INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

PROTEIN KINASE C (PKC) may have multiple and varied effects on myocyte function through phosphorylation of several intracellular sites, including L-type calcium channels (23), ATP-sensitive K⁺ (K_ATP) channels (14), and several myofilament proteins, including troponin I (TnI) (19, 29). Noland et al. (19) have shown in reconstituted myofilaments that PKC may mediate a reduction in the maximal calcium-stimulated ATPase rate of reconstituted actomyosin S-1, and apparent affinity of myosin S-1 for the thin filament. These effects are largely mediated through phosphorylation of serines-43 and -45 of TnI, and replacement of these with alanine results in near-complete preservation in ATPase activity in response to PKC. Also, Takeishi et al. (27) have demonstrated that in vivo phosphorylation of TnI by PKC decreases myofilament responsiveness to calcium and contractility.

In hearts, in which the PKC sites are phosphorylated, it would be expected that the force generated (developed pressure) for a given level of intracellular calcium is reduced. Alteration in the force-calcium relationship may have several functional consequences. During ischemia there is a gradual increase in intracellular calcium and reduction in ATP, with ultimate development of ischemic contracture (25). PKC is activated during ischemia (1), and its effects on the force-calcium relationship mediated through TnI may protect against the development of contracture during states of high intracellular calcium. Another consequence of alteration in the force-calcium relationship may be an increase in energy utilization. When these PKC sites are phosphorylated, the maintenance of high calcium levels with reduced force would be expected to adversely effect the relationship between developed pressure and oxygen consumption, because cycling of calcium may use a significant proportion of ATP consumption and thus oxygen consumption (10).

The present study examined the functional and energetic effects of PKC phosphorylation of serine residues 43 and 45 by generating a transgenic mouse expressing mutant TnI, in which these phosphorylation sites are replaced with alanine.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Generation of transgenic mice. All of the studies were performed in accordance with institutional guidelines. cDNA for mutant TnI, in which the PKC phosphorylation sites (serines-43 and -45) were replaced with alanine (19), was placed under the control of the cardiac-specific -myosin heavy chain promoter (26). To generate transgenic mice, the transgene was microinjected into FVB zygotes with the use of standard methods (12). Microinjected embryos were reimplanted into pseudopregnant FVB female mice. By using the polymerase chain reaction (PCR) with the following primer sequences: CCG AGA TTT CTC CAT CCC AAG 5' and GCA TCG CTG CTT TCA TCA GCC 3', we detected the presence of mutant cardiac TnI transgene, making use of the heterologous junction between the myosin heavy chain promoter and TnI cDNA.

RNA isolation and RT-PCR. Total RNA was isolated from mouse hearts using standard procedures, then digested with RQ1 RNase-free DNase (Promega) to remove any contaminating DNA. Spectrophotometric quantitation of purified RNA was confirmed visually by agarose-formaldehyde electrophoresis of RNA. RT-PCR was performed with the use of an Access System (Promega) with primer pairs designed to the cDNA sequences of mus musculus CD-1 cardiac TnI (Genbank Accession, U09181) and mouse G3PDH (Clontech). The CD-1 primers correspond to the 3' end (+627 to +607: AAACTTTTTCTTGCGGCCTTC) and the wild-type (AAGTCTAAGATCTCCGCCT) or mutant (AAGTCTAAGATCGCCGCCG) +115 to +133 region. We used 100 ng of purified total RNA for the reactions. The primer sets produced predicted product sizes of 512 bp for CD-1 and 452 bp for G3PDH. A linear range of amplification profile for each primer set was established on the basis of the following cycling parameters: reverse transcription step (1 cycle of 48°C for 45 min and 94°C for 2 min), PCR step (up to 40 cycles of 94°C for 30 s, 60°C for 60 s, and 68°C for 120 s), and extension step (1 cycle of 68°C for 7 min; 4°C soak). A midpoint of the linear amplification range was determined for CD-1 and G3PDH RT-PCR products to be at cycles 30 and 24, respectively. Negative controls included tubes that lacked RNA, avian myeloblastosis virus (AMV) reverse transcriptase, or Tfl DNA polymerase. The identity of each PCR product was confirmed by ligating the DNA fragments into the PCRII vector (Invitrogen) and sequencing these by using standard molecular biology techniques. Equal volumes of amplified products were loaded in each lane of a 1% Tris, acetic acid, and EDTA (TAE)-buffered agarose gel and electrophoresed with the 100-bp ladder (Life Technologies). The results are not strictly semiquantitative; however, care was taken to assure that starting total RNA amounts were the same for each sample and the cycling parameters were adjusted for each primer pair to ensure PCR was stopped during the linear segment of the amplification profile. All samples contained an equal amount of G3PDH amplification product (not shown).

