human hearts. These data fit with the hypotheses that signals leading to compensatory hypertrophy and the transition to heart failure involve PKC-dependent phosphorylation of cTnI (7).

Despite strong circumstantial evidence, the role of cTnI phosphorylation, especially at the PKC sites, in the regulation of cross bridges remains unclear. A full understanding of the influence cTnI phosphorylation exerts over myofilament regulation is confounded by the presence of several sites of phosphorylation, multiple kinases and phosphatases, and various receptors acting through the PKC pathway. To approach this question, we generated a transgenic mouse in which two functionally significant cTnI amino acids, which are targets of PKC (S43/S45), are mutated to unphosphorylatable alanines (cTnI S43A/S45A). Previous in vitro studies (29) indicated that PKC-dependent phosphorylation at these near NH₂-terminal residues outweigh the effects of phosphorylation at Thr144, which is situated in the inhibitory region of cTnI. Furthermore, a study by MacGowan et al. (21) has shown that hearts expressing cTnI S43A/S45A have altered pressure-Ca²⁺ relationships, an effect that may originate at the level of the myofilaments. Our previous work (28) has demonstrated a diminished response of heart preparations from these mice to phenylephrine and phorbol esters.

In this study, we compared tension, stiffness, and actomyosin MgATPase rate in detergent-extracted fiber bundles from wild-type and transgenic cTnI S43A/S45A hearts. The preparations were studied in a basal state and after PKC activation by -adrenergic or endothelin (ET) receptor agonists. These agonists are known to stimulate PKC activity in cardiac myocytes (2, 5, 37) and to modify actin-myosin interaction, concomitant with increases in cTnI phosphorylation (6, 24, 44). Our results indicate that phosphorylation of S43 and S45 of cardiac myofilaments alters the economy of muscle contraction and that this region of cTnI influences protein phosphorylation at neighboring sites on cTnT.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Transgenic mouse model. Transgenic mice were generated as described by MacGowan et al. (21). cDNA for mutant cTnI, in which S43 and S45 were replaced with alanines, was expressed under the control of the cardiac-specific -myosin heavy chain promoter. The transgene was microinjected into FVBN zygotes, and embryos reimplanted into pseudopregnant FVBN mice. Mutant cTnI was detected using the polymerase chain reaction. Myofilaments from the transgenic mice contain ~50% cTnI S43A/S45A and 50% cTnI S43/S45 (28).

Isolation of papillary fiber bundles. Hearts were excised from 3- to 6-mo-old mice that were anesthetized with ether. The hearts were rinsed free of blood in ice-cold saline (0.9% NaCl) and left ventricular papillary fibers quickly dissected free. Some papillary fiber bundles were treated with receptor agonists/antagonists (see Agonist/antagonist treatment), whereas others were immediately placed in ice-cold high-relaxing solution (HR) and cut into fiber bundles as previously described (29). Triton X-100 (1% vol/vol final concentration) was added to chemically permeabilize the preparation.

Agonist/antagonist treatment. Left ventricular papillary muscles were placed in oxygenated (95% O₂-5% CO₂) Krebs-Henseleit solution at room temperature. Fibers were treated with 10 µM phenylephrine plus 1 µM propranolol (-adrenergic receptor/PKC activation) (5), 100 nM ET-1 [ET type A (ET_A) receptor/PKC activation] (14, 37), or were untreated (control) for 5 min. After agonist/antagonist treatment, membranes in the fiber bundles were extracted in ice-cold HR solution containing 1% Triton X-100 (vol/vol final concentration). Some papillary muscles were treated with PKC inhibitors chelerythrine chloride (2 µM) or bisindolylmaleimide (100 nM) for 10 min before and during the 5 min of agonist/antagonist treatment.

