Differential Contribution of Troponin I Phosphorylation Sites to the Endothelin-modulated Contractile Response*

Cardiac troponin I is a phosphorylation target for endothelin-activated protein kinase C. Earlier work in cardiac myocytes expressing nonphosphorylatable slow skeletal troponin I provided evidence that protein kinase C-mediated cardiac troponin I phosphorylation accelerates relaxation. However, replacement with the slow skeletal isoform also alters the myofilament pH response and the Ca2+ transient, which could influence endothelin-mediated relaxation. Here, differences in the Ca2+ transient could not explain the divergent relaxation response to endothelin in myocytes expressing cardiac versus slow skeletal troponin I nor could activation of Na+/H+ exchange. Three separate clusters within cardiac troponin I are phosphorylated by protein kinase C, and we set out to determine the contribution of the Thr144 and Ser23/Ser24 clusters to the endothelin-mediated contractile response. Myocyte replacement with a cardiac troponin I containing a Thr144 substituted with the Pro residue found in slow skeletal troponin I resulted in prolonged relaxation in response to acute endothelin compared with control myocytes. Ser23/Ser24 also is a target for protein kinase C phosphorylation of purified cardiac troponin I, and although this cluster was not acutely phosphorylated in intact myocytes, significant phosphorylation developed within 1 h after adding endothelin. Replacement of Ser23/Ser24 with Ala indicated that this cluster contributes significantly to relaxation during more prolonged endothelin stimulation. Overall, results with these mutants provide evidence that Thr144 plays an important role in the acute acceleration of relaxation, whereas Ser23/Ser24 contributes to relaxation during more prolonged activation of protein kinase C by endothelin.

Protein kinase C (PKC) 2 activation is an important pathway involved in modulating cardiac contractile function (1)(2)(3)(4)(5). A number of end target proteins involved in excitation-contraction coupling are phosphorylated in response to PKC activation, including the L-type Ca 2ϩ channel (6,7) and myofilament proteins such as cardiac troponin I (cTnI) (8 -11), cardiac troponin T (10,12), and myosin light chain 2 (MLC 2 ) (8,13,14). The contribution of each end target to the contractile response to PKC is not well understood. Our laboratory is interested in understanding the role of cTnI in the PKC-mediated contractile function response. Increased PKC expression is observed in failing human hearts (15,16), and the relationship between PKC and cTnI phosphorylation may exert an important influence on cardiac function under physiological and pathophysiological conditions (17).
Earlier work demonstrated that cTnI phosphorylation is correlated with the positive inotropic response to ET (4,8,18). Accelerated relaxation also is observed in response to PKC activation by ET in myocytes expressing cTnI (4). PKC phosphorylation of purified cTnI decreases myofilament Ca 2ϩ sensitivity in reconstituted myofilaments (11), and this desensitization is expected to contribute to the accelerated relaxation. In contrast, peak shortening is diminished and relaxation is delayed rather than accelerated in response to ET when endogenous cTnI is replaced with nonphosphorylatable slow skeletal troponin I (ssTnI) in adult myocytes. These results are consistent with the idea that cTnI phosphorylation enhances peak myocyte shortening and accelerates relaxation rate in response to PKC activation by ET. However, ssTnI expression in adult myocytes also increases myofilament Ca 2ϩ sensitivity and slows the Ca 2ϩ transient (19,20). Thus, the differential response to ET observed in myocytes expressing cTnI versus ssTnI may result from the direct actions of cTnI phosphorylation on shortening and relaxation or from alterations in the Ca 2ϩ transient induced by ssTnI expression.
One goal of the present study is to establish whether PKC-mediated cTnI phosphorylation acts directly to accelerate relaxation or is explained by parallel alterations in cellular Ca 2ϩ . Calcium and shortening responses are measured simultaneously during the ET response for these experiments. Stimulation frequency also alters the Ca 2ϩ transient, with accelerated sarcoplasmic reticulum Ca 2ϩ uptake observed at higher pacing frequencies (21). The relative contribution of phosphorylated cTnI to the ET-induced relaxation response may decrease as pacing frequency increases in the event that alterations in cellular Ca 2ϩ are primarily responsible for this relaxation response. Thus, the contribution of cTnI phosphorylation to the ET-induced relaxation response is examined over a range of frequencies.
