|Home | About | Journals | Submit | Contact Us | Français|
Ischemia and heart failure are associated with protein kinase C (PKC) dependent phosphorylation of cardiac troponin I (cTnI). We investigated the effect of phosphorylation of cTnI PKC sites S43, S45 and T144 under normal (pH 7.0) and acidic (pH 6.5) conditions on tension in skinned fiber bundles from mouse heart. To mimic the PKC-phosphorylation, we exchanged troponin (cTn) in these fiber bundles with cTn complex containing either cTnI-(S43E/S45E) or cTnI-(T144E). We determined how pseudo-phosphorylation and acidic pH affect activation of thin filaments by strongly bound crossbridges by use of n-ethyl maleimide (NEM-S1) to mimic rigor. We hypothesized that PKC-phosphorylation of cTnI amplifies the effect of ischemic/hypoxic conditions to depress myofilament force and Ca2+ -responsiveness by reducing the ability of rigor crossbridge to activate force. Pseudo-phosphorylation of cTnI at S43/S45 exacerbated the effect of acidic pH to induce a rightward shift in the Ca2+ -tension relation. Under acidic conditions, fibers regulated by cTnI-(S43E/S45E) demonstrated a significant reduction in the ability of NEM-S1 to recruit cycling crossbridges, when compared to controls regulated by cTnI. Similar effects of pseudo-phosphorylation of cTnI-(T144) occurred, but to a lesser extent that those of pseudo-phosphorylation of S43/S45. We conclude that under acidic conditions PKC-phosphorylation of cTnI residues at S43/S45 and at T144 is likely to have differential, but significant effects in depressing the ability of both Ca2+ and rigor crossbridges to activate force generation. Although these effects of PKC dependent phosphorylation may be maladaptive in heart failure, they may also spare ATP consumption and be cardioprotective in ischemia.
There is substantial evidence that stresses inducing maladaptive cardiac growth [1;2] modify sarcomeric response to Ca2+ by mechanisms involving the protein kinase C (PKC) signaling cascade. A prominent PKC related mechanism potentially altering sarcomeric response to Ca2+ is phosphorylation of cardiac troponin I (cTnI) at S43, S45 and T144. Phosphorylation of cTnI at these sites induces a decrease in maximum tension, a right shift of the Ca2+ force relation, a decreased affinity of myosin S1 for actin and decreased think filament sliding speed in the motility assay [3;4]. Pseudo-phosphorylation of these sites has also been reported to stabilize an inactive state of actin . In the case of S43/S45, we  reported that mice with hearts expressing a mutant cTnI(S43A/S45A) had enhanced contractility. Moreover, when these mice were cross-bred with mice with severe dilated cardiomyopathy induced by over-expression of active PKCε, there was substantial reversal of the depression. A role for cTnI-T144 phosphorylation as part of the constellation of responses to cardiac stresses also appears important. Phosphorylation of this site via activation of PKCβII has been reported , and earlier studies reported an activation of PKCβII in human hearts at end stage failure . Moreover, we  recently reported evidence that with oxidative stress, PKCδ switches its substrate preferences from S23/S24 (sites on cTnI that depress Ca2+ sensitivity with little effect on maximum tension) to T144. Associated with this effect was a depression in maximum tension and rates of entry of crossbridges into a force generating state . There is also evidence that cTn phosphorylation is a critical element in the reduction in maximum of tension of skinned myocytes isolated from rat hearts stressed either by myocardial infarction (MI) or pressure overload (PO) [1;2]. In these studies, an increase in a highly charged form of cTnI (most likely phosphorylated at S43, S45 and T144) was correlated with the decline in tension. Moreover, treatment of skinned myocytes with PKCα mimicked the effect of MI and PO to depress tension and Ca2+ sensitivity .
The reduced blood flow, hypoxia and altered energy metabolism that accompanies ischemia and maladaptive cardiac growth is expected to depress intracellular pH, to alter the chemical environment of the sarcomeres, and to generate a population of strongly bound rigor crossbridges. Yet there have been no investigations determining the effects of these altered conditions on modification of myofilament function by PKC dependent phosphorylation. Additionally, an acidic environment induces enhanced PKC signaling that results in promoting the downward spiral of events that eventually leads to heart failure .