Isometric tension, stiffness, and actomyosin MgATPase activity. Measurements of isometric tension, stiffness, and actomyosin MgATPase activity were done according to the methods of de Tombe and Stienen (8). Permeabilized papillary fibers were attached to aluminum T clips and mounted to a displacement generator (model 6800, Cambridge Technology) at one end and to a force transducer (model AE 801, SensoNor) at the other. Sarcomere length was measured by laser diffraction and adjusted to 2.3 µm. Fiber bundles were equilibrated for 5 min in relaxing solution (see Solutions), followed by 2 min in preactivating solution (see Solutions). The preparation was subsequently immersed in a maximal activating solution (see Solutions). After the contraction, sarcomere length was readjusted to 2.3 µm. Fiber bundles were then exposed again to maximally activating Ca²⁺ concentration ([Ca²⁺]), and these values were taken as maximum levels. Succeeding contractions were done in activating solutions containing submaximal [Ca²⁺]. After each contraction, fiber bundles were incubated for 1 min in relaxing solution, followed by 2 min in preactivating solution. The final contraction was induced in activating solution containing maximally activating [Ca²⁺]. Only those fibers able to generate >80% of initial tension in their final contraction were kept for analysis. Sigmoidal tension-, actomyosin MgATPase activity-, and stiffness-[Ca²⁺] relations were fit by a nonlinear fit procedure to a modified Hill equation
where P is the parameter of interest (isometric tension, actomyosin MgATPase activity, or stiffness), max is the maximum value at saturating [Ca²⁺], EC₅₀ is the [Ca²⁺] at which 50% of maximum is reached, and H is the slope of the relationship (Hill coefficient). Over the course of the experiments, steady-state isometric tension deteriorates to a small extent, as previously reported (8, 17, 40). Steady-state isometric tension was, therefore, "corrected" for this rundown (8).

Myofibrillar protein phosphorylation. Backphosphorylation was carried out using a modified protocol from Karczewski et al. (16). Ventricular preparations were dissected from mouse hearts as described above and treated with receptor agonists/antagonists (see Agonist/antagonist treatment). Ventricular tissue was homogenized in sample lysis buffer (see Solutions), sonicated and vortexed for 1 min, and centrifuged at 16,000 g. The supernatant fraction was retained and stored at 80°C. The reaction mixture consisted of 40 µg total protein, 12.5 µl 2× buffer (see Solutions), 3.18 µl water, 5 mmol/l ATP, 5 µCi [-³²P]ATP (NEN DuPont), 0.1% Triton X-100 (vol/vol), 0.3 mmol/l 1-palmitoyl-2-oleoyl-sn-glycero-3-[phospho-L-serine], 0.2 mmol/l 1,2-dioleoyl-sn-glycerol, and 1.8 mmol/l PKC-. Recombinant PKC- was prepared in our laboratory according to the procedure of Medkova and Cho (26). Reactions were carried out for 75 min at 30°C, which was sufficient for maximum backphosphorylation (data not shown). The reaction was stopped by the addition of an equal amount of sample loading buffer (see Solutions), and proteins were resolved on 4% stacking and 12.5% separating Tris-glycine sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Proteins were visualized by Coomassie blue staining and exposed overnight in a phophor screen. Band densities of proteins from preparations treated with agonists and antagonists were divided with control densities to give a drug-to-control ratio. Band densities were determined with the use of Image software (Scion; Frederick, MD). In a particular preparation of wild-type or transgenic myofilaments, relatively light bands indicate relatively high levels of endogenous phosphorylation.

Solutions. The high-relaxing (HR) solution contained 10 mM EGTA, 25 µM CaCl₂, 20 mM 3-(N-morpholino)propanesulfonic acid, 50 mM potassium propionate, 6.8 mM MgCl₂, 12 mM phosphocreatine, and 5 mM Na₂ATP, pH 7.0. The relaxing solution contained (in mM) 8.37 MgCl₂, 5.80 Na₂ATP, 20 EGTA, and 42.5 potassium propionate, pH 7.1. The preactivating solution contained (in mM) 7.78 MgCl₂, 5.80 Na₂ATP, 0.50 EGTA, 19.5 HDTA, and 43.6 potassium propionate, pH 7.1. The maximum activating solution contained (in mM) 7.63 MgCl₂, 5.87 Na₂ATP, 20 Ca²⁺-EGTA, and 43.6 potassium propionate, pH 7.1. The relaxing, preactivating, and maximum activating solutions also contained 900 µM NADH, 100 mM N,N-bis[2-hydroxyethyl]-2-aminoethanesulfonic acid, 5 mM sodium azide, 10 mM phospho(enol)pyruvate, 1 mg/ml pyruvate kinase (500 U/mg), 0.12 mg/ml lactate dehydrogenase (870 U/mg), 10 µM oligomycin B, 20 µM P¹P⁵-di(adenosine-5')pentaphosphate, and 10 µM leupeptin, 1 µM pepstatin, 1 mM dithiothreitol, and 10 µM phenylmethylsulfonyl fluoride. The Krebs-Henseleit solution contained (in mM) 118.5 NaCl, 5 KCl, 1.2 MgSO₄, 2 NaH₂PO₄, 26.2 NaHCO₃, 10 glucose, and 200 µM CaCl₂. The sample lysis buffer contained (in mmol/l) 5 histidine-HCl (pH 7.4), 750 KCl, 0.2 dithiothreitol, 0.1 phenylmethylsulfonyl fluoride, 50 K₂HPO₄, 25 NaF, and 10 EDTA. The 2× buffer contained (in mmol/l) 80 histidine-HCl, pH 6.8, 20 MgCl₂, 30 NaF, and 2 EGTA. Sample loading buffer contained 20% glycerol, 2% sodium dodecyl sulfate, 0.25 mg bromophenol blue, 0.35 mM -mercaptoethanol, and 125 mM Tris base (pH 6.8).