A second goal of our study was to differentiate between cTnI phosphorylation-induced and alkalosis-induced changes in the ET-mediated contractile performance response. Previous investigators showed that ET increases Na ϩ /H ϩ exchange via PKC (22,23). The ensuing alkalosis increases myofilament Ca 2ϩ sensitivity (24), and ET-induced alkalosis has been shown to contribute to the enhanced peak shortening and increased rate of contraction (22,23). The diminished peak-shortening response to ET observed in ssTnI-expressing myocytes (4) also could result from the enhanced basal myofilament Ca 2ϩ sensitivity and diminished pH sensitivity relative to cTnI-expressing myocytes (19). In the present study, the influence of ET-induced alkalosis on peak contraction was tested by inhibiting Na ϩ /H ϩ exchange in myocytes expressing cTnI and ssTnI.
The final aim of the present work was to determine the contribution of individual phosphorylation sites within cTnI to the PKC-mediated changes in intact adult myocyte relaxation rate in response to ET. Detailed work on purified cTnI indicates that PKC phosphorylates three separate clusters of residues, including Ser 23 /Ser 24 , Ser 43 /Ser 45 , and Thr 144 within cTnI (11,25). However, the myocardial response to phosphorylation of individual sites within the intact myocyte has yet to be thoroughly investigated. Studies here are focused on investigating the influence of the rat cTnI Ser 23 /Ser 24 and Thr 144 sites on contractile function in response to ET. For studies of cTnIThr 144 , the goal is to determine whether Thr 144 influences the peak contraction and relaxation response to PKC activated by ET, by substitution with the nonphosphorylatable Pro residue. The Ser 23 /Ser 24 cluster also is a target for PKC in purified cTnI and is a major myofilament target for protein kinase A (PKA) (26). Phosphorylation of this cluster by PKA significantly decreases myofilament Ca 2ϩ sensitivity (27) and accelerates relaxation during ␤-adrenergic activation (28). A similar increase in relaxation is expected if this Ser cluster is phosphorylated by PKC in intact adult myocytes. For these studies, the Ser 23 /Ser 24 residues were substituted with Ala to investigate whether this cluster influences the ability of cTnI to maintain relaxation in response to ET activation of PKC.
Generation of Adenoviral Vectors-Recombinant adenovirus vectors were constructed by cotransfection of pCA4cTnI cDNA with pJM17 by calcium phosphate into a HEK 293 cell line. Replication-deficient, recombinant adenovirus containing wild-type rat cTnI, wild-type ssTnI, or mutant ssTnI cDNAs, a cytomegalovirus promoter, and an SV40 polyadenylation signal were obtained following homologous recombination (19).
Myocyte Isolation-Adult rat ventricular myocytes were isolated as described earlier (4,30). Myocytes were plated on laminin-coated coverslips in Dulbecco's modified Eagle's medium supplemented with 50 units/ml penicillin, 50 g/ml streptomycin (pen/strep) and 5% serum for 2 h prior to replacing the medium with serum-free Dulbecco's modified Eagle's medium containing recombinant adenovirus vector (multiplicity of infection ϭ 500). Gene transfer of exogenous TnI at a multiplicity of infection of ϳ500 results in virtually all cells expressing and incorporating exogenous TnI into the myofilament (29). After 1 h, Dulbecco's modified Eagle's medium with pen/strep (2 ml) was added, with regular replacement of media over 4 days in culture.
Analysis of Protein Expression-Control and virus-treated adult rat cardiac myocytes were scraped from coverslips into sample buffer 4 days after gene transfer (31). In some experiments, myocytes were permeabilized in 0.1% Triton X-100 for 1 min to evaluate incorporation of TnI protein into the myofilament after gene transfer. In earlier work, expression in intact and permeabilized myocytes was comparable when myofilament proteins were incorporated into the myofilament (29,32). Proteins were then separated by gel electrophoresis and blotted onto polyvinylidene difluoride membrane for 2000 V-h with immunodetection carried out as described earlier (30) using a 1:1000 dilution of MAB 1691 (Chemicon Inc., Temecula, CA), a monoclonal antibody that recognizes all TnI isoforms. A phospho-Ser 23 /Ser 24 -specific polyclonal antibody (1:1000; Cell Signaling Technology, Beverly, MA) and a paired horseradish peroxidase-conjugated secondary antibody were used to detect Ser 23 /Ser 24 phosphorylation. Myocytes collected for detection of phosphorylation were incubated at 37°C with calyculin A (10 nM) alone or with the selected PKA and PKC agonist and/or inhibitor for time intervals ranging from 2.5 min up to 1 h.