Immunoblot analyses. Hearts were snap-frozen in liquid nitrogen and stored at 80°C. To determine total cardiac TnI levels, proteins were isolated from the frozen hearts by Dounce homogenization in a small volume of ice-cold lysis solution, consisting of 150 mM NaCl, 20 mM Tris · HCl pH 7.6, 1 mM EDTA, 0.5% sodium deoxycholate, 70 mM NaF, 1% Nonidet P-40, a protease inhibitor cocktail (Complete; Boehringer Mannheim), 200 µM sodium orthovanadate, and 2 µM phenylmethylsulfonyl fluoride (PMSF). Debris was pelleted and the supernatants were stored at 80°C. Protein concentrations were established using the Pierce BCA protein determination kit. SDS-PAGE and transfer to nitrocellulose were performed by using standard procedures, and identical gels were stained with Coomassie brilliant blue R250 to confirm equal protein loading. Nitrocellulose membranes were rinsed in Tris-buffered saline pH 7.4 containing 0.1% Tween 20 (TBS-T) and blocked in 5% nonfat milk; TBS-T for 1 h at 22°C on a rocking platform. Membranes were rinsed four times in TBS-T then incubated overnight at 4°C on an orbital shaker with primary antibodies diluted in TBS-T. The anti-human cardiac-specific TnI monoclonal antibodies (either 3314 11E1.3 or 3350 2F6.6, a gift from Dr. Jack Ladenson, Washington University, St. Louis, MO) were used at a final concentration of 1 µg/ml and found to react equally well. The primary antibody was detected with horseradish peroxidase-conjugated anti-mouse secondary antibody and the ECL kit (Amersham) according to the manufacturer's directions before exposure to X-Omat AR film (Kodak).

Experimental design and isolated perfused mouse heart studies. Because calcium-induced inotropy is partly mediated by PKC (15, 17), high calcium was used as an activator of protein kinase C. In other studies, global ischemia, which is also associated with high calcium (25) and PKC activation (1), was used.

Ischemia experiments. In another group of experiments, hearts were perfused as described above for a baseline period of 10 min and then underwent 15 min of global no-flow ischemia, followed by 20 min of reperfusion. To determine that PKC- translocated during ischemia, hearts were exposed to the same ischemic conditions and time as above and then flash frozen in liquid nitrogen and compared with control hearts not exposed to ischemia. Left ventricular tissue was homogenized in ice-cold buffer containing 20 mM Tris · HCl, 2 mM EDTA, 0.5 mM EGTA, 1 mM PMSF, 25 µg/ml leupeptin, 0.3 mM sucrose at pH 7.4 and centrifuged at 1,000 g for 5 min to remove tissue debris. Cytosolic and particulate fractions were separated by ultracentrifugation of the supernatant (100,000 g for 60 min). The resulting supernatant (cytosolic fraction) was collected, and the pellet (particulate fraction) was resuspended in sucrose-free buffer containing 1% Triton X, incubated for 60 min at 4°C, and centrifuged at 40,000 g for 15 min. SDS-PAGE and transfer to nitrocellulose was performed as described above. The membranes were incubated with a rabbit polyclonal antibody raised against murine PKC- (Upstate) at 1 µg/ml and detected with a horseradish peroxidase-conjugated anti-rabbit secondary antibody and the ECL kit (Amersham). After exposure to film, quantitative protein analysis was performed by laser densitometry.

Measurement of intracellular calcium in whole perfused mouse hearts with rhod 2. The methods used to measure intracellular calcium with rhod 2 (Molecular Probes, Eugene, OR) in perfused hearts have been previously extensively described (7, 8). Excitation at 524 nm and emission at 589 nm was used for fluorescence measurements. We quantified the relative amount of rhod 2 in the heart with the use of absorbance measurements, by taking the ratio of absorbance at 524 nm (rhod-2 sensitive) to 589 nm (rhod-2 insensitive). This procedure eliminated the effect of motion and scattering changes because both are equally affected by motion and concentration, though only 524 reflects the concentration of rhod 2 (7, 8). Rhod 2 (100 µg) was dissolved with 4 µl of DMSO and 200 µl of H₂O and loaded through the coronary perfusate. At the end of the perfusion protocol, maximal fluorescence, used in the calculation of intracellular calcium concentration, was determined by tetanizing the heart with a 20 mM bolus of calcium chloride and 10 µM of cyclopiazonic acid (Sigma Chemical), which is a potent inhibitor of Ca²⁺-ATPase (3).