Statistical analysis. All values are presented as means ± SE, and P < 0.05 was chosen to indicate statistical significance. Data from agonist/antagonist-treated fibers were analyzed using a one-way analysis of variance (ANOVA) and post hoc Dunnett's t-test or Student-Newman-Keuls test. Basal characteristics were analyzed with a Student's t-test. Backphosphorylation was analyzed with the use of one-way ANOVA and a least-significant difference t-test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of basal cTnI S43/S45 phosphorylation on mechanical and energetic characteristics of left ventricular papillary fiber bundles. In the first series of experiments, we determined the consequences of basal cTnI S43/S45 phosphorylation on myofilament Ca²⁺ activation. To examine these effects, we compared the myofilament activity of detergent extracted papillary fiber bundles from wild-type and transgenic mouse hearts in which cTnI S43/S45 had been mutated to alanines (Fig. 1 and Table 1). Figure 1 shows the relationships between free [Ca²⁺] and (A) isometric tension, (B) stiffness, and (C) actomyosin MgATPase activity. Table 1 summarizes all the data obtained with preparations in the basal state. The data indicate that tension and stiffness were less in transgenic versus wild-type myofilaments, whereas the actomyosin MgATPase rate was higher in the transgenic myofilaments. Maximum isometric tension was decreased by 13% and stiffness by 14% in fiber bundles from cTnI S43A/S45A mice, compared with wild-type controls. Actomyosin MgATPase activity of transgenic myofilaments at maximally activating [Ca²⁺] was 20% higher than in wild-type myofilaments. Wild-type and transgenic bundles demonstrated no significant differences in myofilament Ca²⁺ sensitivity or Hill coefficients for any parameter (Table 1). However, the combination of decreased tension development and increased actomyosin MgATPase in the transgenic preparations resulted in an increase in tension cost. This is illustrated in Fig. 2, which shows that the slope of the relation between actomyosin MgATPase rate and tension is significantly greater in cTnI S43A/S45A myofilaments than in wild-type controls.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Mechanical and energetic characteristics of left ventricular papillary fiber bundles from cardiac troponin I (cTnI) in which serines 43 and 45 were mutated to alanines (S43A/S45A) in transgenic mice and wild-type controls. Steady-state isometric tension (A), stiffness (B), and actomyosin MgATPase activity (C) were simultaneously measured in left ventricular papillary fiber bundles and plotted against Ca2+ concentration ([Ca2+]). Tension and stiffness in maximally activating [Ca2+] were decreased in fiber bundles containing cTnI S45A/S45A. Maximal actomyosin MgATPase activity was elevated in transgenic myocardium. Means ± SE are presented in Table 1.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Tension cost in cTnI S45A/S45A transgenic mice and wild-type mouse left ventricular papillary fiber bundles. Actomyosin MgATPase activity was plotted against steady-state isometric tension and fit by linear regression. Slopes of the fitted lines were taken as tension cost. Tension cost was measured in each fiber bundle and the means ± SE values are presented in Table 1. Figure shows the pooled actomyosin MgATPase activity-steady-state isometric tension relations. Data are pooled in 5 mN/mm2-wide steady-state isometric tension bins and fitted by linear regression.

View larger version (58K): [in this window] [in a new window]   Fig. 3.   Myofilament protein backphosphorylation. Autoradiographs of left ventricular papillary muscles from wild-type or cTnI S43A/S45A transgenic mice after endothelin (ET), phenylephrine plus propranolol (PP), or no treatment (CON) in the presence of [-32P]ATP. Means ± SE are presented in Table 3.