Measurement of Sarcomere Shortening in Single Cardiac
Myocytes-Myocytes to be used for shortening assays were transferred to a Plexiglas stimulation chamber consisting of 8 wells containing platinum electrodes. Myocytes were stimulated electrically (2.5-ms pulse, 0.2 Hz) in medium 199 supplemented with pen/strep (M199), 10 mM HEPES, 0.2 mg/ml bovine serum albumin, and 10 mM glutathione (M199ϩ), and the media was replaced every 12 h for days 2-4 post-gene transfer (33). The voltage was set so that ϳ20% of the myocytes were stimulated on each coverslip. Four days after gene transfer, individual coverslips were transferred to a temperature-controlled (37°C) chamber mounted on a Nikon microscope stage. A video-based detection system (Ionoptix, Milton, MA) was used to detect sarcomere shortening in intact myocytes. For frequency-response studies, myocytes were stimulated at each frequency (0.2, 0.5, 0.75, 1, and 2 Hz) until a steady state contraction was reached. Sarcomere shortening also was recorded at 0.2 Hz before and after the increases in stimulation frequency to ensure there were no changes in the rate of contraction, peak contraction, and rate of relaxation after testing at all frequencies. Experiments were recorded using SarcLen software, and an average of 10 twitches/myocyte was collected under basal conditions. Signal-averaged data were analyzed to determine resting sarcomere length (SL R ), percent change in peak sarcomere shortening above base line (S b ), time to peak shortening (TTP), time to 25, 50, and 75% relaxation (TTR 25 , TTR 50 , and TTR 75 , respectively), and maximum normalized shortening and relaxation velocities (ϩdl/dt max , Ϫdl/dt max , respectively). Media containing ET (10 nM), 5-N(ethyl-N-iospropyl)amiloride (EIPA, 500 nM) or both reagents were perfused via separate individual perfusion tubing to eliminate introduction of residual agonists/inhibitors. Experiments for Ca 2ϩ transient and sarcomere shortening measurements were carried out on myocytes loaded with Fura-2 AM, as described previously (34). Briefly, myocytes were loaded for 5 min with 5 M Fura-2 AM (final Me 2 SO ϭ 0.6%) at 37°C followed by a 5-min wash-out for de-esterification. Myocytes were then stimulated at 0.2 Hz at 37°C, and the base-line sarcomere length and Ca 2ϩ transient were recorded from a single cell. Recordings were then made 5, 10, and 15 min after the addition of ET (10 nM), with the cell maintained in the dark between sampling times. Optimal cTnI phosphorylation was observed in response to 10 -100 nM ET in an earlier study (4). Ten twitches/ myocyte were collected for each sample. In addition to sarcomere shortening measurements, the resting and peak ratios, maximum Ca 2ϩ transient rate (ϩdCa/dt max ), maximum Ca 2ϩ decay rate (ϪdCa/dt max ), and time to 50 and 75% decay (TTD 50 , TTD 75 , respectively) in the Ca 2ϩ transient were measured. The Ca 2ϩ transient and shortening measurements were maintained over this same time interval in myocytes maintained under basal conditions.
Statistical Analysis-Values are expressed as mean Ϯ S.E. Grouped comparisons are analyzed using a Student's t test or analysis of variance testing (ANOVA). Statistical differences detected by ANOVA are compared with a Newman-Keuls multiple comparison test with p Ͻ 0.05 considered significantly different.

Relationship between Ca 2ϩ Transient and Contractile Response to
Endothelin-Sarcomere shortening and Ca 2ϩ transients were measured before and after ET in Fura-2-loaded myocytes expressing cTnI or ssTnI 4 days after gene transfer (Fig. 1, TABLE ONE). The increased peak shortening and enhanced relaxation rate observed in response to 10 nM ET in earlier work on myocytes expressing cTnI (4), was similarly observed in myocytes loaded with Fura-2 ( Fig. 1A). In myocytes expressing nonphosphorylatable ssTnI, the delayed relaxation and loss of the enhanced peak shortening observed in response to ET (4) similarly developed in Fura-2-loaded cells (Fig. 1B, TABLE ONE).
