Impaired cardiomyocyte relaxation and diastolic function in transgenic mice expressing slow skeletal troponin I in the heart

Abstract

1. To assess the specific functions of the cardiac isoform of troponin I (cTnI), we produced transgenic mice that expressed slow skeletal troponin I (ssTnI) specifically in cardiomyocytes. Cardiomyocytes from these mice displayed quantitative replacement of cTnI with transgene-encoded ssTnI.

2. The ssTnI transgenic mice were viable and fertile and did not display increased mortality or detectable cardiovascular histopathology. They exhibited normal ventricular weights and heart rates.

3. Permeabilized transgenic cardiomyocytes demonstrated an increased Ca²⁺ sensitivity of tension and a lack of contractile responsiveness to cAMP-dependent protein kinase (PKA). Isolated cardiomyocytes from transgenic mice had normal velocities of unloaded shortening but unlike wild-type controls exhibited no enhancement of the velocity of shortening in response to treatment with isoprenaline. Transgenic cardiomyocytes exhibited greater extents of shortening than non-transgenic cardiomyocytes at baseline and after treatment with isoprenaline.

4. The rates of rise of intracellular [Ca²⁺] and the peak amplitudes of the intracellular [Ca²⁺] transients were similar in transgenic and wild-type myocytes. However, the half-time of intracellular [Ca²⁺] decay was significantly greater in the transgenic myocytes. This change in decay of intracellular [Ca²⁺] was correlated with an increase in the re-lengthening time of the transgenic cells.

5. These changes in cardiomyocyte function in vitro were manifested in vivo as impaired diastolic function both at baseline and after stimulation with isoprenaline.

6. Thus, cTnI has important roles in regulating the Ca²⁺ sensitivity of cardiac myofibrils and controlling cardiomyocyte relaxation and cardiac diastolic function. cTnI is also required for the normal responsiveness of cardiomyocytes to β-adrenergic receptor stimulation.

Despite sharing structurally similar sarcomeric proteins, skeletal muscle fibres and cardiac myocytes exhibit important differences in contractile properties that reflect the distinct functions of the two muscle lineages in most higher organisms. In contrast to skeletal muscle fibres, cardiomyocytes exhibit reduced responsiveness to Ca²⁺ (i.e. a shallower tension-pCa relationship) and pronounced increases in contractility as length is increased (Frank-Starling properties). Furthermore, in response to β-adrenergic receptor stimulation, cardiomyocytes display decreased myofilament Ca²⁺ sensitivity, enhanced contractility and faster relaxation compared with skeletal muscle fibres.

There is considerable evidence that many of these differences in skeletal and cardiac muscle function reflect the expression of distinct myofibrillar protein isoforms in these two muscle lineages. Each of the myofibrillar proteins is encoded by multiple genes whose expression is dynamic and may not be restricted to one muscle type (Schiaffino & Reggiani, 1996). Changes in protein isoform expression often occur within the same muscle lineage during normal embryonic and postnatal development as well as in response to both physiological and pathophysiological stimuli in adult muscle cells. A molecular understanding of the role of specific contractile protein isoforms in determining the phenotypic differences between cardiac and skeletal muscle will yield novel insights concerning sarcomere function and may also have important implications for the treatment of human cardiac diseases.

Several complementary approaches have been used to study the roles of individual contractile protein isoforms in sarcomere function. These include ultrastructural studies, protein biochemistry and biophysical analyses of permeabilized and intact single myocyte and multicellular preparations (Schiaffino & Reggiani, 1996; Solaro & Rarick, 1998). More recently, a number of groups have utilized transgenic technologies to produce targeted alterations in contractile protein isoform expression in cardiac myocytes in mice (Metzger et al. 1993; Muthuchamy et al. 1995; Palermo et al. 1995; Oberst et al. 1998; Tardiff et al. 1998). Genetically altered mice are particularly useful because they allow the correlation of biochemical and cellular contractile properties with acute and long term changes in cardiovascular function at the level of both the organ and the whole organism.