In experiments reported here, we investigated the effect of pseudo-phosphorylation of cTnI sites S43, S45 and T144 on tension of mouse cardiac skinned fiber bundles at pH 7.0 and pH 6.5. We measured the force-Ca2+ relations in detergent extracted fibers in which we exchanged native cTn with cTn containing either cTnI-(S43E/S45E) or cTnI-(T144E) to mimic PKC induced phosphorylation. We also determined the effect of pseudo-phosphorylation of cTnI at pH 7.0 and pH 6.5 on the ability of NEM-S1, a mimic of rigor crossbridges, to induce activation of skinned fiber tension. Compared to controls regulated by cTnI, fibers regulated by cTnI-(S43E/S45E) demonstrated a right shift of the Ca2+ -force relation and depression in maximum tension at pH 7.0 and pH 6.5. This desensitization to Ca2+ was greater at pH 6.5 than at pH 7.0.
In addition at acidic pH, the ability of strongly bound crossbridges to activate the thin filament was also significantly reduced in myofilaments controlled by cTnI-(S43E/S45E). The relative depression of these activities was greater with cTnI-(S43E/S45E) that with cTnI-(T144E). We conclude that in cardiac stress associated with ischemia and with activation of PKC, there is an exacerbation of the inhibitory effects of phosphorylation of cTnI that is likely to be an important mechanism in ischemia/reperfusion and in the progression to failure.
Human cardiac TnC, adult mouse cardiac TnT and wild-type (WT) mouse cardiac TnI were expressed and purified as previously described . Adult mouse cTnI was cloned into a pET3d vector and selectively mutated using the Quick-Change Site-Directed Mutagenesis kit (Stratagene). The constructs were verified by DNA sequencing and designated as cTnI-(S43E/S45E) or cTnI-(T144E). Reconstitution of hetero-trimeric Tn complexes was done as previously described .
Male CD-1 mice (age 3-4 months) were deeply anesthetized by intra-peritoneal injection of pentobarbital sodium (500 mg/kg body weight). Hearts were quickly excised and left ventricular papillary muscles were dissected into fiber bundles (4-5 mm long and 150-250 μm in diameter) in high relax buffer (HR) containing 20 mm MOPS pH 7.0, 10 mM EGTA, 1 mM free Mg2+, 5 mM MgATP-2, 12 mM creatine phosphate, 10 IU/ml creatine phosphokinase (Sigma) with free Ca2+ concentration at 10-9 M and ionic strength adjusted to 150 mM with KCl . To extract membranes, the fiber bundles were placed at 4°C in HR with 1% Triton X-100 for 4 h and then transferred to exchange buffer containing 20 mM MOPS pH 6.5, 20 mM KCl, 5 mM EGTA, 5 mM MgCl2, 1 mM DTT, with approximately 18 μM of recombinant Tn complex and incubated overnight at 4°C. All solutions contained a cocktail of protease inhibitors (1μg/ml pepstatin, 5 μg/ml leupeptin and 0.2 mM 4-(2-aminoethyl) benzenesulfonyl fluoride or AEBSF) .
Fiber bundles were mounted between a force transducer and a micro-manipulator with cellulose-acetate glue. Resting sarcomere length was set at 2.0 μm as determined using laser diffraction, cross sectional measurements were made, and isometric tension was recorded on a chart recorder . All experiments were carried out at 22°C. The myofilaments were initially placed in HR at pH 7.0 or pH 6.5 then switched to a maximally activating Ca2+ solution (10-4.5 M), and then fully relaxed in HR. We then measured force over a series of Ca2+ concentrations (10-8 M – 10-4..5 M). The fiber bundles were then immersed in a HR bath containing 6 μM NEM-S1. The effects of NEM-S1 on Ca2+ activated force have been previously described by Swartz and Moss . We isolated myosin S1 from rabbit fast skeletal muscle and modified it with NEM as previously described . After 15 min incubation in HR + NEM-S1, we carried out a second full series of Ca2+ activations at either pH 7.0 or pH 6.5. For force-Ca2+ measurements, Ca2+ was varied by mixing HR with HR to which CaCl2 had been added to achieve a range of concentrations at ionic strength 0.15 M and 10-9 M Ca2+. The multi-equilibria in the solutions used in these experiments at pH 7.0 and pH 6.5 were computed as previously described  using binding constants given by Godt and Lindley .