Ca 2ϩ transient measurements also were examined in these myocytes. The time to 50% decay for basal Ca 2ϩ transients reportedly slows in transgenic mouse myocytes expressing ssTnI (20), and similar results developed in rat myocytes expressing ssTnI (TABLE TWO). Peak Ca 2ϩ and the TTP remained unchanged in the ssTnI-expressing myocytes compared with values observed in controls at 0.2 Hz. The Ca 2ϩ transient change observed in response to ET was then analyzed in myocytes expressing cTnI and ssTnI (TABLE ONE). The percent change in peak Ca 2ϩ amplitude and maximum rate of Ca 2ϩ decay (ϪdCa/dt max ) in response to ET was not significantly changed in myocytes expressing either cTnI or ssTnI (Fig. 1, A and B; TABLE ONE). However, significant delays were observed in the time to 75% Ca 2ϩ decay in both groups of myocytes. The similar Ca 2ϩ transient responses to ET coupled with the divergent shortening response of cTnIversus ssTnI-expressing myocytes indicates that cTnI contributes significantly to the enhanced peak shortening and accelerated relaxation rate observed in response to PKC activation by ET compared with myocytes expressing ssTnI.
Additional studies were performed to investigate whether the contribution of phosphorylated TnI to the ET-mediated change in contractile function is diminished with higher frequencies. The shortening Ϫ frequency relationship described previously in rodents (33,35) was similar under basal conditions and in response to ET over a range of frequencies (0.2-2 Hz), (TABLE THREE) in control myocytes. The percent decrease in peak shortening, ϩdl/dt max and Ϫdl/dt max in response to increased frequency was similar before and after ET. The consistency of this response provides evidence that the contribution of cTnI phosphoryla-tion to contractile function is not altered by frequency-induced changes in the Ca 2ϩ cycle.
Contribution of Cellular Alkalosis to the Myofilament Response to Endothelin-An important component of the contractile response to ET depends on activation of Na ϩ /H ϩ exchange and the generation of a cellular alkalosis (22,23). Experiments were carried out in the presence of the amiloride analog EIPA, an inhibitor of Na ϩ /H ϩ exchange (23,36), to identify responses resulting from alkalosis. The myocyte response to EIPA is comparable in nontreated myocytes and myocytes expressing cTnI after gene transfer (results not shown); the pooled results are shown in Fig. 2. Myocyte shortening amplitude in myocytes expressing Comparison of the percent change in sarcomere shortening and Ca 2؉ transient measurements in response to 10-min exposure to ET (10 nM) for myocytes expressing cTnI, ssTnI, or cTnIT144P Percent change in resting sarcomere length (SL R ), maximum normalized departure velocity (dl/dt max ), peak shortening normalized for resting length (S b ), maximum normalized relaxation velocity (Ϫdl/dt max ), and time to 50 and 75% relaxation from peak (TTR 50 , TTR 75 ) are shown for the shortening measurements. Ca 2ϩ measurements shown include percent changes in resting Ca 2ϩ fluorescence, maximum normalized Ca 2ϩ rise (dCa/dt max ), peak Ca 2ϩ normalized for resting Ca 2ϩ (peak Ca 2ϩ ), maximum normalized Ca 2ϩ decay (ϪdCa/dt max ), and time to 50 and 75% decay from peak (TTD 50 , TTD 75 ). Results are expressed as mean Ϯ S.E. Statistical comparisons were made between the control and ssTnI responses and were performed using ANOVA (p Ͻ 0.05) and post-hoc Newman-Keuls tests (see Footnotes a and b).   cTnI decreased slightly after the addition of EIPA but did not significantly influence peak amplitude in myocytes expressing ssTnI (Fig. 2, A  and B). Resting sarcomere length (SL R ), ϩdl/dt max , and Ϫdl/dt max (ϩdl/dt and Ϫdl/dt in Fig. 2, respectively) were not significantly altered by EIPA in myocytes expressing cTnI or ssTnI (Fig. 2B). However, EIPA blocked a major portion (Ͼ70%) of the ET-induced increase in peak shortening amplitude observed in cTnI-expressing myocytes. In contrast, EIPA did not significantly modify the accelerated relaxation response to ET in myocytes expressing cTnI, and the delayed relaxation observed in response to ET was comparable with and without EIPA in myocytes expressing ssTnI ( Figs. 1 and 2). These results support the idea that Na ϩ /H ϩ exchange and the ensuing alkalosis is important for the enhanced peak shortening response, whereas cTnI plays a more critical role in the accelerated relaxation rate response to ET.