View this table: [in this window] [in a new window]   Table 2.   Effects of receptor agonists/antagonists on PKC--dependent incorporation of 32P into myofilament proteins

Role of cTnI S43/S45 in PKC-dependent myofilament regulation. In a second series of experiments, we investigated the role of cTnI S43/S45 in PKC-dependent myofilament regulation (Fig. 4 and Table 3). In the wild-type myofilaments, PKC activation after treatment with ET or phenylephrine plus propranolol decreased maximum isometric tension, stiffness, and actomyosin MgATPase activity (Fig. 4, A-C, respectively). PKC-coupled receptor stimulation did not affect myofilament Ca²⁺ sensitivity or Hill coefficients of any parameter in the wild-type myocardium (Table 3). The PKC inhibitor chelerythrine chloride blocked the effects of -adrenergic and ET receptor stimulation in wild-type cardiac muscle (Table 4). In contrast, fiber bundles containing cTnI S43A/S45A (Fig. 5) responded to PKC activation with ET or phenylephrine plus propranolol with increased maximum tension (Fig. 5A) and stiffness (Fig. 5B), compared with untreated transgenic fibers. Although phenylephrine plus propranolol treatment induced an increase in actomyosin MgATPase activity, ET receptor activation did not alter maximum MgATPase activity of transgenic papillary fiber bundles (Fig. 4C). Myofilament Ca²⁺ sensitivity and Hill coefficients were not altered after agonist/antagonist treatment in cTnI S43A/S45A myocardium. Chelerythrine chloride abolished the effects of ET treatment on transgenic cardiac muscle (Table 4), whereas neither chelerythrine chloride nor a second PKC inhibitor, bisindolylmaleimide (data not shown), altered the effects of -adrenergic receptor activation.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Mechanical and energetic characteristics of left ventricular papillary fiber bundles from wild-type mice treated with receptor agonists/antagonists. Steady-state isometric tension (A), stiffness (B), and actomyosin MgATPase activity (C) were simultaneously measured in left ventricular papillary fiber bundles treated with receptor agonists/antagonists and plotted against [Ca2+]. Maximum isometric tension, actomyosin MgATPase activity, and stiffness were all decreased after phenylephrine plus propranolol or ET treatment, compared with controls. Means ± SE are presented in Table 3.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Mechanical and energetic characteristics of left ventricular papillary fiber bundles from cTnI S45A/S45A transgenic mice treated with receptor agonists/antagonists. Steady-state isometric tension (A), stiffness (B), and actomyosin MgATPase activity (C) were simultaneously measured in left ventricular papillary fiber bundles treated with receptor agonists/antagonists and plotted against [Ca2+]. Maximum isometric tension and stiffness were increased after phenylephrine plus propranolol or ET treatment, compared with untreated cTnI S45A/S45A transgenic fibers. Actomyosin MgATPase activity in the presence of maximally activating [Ca2+] was increased by phenylephrine plus propranolol, but unchanged after ET treatment. Means ± SE are presented in Table 3.

Effects of cTnI S43A/S45A on PKC-dependent myofilament phosphorylation. The backphosphorylation protocol revealed that ventricular strips from wild-type hearts had an increase in cTnT and cTnI phosphorylation after ET treatment and cTnI phosphorylation after phenylephrine plus propranolol treatment (Fig. 3 and Table 2). As described in METHODS, bands of lower density are indicative of higher levels of protein phosphorylation. ET treatment also increased the phosphorylation of cTnT and cTnI in the transgenic myocardium expressing cTnI S43A/S45A. -Adrenergic activation increased cTnI and cTnT phosphorylation in transgenic tissue, although these increases were less than that observed after -adrenergic receptor activation in wild-type myocardium. Antagonism of PKC activation with chelerythrine chloride abolished the effects of -adrenergic and ET receptor activation in both wild-type and transgenic cardiac muscle (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There are several important and novel findings of the experiments reported here. Our data demonstrate that changes in the state of cTnI sites at S43 and S45 are associated with alterations in the economy of force development as revealed in our simultaneous determination of myofilament tension and ATPase rate. These results indicate that cross-bridge cycling may be regulated by covalent modifications in thin filament proteins. Our data also suggest that the state of the phosphorylation sites at S43 and S45 is an important determinant of the response of the myocardium to agonists that stimulate the PKC pathway. Moreover, in support of our previous finding (29), data presented here also show that modifications of phosphorylation sites on one troponin component may influence the state of phosphorylation of a near neighbor.