Statistical analysis. Unpaired comparisons were performed using Student's t-test, and when multiple comparisons were made, we used ANOVA with the Scheffé's test for subgroup analysis. Data are expressed as means ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Description of mice. All studies were performed in isolated perfused mouse hearts from transgenic mice and FVB age-matched controls. Mice were aged between 8 and 40 wk. All experiments were done in males, except for three females in the wild-type group, and one female in the transgenic group. Founders 2, 7, and 9 were identified as the highest expressers by mRNA levels. In founders 2 and 7 amounts of wild-type and mutant mRNA were approximately equal, and in founder 9 the amount of mutant mRNA was less than wild-type mRNA (Fig. 1). In preliminary studies no functional differences were noted between these founders in terms of developed pressures, heart-to-body weight ratios, ischemic contracture pressure, myocardial oxygen consumption, or intracellular calcium (data not shown). All of the founders were used in the ischemia experiments, and founder 7 was used for the myocardial oxygen consumption and founder 9 for the intracellular calcium experiments. TnI antibodies did not selectively identify the mutant protein from wild-type mice, though total immunoreactive TnI was similar in wild-type and transgenic mice (Fig. 2), suggesting that the total amount of TnI was closely regulated, and that any excess protein was degraded. Mice expressing mutant TnI appeared normal, and there was no significant difference in heart-to-body weight ratios (Table 1). Also, systolic and diastolic pressures were not significantly different between transgenic and wild-type mice (Table 1), and histology revealed no abnormalities in transgenic mouse hearts (Fig. 3).

View larger version (37K): [in this window] [in a new window]   Fig. 1.   Agarose gel of RT-PCR products from wild-type (WT) and transgenic (TG) mice. Lanes 2-5 are from 2 WT controls, lanes 6 and 7 from founder 2 (m2), lanes 8 and 9 from founder 7 (m7), and 10 and 11 from founder 9 (m9). +, WT mRNA and , mutant mRNA. In founders 2 and 7 amounts of mutant and WT mRNA are approximately equal, and in founder 9 mutant mRNA is less than in WT mice. cTnI, cardiac troponin I; bp, base pair.

View larger version (47K): [in this window] [in a new window]   Fig. 2.   A representative immunoblot of total cTnI levels from WT and TG mice. Lanes 1 and 2 are from 2 WT controls, lane 3 from m2, lane 4 from m7, and lane 5 from m9.

View larger version (117K): [in this window] [in a new window]   Fig. 3.   Histology (hematoxylin and eosin) of WT (A) and TG (B) at ×40 magnification. Normal histology is seen in the TG mice.

Intracellular calcium. Systolic and diastolic intracellular calcium was significantly lower in transgenic mice (ANOVA, P  0.001). Levels of intracellular calcium were reduced by 35% and 32% at peak systole and diastole, respectively, at perfusate calcium of 3.5 mM (Table 2, and Fig. 4). Developed pressure was not significantly different between wild-type and controls, indicating that the pressure-calcium relationship in the transgenic mice was significantly altered compared with controls. The duration of the calcium transient (measured from the initial increase in intracellular calcium to 90% of return to baseline) was increased by 27% in the transgenic mice at perfusate calcium 2.5 mM, and 12% at perfusate calcium 3.5 mM (ANOVA, P < 0.01). Further information regarding the pressure-calcium relationship was obtained from pressure-calcium plots (Fig. 4). These were created from simultaneous pressure and calcium (derived from fluorescence) measurements, averaged over five to seven cardiac cycles. Consistent with the altered pressure-calcium relationship, pressure at the peak of the calcium transient was increased in the transgenic mice despite the prolongation of the calcium transient. The subsequent slope of calcium decay versus pressure (taken from peak calcium to peak pressure) was decreased in the transgenic mice reflecting both the prolonged calcium transient and altered pressure-calcium relationship (Table 2, Fig. 4).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Intracellular calcium and calcium-pressure relations from WT (A and B) and TG (C and D) mice. Reduced levels of intracellular calcium at perfusate calcium 3.5 mM and prolonged calcium transients are seen in the TG mice.

Myocardial oxygen consumption. Myocardial oxygen consumption was measured as a decrease in calcium cycling that may result in a decrease in myocardial oxygen consumption. There was a significant reduction in myocardial oxygen consumption in transgenic mice compared with wild-type mice (ANOVA, P < 0.01; Table 1). Developed pressures were not significantly different between the two groups, though the ratio of developed pressure to myocardial oxygen consumption [in mmHg/(µmol · min¹ · g dry wt¹)], an index of energetic efficiency, was significantly higher in the transgenic mice at perfusate calcium of 3.5 mM (P < 0.05).