Role of cTnI Phosphorylation Residues in the Endothelin Response-
To determine the role of the cTnIThr 144 phosphorylation site in the relaxation response, the ssTnI Pro residue was substituted for Thr 144 to form cTnIT144P, and replacement was measured 4 days after gene transfer. The stoichiometric replacement of cTnI with cTnIT144P-FLAG was 46 Ϯ 2% (n ϭ 13) and comparable with the 53 Ϯ 2% (n ϭ 14) replacement with cTnI-FLAG 4 days after gene transfer (Fig. 3). Replacement of nontagged cTnI and cTnIT144P is expected to be comparable in myocytes over the same time interval (29,37). Total TnI expression remained constant in myocytes expressing cTnI and cTnIT144P based on the ratio of TnI intensity to the intensity of a silver-stained band on the gel (cTnI set at 1.0; cTnIT144P ϭ 1.13 Ϯ 0.10, n ϭ 11), and isoform expression of tropomyosin and troponin T was comparable in myocytes expressing cTnI and cTnIT144P (results not shown). Expression of total TnI also was comparable in intact (I) and permeabilized (P) myocytes (Fig. 3, I/P ratio for cTnI ϭ 1.10 Ϯ 0.08, n ϭ 5; cTnIT144P ϭ 1.05 Ϯ 0.06, n ϭ 5), which is consistent with myofilament incorporation of the delivered TnI.
Measures of basal contractile function in myocytes expressing cTnIT144P are not significantly different from values obtained from myocytes expressing cTnI (TABLE TWO, Fig. 4A). These results are in agreement with the earlier reverse substitution of Pro 110 with Thr in fast skeletal TnI (38). Base-line Ca 2ϩ transients in myocytes expressing cTnIT144P (Fig. 4B) also showed no significant differences from control values. Peak shortening increased significantly in response to ET in cTnIT144P-expressing myocytes (Fig. 4) and was comparable with the ET-induced enhanced peak shortening in control myocytes (Fig. 1B). However, several indices of relaxation, including Ϫdl/dt max and TTR 75 , were significantly delayed in response to ET in myocytes expressing cTnIT144P compared with controls (  Fig. 4). As with control myocytes, a frequency change from 0.2 to 2 Hz produced similar changes before and after ET in myocytes expressing cTnIT144P (results not shown). The major difference between myocytes expressing cTnIT144P versus cTnI was the decrease in relaxation rate in response to ET; this finding indicates a specific role for Thr 144 in accelerating relaxation in response to ET.
The contribution of Ser 23 /Ser 24 phosphorylation to the ET-induced change in contractile function also was determined. A phospho-specific Ser 23 /Ser 24 antibody was used for these studies. Myocytes were incubated with the PKA agonist dobutamine (DOB) for 2 min to verify detection of the well documented PKA-mediated cTnISer 23 /Ser 24 phos-

Frequency response (0.2 and 2 Hz) of electrically stimulated adult rat cardiac myocytes observed before and after ET (10 nM) in control myocytes 4 days after myocyte isolation
Results are expressed as mean Ϯ S.E. of the percent change in sarcomere shortening measurements made between 0.2 and 2.0 Hz under basal conditions and in response to 10 -15-min exposure to ET. There were no significant differences between the frequency response under basal conditions and in response to ET using a Student's t test with p Ͻ 0.05 considered significant.  . Results from myocytes expressing cTnI with or without adenovirus-mediated cTnI gene transfer were not significantly different from each other (results not shown) and were therefore pooled for these tracings. B, percent change in shortening measurements from basal to EIPA and from EIPA to EIPA ϩ ET. Results are shown for resting sarcomere length (SL R ), peak shortening amplitude (Peak), maximum shortening rate (ϩdl/dt), and maximum relaxation rate (Ϫdl/dt) in myocytes expressing cTnI (n ϭ 38) or ssTnI (n ϭ 16) after EIPA (upper panel) and EIPA ϩ ET (lower panel). Myocyte shortening was analyzed 4 days after gene transfer, and measurements were carried out 15 min after the addition of EIPA and 10 min after the addition of EIPA ϩ ET. Although SL R , ϩdl/dt, and Ϫdl/dt were not significantly changed by EIPA, the percent change in peak shortening decreased significantly in myocytes expressing cTnI (p Ͻ 0.05 versus base-line level by ANOVA and post-hoc Newman-Keuls tests). There were no significant differences in the response to EIPA when comparing myocytes expressing cTnI versus ssTnI. SL R , peak, and ϩdl/dt were not significantly changed by EIPA ϩ ET compared with EIPA alone. However, Ϫdl/dt increased significantly in myocytes expressing cTnI (*, p Ͻ 0.05), and this increased rate of relaxation was significantly higher than the change in relaxation observed in myocytes expressing ssTnI (ϩ, p Ͻ 0.05 versus control).