The myofibrillar thin filament is composed of repeating functional units of seven actin monomers, a coiled-coil tropomyosin dimer and one troponin complex (Farah & Reinach, 1995; Tobacman, 1996; Solaro & Rarick, 1998). The troponin complex is composed of three subunits: troponin C (TnC), troponin I (TnI) and troponin T (TnT). TnI, the inhibitory component of the complex, is a 27-31 kDa polypeptide that can bind to actin-tropomyosin and inhibit actomyosin ATPase activity. This TnI-mediated inhibition of contraction is relieved by a complex allosteric change in the thin filament that occurs upon Ca²⁺ binding to the regulatory sites of the TnC subunit of the troponin complex (Solaro & Rarick, 1998). There are three known isoforms of TnI, each encoded by a separate gene and each displaying unique spatial and temporal patterns of expression (Schiaffino & Reggiani, 1996). In the heart, the slow skeletal isoform of TnI (ssTnI) is expressed during embryonic and early postnatal life and is then replaced entirely by the cardiac-specific isoform (cTnI) between 2 and 3 weeks after birth (and for the remainder of adult life) (Bhavsar et al. 1991; Hunkeler et al. 1991; Sasse et al. 1993). ssTnI is also expressed in slow skeletal muscle fibres (Wade et al. 1990; Corin et al. 1994). The third isoform of TnI, fsTnI, is expressed in fast skeletal muscle fibres (Dhoot & Perry, 1979; Koppe et al. 1989). Despite overall similarity, there are a number of significant structural differences between ssTnI and cTnI. Perhaps most importantly, cTnI has a 27-33 amino acid N-terminal extension which is absent in ssTnI and which contains two cAMP-dependent protein kinase (PKA) phosphorylation sites. These sites have been shown to be phosphorylated in response to β-adrenergic receptor stimulation and have been postulated to play important roles in regulating TnI-TnC interactions, the co-operative binding of troponin to actomyosin, and the affinity of the regulatory sites of TnC for Ca²⁺ (Robertson et al. 1982; Zhang et al. 1995; Chandra et al. 1997).

To better assess the role of cTnI in determining the unique contractile properties of cardiac muscle, we have produced transgenic mice in which cTnI expression in adult cardiomyocytes has been quantitatively replaced by cardiomyocyte-specific expression of a ssTnI transgene driven by the α-myosin heavy chain (α-MHC) promoter. Such transgenic mice are viable and fertile and do not display detectable cardiovascular pathology or increased mortality up to 18 months of age. Studies of the contractile properties of isolated cardiomyocytes and whole hearts from these ssTnI transgenic mice demonstrate important roles for cTnI in regulating the Ca²⁺ sensitivity of cardiac myofibrils and in controlling cardiomyocyte relaxation in vitro and cardiac diastolic function in vivo. The results also suggest that the cTnI-specific effects on contractile properties in response to β-adrenergic receptor stimulation are in large part due to phosphorylation of cTnI.

METHODS

Generation of ssTnI transgenic mice

The α-MHC ssTnI-FLAG and α-MHC ssTnI transgenes contain the murine slow skeletal troponin I cDNA (including a 5′ FLAG epitope tag in the α-MHC ssTnI-FLAG transgene; Eastman Kodak, New Haven, CT, USA) cloned into the Not I-Sal I-digested pMHC poly(A) vector (Subramaniam et al. 1991). This vector contains a 5.8 kb BamHI-MaeIII fragment of the murine α-MHC gene that includes the promoter and exons 1-3 from the 5′ untranslated region of the gene as well as an SV40 polyadenylation site (nucleotides 2500-2700 of the SV40 genome). The unique HindIII site in the pMHC poly(A) vector was modified with an adapter containing a XhoI site. A 5′NotI site (including the FLAG epitope sequence in the α-MHC ssTnI-FLAG transgene) and a 3′SalI site were generated outside of the coding sequence of the cDNA by PCR amplification. The identities of both transgenes were confirmed by DNA sequence analysis. Female CD1 mice (3.5 days postcopulation) were deeply anaesthetized with methoxyflurane and killed by cervical dislocation. Single cell embryos obtained from the oviducts of these animals were injected with Xho I-digested transgenes to produce the ssTnI and ssTnI-FLAG transgenic mice. Transgenic founders were identified by Southern blot analysis of tail DNA. PstI-digested genomic DNA was transferred to nitrocellulose membranes and hybridized with a 442 bp radiolabelled XbaI-NotI fragment containing exons 2 and 3 from the 5′ untranslated region of the murine α-MHC gene. This probe detects both the endogenous α-MHC gene and the transgene. All animal experimentation was performed in accordance with National Institutes of Health guidelines, and approved by the Animal Care and Use Committee of the University of Chicago.