As demonstrated from the backphosphorylation data discussed in RESULTS, differences in tension cost between transgenic and wild-type myofilaments in the basal state are likely to be due to a difference in basal phosphorylation levels of cTnI and possibly cTnT. A precise molecular mechanism for this effect of cTnI phosphorylation at the PKC sites awaits further experiments. Possible mechanisms include changes in the step size of the cross bridges reacting, changes in duration of the displacement, or both. By analogy with changes in tension cost between V3 and V1 cardiac myosin heavy chain isoforms (34, 35, 41, 43), our hypothesis is that alterations in the region of S43 and S45 induce changes in the duration of unitary displacement with no change in unitary force. In optical trap experiments, the unitary step displacement of cardiac myosin, which has been estimated to be between 7 and 10 nm, was the same for V1 and V3 myosin (35, 43), as well as for R403Q and L408V myosin mutants (36, 45). There were, however, differences in the duration of unitary displacement ("cross-bridge duty cycle"). The duration was longer for V3 than for V1 myosin, consistent with data showing a reduced tension cost for heart muscle containing a preponderance of V3 myosin. Thus our data indicate that in addition to shifts in isoform population of myosin heavy chains, phosphorylation of cTnI at PKC sites may also alter the duration of unitary cross-bridge displacement. This is an important idea to test in that the isoform population of myosin heavy chains correlates linearly with power output of hearts and myofilaments (12, 25).

Our data support and extend results from earlier studies indicating that phosphorylation of S43 and S45 of cTnI is an important mechanism in the response of the heart to agonists that promote activity of the PKC pathway (15, 29-32, 44). Using cTnI mutants in which phosphorylation sites were changed to alanines, Noland et al. (29, 32) and Venema and Kuo (44) characterized the influence of specific cTnI phosphorylation sites on MgATPase activity of heavy meromyosin reacting with fully reconstituted thin filaments. These studies concluded that phosphorylation of cTnI S43/S45 by a variety of PKC isoforms reduces cross-bridge affinity for the thin filament, thereby inhibiting actomyosin MgATPase activity. At the time, it was unclear whether these PKC sites were important in situ. The phosphorylation had been accomplished using exogenous PKC in vitro in a protocol that potentially permitted phosphorylation of sites normally inaccessible in the intact myofilament lattice (29, 32) or were made on unloaded preparations containing multiple PKC substrates, which may have also affected cross-bridge cycling rate (44).

On the basis of this strong circumstantial evidence regarding the functional significance of cTnI S43 and S45, we generated mice harboring a transgene expressing cTnI S43A/S45A (24, 30). MacGowan et al. (21) reported that, compared with controls, hearts isolated from these transgenic mice had a significant alteration in the Ca²⁺-pressure relationship and increased susceptibility to a contracture induced by ischemia. This result fit with the study of Montgomery et al. (27), who reported a blunting of the effects of phenylephrine on tension generated by papillary muscle preparations from transgenic hearts. Montgomery et al. (28) also reported an attenuation of the reduction in maximum tension of skinned fiber bundles prepared from these papillary muscles after treatment with phorbol esters. The present study significantly extends these findings. Apart from our determinations of tension and ATPase rate of strained cross bridges, our protocol involved PKC activation by stimulation of receptors linked to PKC. This pathway of activation is more likely to produce increases in PKC activity within physiological limits and to limit the action of PKC phosphorylation to sites that are exposed in situ in the myofilament lattice.

The results of our studies using this mode of activation of the PKC pathway demonstrated different responses to phenylephrine plus propranolol and ET. In wild-type preparations, phenylephrine and ET treatment both decreased tension and actomyosin MgATPase rate with no change in tension cost. In the transgenic myocardium, however, phenylephrine treatment increased tension and cross-bridge cycling rate, whereas ET produced an increase in tension with a small decrease in actomyosin MgATPase rate. One explanation for these effects may be the activation of different PKC isoforms with the stimulation of different receptors. Consistent with the idea of differential PKC isoform activation accounting for varying effects on actomyosin MgATPase activities are our findings that -adrenergic receptor activation increased the phosphorylation of only cTnI in transgenic myocardium, whereas ET receptor activation increased cTnI phosphorylation as well as cTnT phosphorylation. A study by Clerk et al. (5) showed that the stimulation of -adrenergic or ET receptors in mouse hearts activates dissimilar combinations of PKC isoforms. Discrepancies in the literature with regard to the effects of PKC activation on contraction may also be due to differences in PKC isoform and trafficking in the activation process (22).