Measurements
phorylation site (39,40) (Fig. 5A). Detection of cTnI phosphorylation by this antibody was specific, as it was blocked when, in addition to DOB, the ␤-adrenergic inhibitor propranolol was added to myocytes (results not shown). There was no significant phosphorylation detected in myocytes incubated with ET for 10 min with (Fig. 5A) or without the PKC inhibitor bis-indolylmaleimide-1 (bis-1, results not shown). However, significant phosphorylation of Ser 23 /Ser 24 developed in control myo-cytes over a 1-h time span, and this ET-mediated phosphorylation was significantly reduced in the presence of the PKC inhibitor bis-1 but not by the ␤-adrenergic antagonist propranolol (Fig. 5A). Treatment of myocytes with other PKC agonists, including phenylephrine and phorbol 12-myristate 13-acetate for 1 h also resulted in significant cTnI phosphorylation of Ser 23 /Ser 24 . Incubation with fibroblast growth factor-2 was previously shown to activate PKC (41), yet this agonist did not significantly influence cTnI phosphorylation over a 1-h time span (results not shown). These results provide evidence for temporal phosphorylation of cTnI Ser 23 /Ser 24 in response to PKC activation by several PKC agonists.
To directly determine the contribution of Ser 23 /Ser 24 to the contractile function response to ET, cTnIS23A/S24A was expressed in adult myocytes. By 4 days after gene transfer, 44 Ϯ 5% of cTnI was replaced with the cTnIS23A/S24A (n ϭ 15, Fig. 5B). DOB phosphorylation of cTnI was significantly reduced in myocytes expressing cTnIS23A/S24A or cTnIS23A/S24A-FLAG (results not shown). In addition, Ser 23 /Ser 24 phosphorylation was significantly attenuated after 1 h of ET in myocytes expressing tagged and nontagged cTnIS23A/S24A (Fig. 5C). This S23A/ S24A substitution has previously been shown to have little influence on basal myofilament function (42), and no significant differences were observed in SL R , peak amplitude, ϩdl/dt max , or Ϫdl/dt max in myocytes expressing cTnIS23A/S24A compared with controls (results not shown). Control myocytes expressing cTnI or the substituted cTnIS23A/S24A responded similarly to ET over the first 15 min (Fig. 6). However, a significant difference in the response was observed 1 h after the addition of ET in myocytes expressing cTnIS23A/S24A compared   with control values (Fig. 6). Specifically, peak shortening amplitude was not increased and both Ϫdl/dt max and TTR 75 indicated relaxation was significantly prolonged in myocytes expressing cTnIS23A/S24A compared with control myocytes and were comparable with the results obtained in myocytes expressing ssTnI (Fig. 6). These results demonstrate Ser 23 /Ser 24 phosphorylation contributes to the accelerated relaxation during more prolonged myocyte PKC activation by ET.

DISCUSSION
Results from the present study demonstrate the important direct role for cTnI phosphorylation in accelerating relaxation in response to PKC activated by ET in adult myocytes (Figs. 1-3). Our results now demonstrate a clear role for cTnI Thr 144 and Ser 23 /Ser 24 in contributing to the accelerated relaxation in response to ET (Figs. 4 and 6). Substitution with a Pro residue at Thr 144 results in the delayed relaxation previously observed in myocytes expressing nonphosphorylatable ssTnI and provides evidence that the Thr 144 site contributes to accelerated relaxation during the ET response. The Ser 23 /Ser 24 cluster does not appear to contribute to the initial response to ET but, instead, is important for the accelerated relaxation during more prolonged stimulation by ET. This finding indicates that the relaxation rate response to ET-induced PKC pathway activation depends on temporal phosphorylation of individual clusters within cTnI.