Statistical differences in tension developed before and after NEM-S1 treatment was determined by paired Student's t-test. All values are presented as mean ± S.E.M. with significance set at p < 0.05.
To determine the effects of phosphorylation of cTnI on Ca2+ dependent activation at neutral and acidic pH, we used two mutants, cTnI-(S43E/S45E) and cTnI-(T144E), which represent pseudo-phosphorylated forms of cTnI at PKC sites. We have previously provided evidence that these pseudo-phosphorylations mimic effects of authentic phosphorylation . Data, which are summarized in Table 1 and shown in Figure 1A, demonstrate the well-known effect of acidic pH to induce a rightward shift of the Ca2+-force relation in fiber bundles regulated by wild-type cTn. Data shown in Figure 1B, demonstrate that myofilaments in which endogenous cTn was exchanged with cTn reconstituted with cTnI-(S43E/S45E) exhibited both a rightward shift and a decrease in maximum tension at pH 7.0. At pH 6.5, the combination of acidic pH and pseudo-phosphorylation severely depressed activation in cTnI-(S43E/S45E) myofilaments (Figure 1B and Table 1). As shown in Figure 1C, myofilaments regulated by cTnI-(T144E) demonstrated a small but significant desensitization to Ca2+ compared to cTnI myofilaments at pH 7.0. However, this desensitization increased significantly with a drop to pH 6.5 (Table 1).
To determine the ability of strongly bound crossbridges to recruit tension generating endogenous crossbridges, we employed NEM-S1, a strongly binding mimetic of rigor bound crossbridges . Data in Figures 2 and and33 are displayed as a plot of tension versus sub-maximal Ca2+ concentration inasmuch as this is where normal heart function occurs and where NEM-dependent force activation is most prevalent due to a relatively large pool of non-cycling crossbridges available for recruitment to force generating crossbridges. In these figures a value of fold activation in the inset of each graph indicates the relative ability of NEM-S1 to activate tension by recruiting additional crossbridges in myofilaments regulated by either cTnI or the mutants, cTnI-(S43E/S45E) or cTnI-(T144E). Fold activation was computed by dividing the tension developed at a given [Ca2+] (0.32 μM in each example) after NEM-S1 treatment by the tension developed at that [Ca2+] before NEM-S1. In Figure 2A, we show tension data for cTnI-myofilaments before (solid line) and after (dotted line) NEM-S1 treatment at pH 7.0. A 4.2 fold activation of tension is indicated. Figure 2B shows the data obtained at pH 6.5. In this case there was a 12.7 fold activation of tension at the selected [Ca2+]. These data indicate that with a drop from pH 7.0 to pH 6.5, there was diminished tension development as expected, but there was also an enhanced ability of NEM-S1 to recruit force generating crossbridges in myofilaments regulated by cTnI. As shown in Figure 2C, replacement of cTnI with cTnI-(S43E/S45E) induced a significantly increased ability of NEM-S1 to recruit force generating crossbridges (21.1 fold activation) compared to cTnI controls at pH 7.0 (4.2 fold activation). We interpret this enhanced ability of NEM-S1 to activate tension of cTnI-(S43E/S45E) myofilaments to be due to a relative increase in the pool of non-cycling crossbridges available for recruitment into force generating crossbridges. However as shown in Figure 2D at pH 6.5, there was a severe depression in the ability of NEM-S1 to induce tension development in cTnI-(S43E/S45E) myofilaments (8.4 fold activation). These data indicate that despite the availability of endogenous crossbridges to be recruited into cycling force generating crossbridges under these acidic conditions, the presence of cTnI-(S43E/S45E) inhibited thin filament activation by NEM-S1.