Ischemia and reperfusion. Figure 5A shows representative Western immunoblots with translocation of PKC- to the particulate fraction after 15 min of ischemia in both wild-type and transgenic mice, and Fig. 5B demonstrates the laser densitometry quantification of PKC- content in fractionated tissue with significant (P < 0.05, n = 6 measurements per group) increase in the particulate fraction after 15 min of ischemia in both wild-type and transgenic mice.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   A: representative Western immunoblots demonstrating translocation of protein kinase C (PKC-) to the particulate fraction during 15-min ischemia in WT and TnI mutant mice. B: laser densitometry quantification of PKC- content in fractionated tissue samples. Bars represent means ± SD of 6 measurements per group. Solid bars, WT mice. Open bars, TnI mutant mice. *P < 0.05 vs. no ischemia.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Examples of pressure tracings (slow time based) during development of ischemic contracture in WT (A) and TG (B) mice; perfusate calcium 3.5 mM. Arrows indicate time to development of ischemic contracture, an increase in diastolic pressure of 5 mmHg above baseline, which was earlier in TG mice at perfusate calcium 3.5 mM. In the early reperfusion period, diastolic pressures increase to a greater extent in the TG mice.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Systolic and diastolic pressures at baseline, ischemia, and during reperfusion for WT and TG mice at perfusate calcium 2.5 mM. WT, n = 7; TG, n = 8; *P < 0.05 WT vs. TG.

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Systolic and diastolic pressures at baseline, ischemia, and during reperfusion for WT and TG mice at perfusate calcium 3.5 mM. WT, n = 6; TG, n = 6; **P < 0.01 WT vs. TG.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In transgenic mice that express mutant TnI lacking functional PKC phosphorylation sites, there is a marked reduction in both systolic and diastolic levels of intracellular calcium at similar developed pressures indicating an altered pressure-calcium relation in the transgenic hearts. There is also significant prolongation of the duration of the calcium transient. These transgenic mice are susceptible to global ischemia with more rapid and extensive development of ischemic contracture and postischemic dysfunction, most likely because of the inability of PKC to phosphorylate TnI and decrease the force-calcium relation. Finally, these mice have a significantly elevated ratio of developed pressure to myocardial oxygen consumption at high-perfusate calcium, which may signify an energetic benefit of decreased calcium cycling. These findings are from perfused hearts at high-perfusate calcium (3.5 mM) (17) or during global ischemia (1), which have been used to activate PKC. We have previously demonstrated in this perfused mouse heart model that the selective PKC antagonist chelerythrine can inhibit calcium-induced inotropy, suggesting that PKC is activated by high-perfusate calcium (15).

Altered pressure-calcium relationship in the perfused mouse heart. On the basis of in vitro studies demonstrating a reduction in maximal calcium-stimulated ATPase activity produced by PKC phosphorylation of TnI (19), an alteration in the pressure-calcium relationship is predicted in these mice. Though this prediction is clearly demonstrated, the surprising finding is that developed pressure is not increased but normal in these mice. The markedly reduced intracellular calcium levels indicate that the transgenic heart has adapted to the increased calcium sensitivity by decreasing the calcium transient, suggesting there is direct feedback from the myofilament to the processes controlling calcium homeostasis. This is further suggested by the prolongation of the calcium transient duration, which is seen at both levels of perfusate calcium. A precedent for such an effect may be seen in congestive heart failure. It has been well described that in heart muscle from patients with congestive heart failure there are alterations in calcium handling, with prolongation in calcium transients and increased diastolic levels of intracellular calcium with higher levels of perfusate calcium (11). However, Perez et al. (20) have recently studied the role of calcium cycling versus myofilament dysfunction in the spontaneous hypertension and heart failure rat model. They also demonstrated slowing of calcium cycling, but a much greater reduction in maximal calcium activated force. Their analysis revealed that the slowing of calcium cycling prolonged the time available for calcium to activate the contractile apparatus, as a compensatory response to reduced myofilament function. Thus myofilament function may be able to effect calcium cycling, a possible explanation for the reduced calcium levels and prolonged transients seen in the present study. The mechanism for the feedback from myofilament dysfunction to altered calcium cycling is not known. Minamisawa et al. (18) have recently shown that phospholamban ablation reverses the dilated cardiomyopathy phenotype of the mouse deficient in the Lin-11, Isl-1, and Mec-3 (LIM) cytoskeletal protein. It would be interesting to study sarcoplasmic reticulum regulatory proteins in these transgenic mice.