TnI Phosphorylation Plays a Direct Role in the Relaxation Response to Endothelin-Results from the present study demonstrate the important role of cTnI phosphorylation in the relaxation response during the positive inotropic response to ET, independent of cellular Ca 2ϩ handling. The only aspect of Ca 2ϩ transient that changed significantly in response to 10 nM ET was the percent change in TTD 75 , which decreased in both control and ssTnI-expressing myocytes (Fig. 1, TABLE ONE). All other measurements of the Ca 2ϩ transient remained unchanged during the ET response, and work by other laboratories supports this finding (43,44). There are reports that ET activation of PKC enhances the Ca 2ϩ transient at higher doses of ET (45), which may be due to increased Ca 2ϩ entry via the L-type Ca 2ϩ channel (46). However, results from Ebihara et al. (47) indicate the increased Ca 2ϩ transient is not necessary for the enhanced contractile response to ET.
Experiments with the NHE1 inhibitor EIPA also demonstrate that inhibition of NHE1 prevents the ET-mediated enhancement in peak shortening amplitude (Fig. 2). ET in the concentration range used here activates PKC-induced NHE1 (48) and causes a cellular alkalosis (22,23). The alkalosis-mediated increase in myofilament Ca 2ϩ sensitivity (24) is responsible for the increased peak contractile shortening (22,23). This enhancement in peak shortening is absent from myocytes expressing ssTnI (Fig. 1), which is likely due to the combination of increased basal myofilament Ca 2ϩ sensitivity (19,20) and a decreased pH response to alkalosis in these myocytes compared with cTnI-expressing myocytes (19).
In contrast to peak shortening, the divergent influence of cTnI versus ssTnI on the relaxation response to ET was not influenced significantly by EIPA (Fig. 2). Although the slowing of relaxation in myocytes expressing ssTnI was attenuated by the presence of EIPA during the ET response, relaxation was significantly accelerated in myocytes expressing cTnI versus ssTnI during this response. Our results now provide evidence that phosphorylated cTnI maintains relaxation regardless of the alkalosis. The delayed relaxation response to ET observed in myocytes expressing cTnIT144P also supports this finding (Fig. 4).
What Role Does cTnI Phosphorylation Play in the Contractile Response to ET?-The role of cTnI phosphorylation in the contractile function response to PKC agonists remains controversial. Earlier in vitro work showed that PKC phosphorylation of cTnI decreases myofilament Ca 2ϩ sensitivity as well as maximum actomyosin ATPase activity (11,25). PKA phosphorylation of cTnI also decreases myofilament Ca 2ϩ sensitivity and results in accelerated myocyte relaxation (27,28), and substitution of Ser 23 /Ser 24 with Asp to mimic phosphorylation accelerates relaxation in transgenic mice (39). However, PKC phosphorylates Ser 43 /Ser 45 and Thr 144 in addition to the Ser 23 /Ser 24 cluster (11). More recent studies in which the Ser 23 /Ser 24 , Ser 43 /Ser 45 , and Thr 144 clusters are all replaced with Ala in knock-in mice (49,50) or substituted with Asp to mimic phosphorylation in transgenic mice (37) suggest that cTnI works primarily to delay cardiac relaxation during PKC activation. These results are in contrast to results presented here demonstrating an accelerated relaxation response attributable to cTnI phosphorylation (TABLE ONE), and the in vitro work demonstrating decreased myofilament Ca 2ϩ sensitivity (11). Differences in experimental conditions, such as load and/or temperature in these various models, do not appear to account for the divergent functional response. We propose that these seemingly divergent responses may be due to a hierarchy of cTnI phosphorylation in response to a given agonist, with different cTnI clusters phosphorylated in response to PKC activation. Ser 43 /Ser 45 phosphoryl- Results are expressed as percent change in sarcomere shortening measurements. Measurements are shown for percent change in base-line sarcomere length (baseline SL), peak shortening, Ϫdl/dt max , and time to relaxation 75% (TTR 75 ) and are compared in myocytes expressing cTnI (control; n ϭ 24), ssTnI (n ϭ 15), and S23A/S24A (n ϭ 16) 4 days after gene transfer. No significant differences between myocytes expressing cTnI and cTnIS23A/S24A were observed in response to 15 min ET, but maintenance of relaxation at 60 min was significantly decreased in myocytes expressing ssTnI and cTnIS23A/S24A compared with controls. *, p Ͻ 0.05 versus control by ANOVA and post-hoc Newman-Keuls tests.