Data in Figure 3 show results of a separate series of measurements comparing effects of Ca2+ and NEM-S1 induced activation on force generation of skinned fibers regulated by wild-type cTnI and by cTnI-(T144E) at pH 7.0 and pH 6.5. As with the data reported in Figure 2, data in Figs. 3A and 3B show that in cTnI myofilaments NEM-S1 induced recruitment of force generating crossbridges that was significantly enhanced at pH 6.5 (fold increase 13.4) compared with the effect at pH 7.0 (fold increase 1.9). In Figure 3C, the data show that myofilaments exchanged with cTnI-(T144E) displayed NEM-S1 induced activation of tension similar to the controls at pH 7.0. At pH 6.5 there was little change in the fold activation by NEM-S1 in the fibers containing cTnI-(T144E) (Figure 3D). The corresponding fold activation of 2.2 at pH 7.0 and 3.1 at pH 6.5 were not significantly different when compared to the fold activation of 1.9 in cTnI myofilaments at pH 7.0 (Figure 3A). Thus, in contrast to the fibers regulated by cTnI-(S43E/S45E) in which NEM-S1 induced activation increased significantly compared to controls at pH 7.0, this was not the case with cTnI-(T144E) myofilaments, which also had a relatively small inhibitory effect on Ca2+ activation. Yet, despite the enhanced inhibition of tension at pH 6.5 by cTnI-(T144E) myofilaments, there was a reduced ability of NEM-S1 to activate tension.
Our data are the first to demonstrate that acidic pH induces an enhanced inhibitory effect of cTnI pseudo-phosphorylated at PKC sites S43, S45 and T144 on Ca2+ and rigor-like crossbridge dependent thin filament activation. Our data also indicate that under these conditions there are differential effects of PKC phosphorylation of the near N-terminal region of cTnI and in the inhibitory peptide.
Potential mechanisms, which are responsible for the effects we have reported here, are couched in terms of current models of thin filament control mechanisms. The N-terminal region of cTnI-(39-58), containing the PKC sites S43/S45, binds with nano-molar affinity to cTnC . This region of cTnI also interacts with charged residues in a highly conserved region of cTnT that intertwines with cTnI to form the IT arm . As with the charge modifications associated with pseudo-phosphorylation of S43/S45, we have previously reported that a pH dependent charge modification (cTnI-A66H) in this region is critical in modifying the ability of NEM-S1 to activate the thin filament . Our binding data demonstrated that the charge change in this region modified the interaction of cTnI with cTnT. We speculated that this altered interaction with cTnI is transmitted to Tm, thereby affecting the ability of NEM-S1 to recruit force generating crossbridges. This may be the mechanism by which acidic pH and pseudo-phosphorylation at S43/S45 depress tension and NEM-S1 dependent activation. It is apparent that the altered response of skinned fibers regulated by cTnI-(S43E/S45E) to Ca2+ is also due in part to a depression of the affinity of cTnC for Ca2+ as reported by Mathur et al. . These investigators also reported in reconstituted preparations that pseudo-phosphorylation of S43 in particular promotes an inhibited state of the thin filament.
Interpretation of the mechanism of the effects of cTnI-(T144E) reported here are related to recent data, which have provided a model with new insights into potential mechanisms by which C-terminal regions of cTnI (131-210) control thin filaments . Reconstructions of electron micrographs of negatively stained filaments indicate that the C-terminal region of cTnI (131-210) (containing the inhibitory peptide and thus T144) binds to both actin and Tm in relaxing conditions thereby constraining Tm movements at low Ca2+. Thus, our data demonstrating loss of NEM-S1 induced activation in myofilaments controlled by cTnI-(T144E) at an acidic pH indicate that in addition to an altered interaction of cTnI with actin, its interaction with Tm may be altered under these conditions. Previous studies by Mathur et al.  reported that pseudo-phosphorylation of T144 in reconstituted preparations had little effect on thin filament activation or Ca2+ binding at pH 7.0. Our data generally agree with this finding, although we studied force generation in the myofilament lattice. Our data extend the findings of Mathur et al.  by the demonstration that the effects of cTnI-(T144E) phosphorylation to desensitize the myofilaments to Ca2+ are especially evident under acidic conditions.