Ischemia-reperfusion. Ischemia is associated with a progressive rise in intracellular calcium (25) and activation of PKC (1), which we have confirmed in this perfused mouse heart model. The resulting phosphorylation of PKC sites on TnI is expected to reduce maximal ATPase activity, which may protect against the development of ischemic contracture. Thus we predicted that these transgenic mice would be susceptible to development of ischemic contracture, and that this effect would be greater with higher levels of perfusate calcium. The effects of PKC on actomyosin ATPase activity are greater during acidosis (as is expected in global ischemia), and this may explain why marked differences were seen at 2.5 mM perfusate calcium in the ischemia experiments, though differences were less marked in the intracellular calcium or myocardial oxygen consumption experiments at the same perfusate calcium. Importantly, this effect is not a nonspecific effect of overexpression of TnI, because total TnI levels in wild-type and transgenic mice are similar. Whereas the experiments studying intracellular calcium levels were performed during normoxia, the altered pressure-calcium relationship is a plausible explanation of the predisposition to ischemic contracture.

Energetic effects of mutant TnI. Recent studies have demonstrated that calcium cycling may account for a significant proportion of the energy consumption (thus oxygen consumption) of the heart. For instance, in the perfused rat heart, Grandis et al. (10) have compared the inotropic and energetic effects of the calcium-sensitizing agent EMD-57033, and the -adrenergic agonist dobutamine. Whereas dobutamine produces a positive inotropic effect by increasing the amplitude of the calcium transient (22), EMD-57033 does not increase calcium cycling (24). At doses chosen to result in equal increases in developed pressure, there was a marked increase in myocardial oxygen consumption with dobutamine, though only a small increase in myocardial oxygen consumption with EMD-57033, not significantly different from control. Similar results have also been obtained in isolated myocyte studies (21). Thus we predicted that in these transgenic mice with reduced calcium cycling that myocardial oxygen consumption would be also reduced. Consistent with this, the reduced calcium transient and energetic advantage (developed pressure/myocardial oxygen consumption, Table 1) are both seen only with high-perfusate calcium. Recently, Brandes and Bers (4) have analyzed the contribution of mechanical work and calcium cycling to total mitochondrial ATP hydrolysis as measured by NADH levels by using fluorescent spectroscopy of rat cardiac trabeculas. They concluded that there was an equal contribution of mechanical work and calcium cycling to total ATP hydrolysis. Thus, given a 35% reduction in peak systolic calcium in the present study of transgenic mice, a 17.5% reduction in ATP hydrolysis would be expected, which compares favorably to the measured 15% reduction in the ratio of developed pressure to myocardial oxygen consumption.

Limitations. We do not know the wild-type-to-mutant protein ratio in transgenic mice because the TnI antibodies did not selectively recognize mutant protein. This uncertainty makes it difficult to assess whether the phenotype demonstrated is a result of the mutant protein, or whether it is also an effect of the mutant protein on wild-type protein function. However, we do know that in two of the three transgenic lines used in these experiments that the ratio of mutant to wild-type mRNA is ~50:50. Nevertheless, transgenic techniques that change the endogenous gene would avoid this uncertainty relating to the efficiency of mutant protein production. Also, total TnI levels including mutant and wild-type proteins are similar in wild-type and transgenic mice, so the results presented are not the result of overexpression of TnI. All of the the experiments in the current study are performed in isolated isovolumically contracting hearts, and it must be recognized that this experimental preparation may produce different results to ejecting, or in vivo preparations. Nevertheless, in the present study, the data regarding developed pressure, myocardial oxygen consumption, and intracellular calcium are all derived from the same experimental methods, i.e., the perfused mouse heart, allowing correlations between these data. Also, we have used the physiological heart rate of 8 Hz. Because the mouse has a negative force-frequency relationship (5), higher pressures and altered calcium levels at lower stimulation rates may be expected. With respect to rhod 2 measurements of intracellular calcium, other investigators using myocyte studies and different loading procedures to those described in the present study have used rhod 2 as an indicator of mitochondrial calcium (28). Though we cannot exclude a small proportion of rhod 2 leaking into organelles, we have previously shown that there is predominant cytosolic localization of rhod 2 (7, and unpublished data by MacGowan and Koretsky). The differences in rhod 2 localization probably relate to different loading conditions used by different investigators.