ation may have a dominant influence on contractile relaxation via its influence on maximum ATPase activity when all five phosphorylation sites are replaced by Ala (49) or replaced with a negatively charged Asp (37). However, PKC activation by a specific agonist may differentially and/or temporally phosphorylate one or more of the cTnI residues rather than phosphorylate all three clusters. In support of this idea, in vitro motility assays carried out with cTnISer 43 /Ser 45 containing charged Glu substitutions demonstrated a decrease in myofilament Ca 2ϩ sensitivity and decreased sliding speed, whereas cTnIT144G only decreased myofilament Ca 2ϩ sensitivity (51). NMR analysis of Ser 43 / Ser 45 substituted with Asp provides evidence that phosphorylation of this site may be dominant in that it influences the Ca 2ϩ binding affinity to TnC as well as protein-protein interactions between TnI and TnC (52). Thus, the divergent results are predicted to stem from differential phosphorylation of the three cTnI phosphorylation clusters and/or the relative dominance of one cluster in response to activation and translocation of specific PKC isoforms by a given agonist. This idea will be important for rigorous testing, with a variety of agonists under physiological and pathophysiological conditions, to better understand the PKC signaling pathway and its ability to modulate both cTnI phosphorylation and contractile performance.
Role of Individual cTnI-phosphorylated Residues in the Contractile Response to ET-Detailed studies on purified cTnI support the idea that the residue phosphorylated by PKC is important for the resulting functional relaxation response. Phosphorylation of Thr 144 in cTnI produces a rightward shift in myofilament Ca 2ϩ sensitivity for both actomyosin ATPase activity (11) and force generation in permeabilized trabeculae (51,53). The anticipated cellular response is an increase in myocyte relaxation rate. Our results now show that substitution of Thr 144 with Pro significantly delays relaxation during the ET response (Fig. 4, TABLE THREE). The Thr 144 site is particularly important because the inhibitory peptide region encompassing this phosphorylation target (amino acids 129 -149) is important for determining TnI binding to actin in the absence of Ca 2ϩ and diminished binding to actin in the presence of Ca 2ϩ (54). Thus, the inhibitory peptide region is important for the molecular switch activity of TnI (55,56). The delayed relaxation response to ET in myocytes expressing cTnIT144P compared with control myocytes is consistent with the idea that phosphorylation of this residue in response to ET plays a significant role in the accompanying relaxation rate response.
Experiments with the phospho-cTnISer 23 /Ser 24 antibody and with cTnIS23A/S24A also demonstrate a key role for this cluster during more extended activation of PKC by ET. Phosphorylation of Ser 23 /Ser 24 decreases myofilament Ca 2ϩ sensitivity and accelerates relaxation in response to PKA activation (27,28), and similar shifts in myofilament Ca 2ϩ sensitivity are reported in response to PKC phosphorylation of Ser 23 /Ser 24 (11). Our results now demonstrate a temporal pattern of Ser 23 /Ser 24 phosphorylation that is important for enhanced relaxation in response to longer intervals of ET in intact myocytes (Fig. 6). In biochemical experiments with purified cTnI, longer treatment with PKC also was required for Ser 23 /Ser 24 phosphorylation (11), yet it remained unclear whether this result indicated that there was temporal phosphorylation in the intact cell or instead indicated that Ser 23 /Ser 24 was not a typical in vivo target for activated PKC. Results with myocytes expressing cTnIS23A/S24A (Fig. 6) clearly demonstrate that there is temporal phosphorylation of this site by PKC activated in response to ET. Several PKC agonists, including ET, are released in a paracrine/ autocrine pattern (57,58), and phosphorylation of Ser 23 /Ser 24 during more prolonged intervals of PKC stimulation would then be important for preventing delays in cellular relaxation. Taken together, the results obtained in experiments to investigate Thr 144 and Ser 23 /Ser 24 provide evidence that both clusters are important in maintaining relaxation during PKC activation. The relative importance, as well as the interactive effects, of these sites and the Ser 43 /Ser 45 site in response to PKC agonists, which appears primarily to decrease peak contraction and slow relaxation (51), will be important to investigate in the future for a complete understanding of the role played by cTnI phosphorylation in modulating cardiac function.