Thus, together with our previous evidence that changes in the N-terminal region of cTnI  affected crossbridge dependent activation, we think that PKC phosphorylation of sites in the N-terminal region (S43, S45) influence the interaction between cTnT and Tm. In contrast, PKC phosphorylation of T144, which resides in the inhibitory peptide, influences the interaction between the C-terminal region of cTnI with Tm and enhances inhibition of crossbridge binding to the thin filament, resulting in differential mechanisms that would alter activation by Ca2+, but especially activation by strongly bound crossbridges.
The effect of PKC phosphorylation is associated with compromised cardiac function. Strongly bound rigor crossbridges an acidic pH are conditions associated with cardiac ischemia. Our findings are relevant with regard to heart failure in that acidic pH typically results in tissue damage and cell death due to ischemic contracture when rigor crossbridges form as levels of ATP are depleted . MacGowan et al.  also used transgenic mice with S43 and S45 mutated to Ala to study the effects of global ischemia and reperfusion. They reported that hearts of TG-TnI(S43A,S45A) were much more susceptible to ischemic contracture than controls and interestingly showed an elevation of diastolic pressure during early reperfusion. It has been suggested that this additional contracture during early reperfusion is due to replenishment of ATP which then immediately releases rigor crossbridges from the thin filament, but also allows them to cycle briefly resulting in increased contracture . In contrast, control animals exhibited less ischemic contracture due to the ability of PKC to phosphorylate these sites to diminish crossbridge dependent activation. Our data fit with these results and extend information with regard to the effect seen on crossbridge activation with phosphorylation of S43 and S45 in that myofilaments regulated by cTnI-(S43E/S45E) demonstrated significant inhibition of thin filament activation by strongly bound crossbridges at acidic pH (Figure 1B).
Our data are also relevant with regard to up-regulation of PKC expression found in failing hearts [1;24]. Belin et al.  studied skinned myocytes from rat hearts stressed by pressure overload-induced left ventricular hypertrophy (PO) and by myocardial infarction-elicited congestive heart failure (MI). In both cases there was a significant depression in maximum tension and Ca2+ sensitivity. They observed that there were four basal cTnI phosphorylation states in control myocytes, which increased significantly to seven phosphorylation states in myocytes from either PO or MI ventricular tissue indicating that increased phosphorylation induced depressed cardiac function leading to heart failure. Exchange of purified Tn taken from MI and PO hearts into myocytes from control hearts induced a depression of Ca2+ sensitivity, whereas purified Tn from control hearts exchanged into the diseased hearts improved Ca2+ sensitivity similar to levels seen in controls. Our data extend our understanding of the mechanism by which phosphorylations are likely to occur in heart failure by providing evidence for a role of diminished crossbridge dependent activation in response to PKC phosphorylation of cTnI under acidic conditions. Our data also extend our understanding of the role of PKC dependent phosphorylation in ischemia/reperfusion. Previous studies  have reported that a PKC induced depression of ATP consumption in ischemia provides cardio-protection. Our data indicate part of this protection is due to reduction in the ability of rigor crossbridges to recruit cycling crossbridges. Thus we propose that our data indicate a primary role of PKC dependent phosphorylation of cTnI in control of cardiac function with adaptive and cardio-protective effects during acute ischemia/reperfusion, but with maladaptive effects in the progression to heart failure.
In summary, we present novel information regarding a differential effect of PKC phosphorylation on separate regions of cTnI on the regulation of cardiac muscle function. This further understanding is important with regard to therapeutic interventions involving inhibition of kinases  and direct modification of the crossbridge thin filament reaction .
This research was supported by NIH T32 HL07692 (PLE, AH) and by NIH PO1 62426 (RJS) and RO1 HL 22231 (RJS). We thank Chad Warren for the preparation of NEM-S1 protein.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.