An interesting and unexpected finding of this work was the inability of two PKC inhibitors (chelerythrine chloride and bisindolylmaleimide) to abolish the functional effects of -adrenergic receptor stimulation in transgenic myocardium. However, backphosphorylation experiments (Table 4) showed that PKC inhibition blocks the -adrenergic receptor-dependent changes in myofilament protein phosphorylation. These two sets of results suggest that an additional PKC-independent pathway is activated by -adrenergic receptor stimulation in the transgenic heart. The key molecule involved in this signaling cascade has not been definitively identified. However, in addition to PKC, we list PKA as an unlikely candidate. We were unable to find any changes in the PKA-dependent phosphorylation of myofilament proteins (data not shown) and we did not note any decrease in myofilament Ca²⁺ sensitivity, a hallmark of PKA activation in the cardiac myocyte (8, 36).

Complex interactions among the thin filament proteins at the region surrounding cTnI S43/S45 may also account for a variety of responses to various agonists. This region of cTnI interacts with the COOH-terminal lobe of cTnC and the COOH-terminal globular region of cTnT. Studies on the myofilament Ca²⁺ sensitizer EMD 57003 revealed the significance of this region as a regulator of cross-bridge interactions with the thin filament. Wang et al. (45) reported that the docking site for EMD is at the C-lobe of cTnC in a region of interaction with a cTnI peptide containing S43/S45. Our finding that myofilaments containing cTnI S43A/S45A generated less maximum tension than wild-type myofilaments also supports the importance of this region of cTnI in tension regulation. The complexity of the protein-protein interactions also extends to effects of modifications of cTnI on phosphorylation of sites within cTnI and phosphorylation of its neighbor on the thin filament cTnT. In reconstituted troponin containing cTnI S43A/S45A, the PKC-dependent phosphorylation of cTnI T144 (29, 32) and an unidentified site (29) are decreased. Furthermore, cTnI S23/S24, the PKA-favored phosphorylation sites, can be cross phosphorylated by PKC- or - when S43/S45 are replaced with alanines (32). We suggest that the cross phosphorylation of S23/S24 by PKC is unlikely to explain the results presented here because we failed to observe any changes in myofilament Ca²⁺ sensitivity. Montgomery et al. (27) reported that switching of TnT isoforms from cTnT to fast skeletal TnT in cardiac myofilaments reduced the ability of PKC to phosphorylate cTnI. These reports indicate that changes in cTnI phosphorylation sites may influence the ability of kinases to phosphorylate sites within TnT and that changes in the phosphorylation of one myofilament protein can impact the covalent modification of other myofilament proteins. Whereas PKC activation decreases maximum isometric tension development in wild-type myocardium, the expression of cTnI S43A/S45A may influence the selectivity of PKC isoforms for specific troponin targets, allowing for significant phosphorylation of previously unfavored sites while altering interaction with preferred residues. Such changes would lead to unusual phosphorylation patterns on cTnI and possibly other myofilament proteins, modifying the effects of PKC-dependent phosphorylation on cross-bridge activity. In addition to the differences in phosphorylation sites, these complex effects may be responsible in part for our finding that -adrenergic or ET receptor stimulation of myocardium increased maximum isometric tension in myofilaments containing cTnI S43A/S45A but decreased maximum tension in wild-type preparations.

Understanding the link between PKC activation and cTnI phosphorylation is crucial to fully comprehend myofilament regulation, the progression of heart failure, and the underlying mechanism of cardioprotection. Whereas myofilament protein phosphorylation plays an important physiological role in regulating cardiac function, changes in cTnI phosphorylation also correlate with myocardial dysfunction (1, 19, 23, 46). Furthermore, the activation of pathways known to protect against postischemic cardiac dysfunction leads to increases in cTnI phosphorylation (38). PKC is critical in the mediation of cardioprotection (9, 10), as well as the development of heart failure (3, 11, 13, 18, 42). How PKC and cTnI interact to mediate this diverse range of effects is unknown. The resolution of this intracellular puzzle may be key to the development of therapies for cardiac disease.