Western blot analysis

Freshly isolated heart, soleus, lung, liver and kidney samples were homogenized in NETN (20 mM Tris-HCl pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.1 % Nonidet P-40) with 1 mM phenylmethylsulphonic fluoride (PMSF), 1 mM benzamidine, 5 μg ml⁻¹ aprotinin, 1 μg ml⁻¹ pepstatin A and 1 μg ml⁻¹ leupeptin (Sigma). Homogenized samples were boiled in Laemmli loading buffer, subjected to SDS-PAGE, and transferred to nitrocelluose membranes (Schleicher and Schuell, Keene, NH, USA). TnI was detected using a 1 : 1000 dilution of a monoclonal antibody directed against skeletal muscle TnI (no. 10T81, Fitzgerald Industries, Concord, MA, USA) at 4°C in blotto (5 % dry milk, 10 mM Tris-HCl pH 7.5, 140 mM NaCl). This antibody recognizes all three isoforms of murine TnI. Immunoreactivity was detected using horseradish peroxidase-conjugated goat anti-mouse IgG (Life Technologies, Grand Island, NY, USA) at room temperature in blotto in conjunction with an enhanced chemiluminescence system (Kirkegaard and Perry Laboratories, Gaithersburg, MD, USA).

Myocyte protein electrophoresis

Ventricular myocytes from transgenic mice and wild-type littermate controls (see below) were diluted in urea/thiourea sample buffer (Strang et al. 1994) and stored at -80°C. Samples were thawed and heated immediately before use. Cardiomyocyte proteins were separated by SDS-PAGE in 15 % acrylamide with a 200 : 1 acrylamide : bisacrylamide ratio using a multiphasic Laemmli buffer system (Hoefer SE-260) as described previously. The resulting gels were fixed overnight in glutaraldehyde, washed, silver stained, and dried between Mylar sheets. Gels were then scanned using an image densitometer (Molecular Analyst, BioRad) and commercially available software.

Histological analysis

Mouse hearts were fixed in 4 % paraformaldehyde overnight at 4°C and embedded in paraffin, and 4 μm sections were cut on a Jung Histocut 820 microtome. Sections were stained with Haematoxylin and Eosin or Masson's trichrome stain. Photomicrographs were obtained using Nikon SMZ-U or Zeiss Axiophot microscopes.

Preparation of skinned myocytes

The isolation and attachment of skinned (chemically permeabilized) ventricular myocytes were performed as described previously (Strang et al. 1994) with slight modification. Hearts were excised from transgenic mice and wild-type littermate controls after the mice had been killed with inhaled methoxyflurane in accordance with institutional guidelines. The hearts were placed in ice-cold relaxing solution (containing (mM): free Mg²⁺, 1; KCl, 100; EGTA, 2; ATP, 4; and imidazole, 10; pH 7.0), the atria were removed, and the ventricular tissue was minced into five to six pieces. The minced ventricular tissue was then mechanically disrupted using a polytron (Kinematica). The resulting suspension of cells and cell fragments was centrifuged at 165 g for 120 s and the pellet was then resuspended in 0.3 % Triton X-100 for 6 min to permeabilize sarcolemmal, mitochondrial and sarcoplasmic reticulum membranes. After washing in fresh relaxing solution, myocytes were resuspended in relaxing solution and kept at 4°C until used (within 8 h). On the stage of an inverted microscope, single permeabilized ventricular myocytes were attached with silicone adhesive (Dow Corning) to stainless steel pins fastened to the active elements of a force transducer (model 403, Cambridge Technology). A motor (model 6350, Cambridge Technology) enabled length step changes within 1.5 ms. After curing of the silicone attachments, the attached myocyte preparation was transferred to relaxing solution and sarcomere length was adjusted to 2.3 μm using on-line videomicroscopy.

Velocities of unloaded shortening in skinned myocytes

Velocities of unloaded shortening were determined at 15°C in a maximally activating solution (pCa 4.5, pH 7.0; containing (mM): EGTA, 7; free Mg²⁺, 1; MgATP, 4; creatine phosphate, 14.5; and imidazole, 20; ionic strength, 180 mM) using the slack-test method as described previously (Strang et al. 1994). After steady tension was reached in maximally activating Ca²⁺ solution, the preparation was rapidly slackened; the time required to take up the imposed slack was measured as the interval between the beginning of the imposed slack length step and the onset of tension redevelopment. Plots of slack length versus duration of unloaded shortening were included if well fitted by a straight line (r= 0.95). In experiments investigating the effects of cAMP-dependent protein kinase (PKA) on unloaded shortening velocities, the myocyte preparation was incubated after attachment to the force transducer for 40 min at 20°C in relaxing solution containing the catalytic subunit of porcine cardiac PKA (3.0 μg ml⁻¹).

Ca²⁺ sensitivity of isometric tension in skinned myocytes

Developed isometric tension was measured at a sarcomere length of 2.3 μm (15°C) in maximally activating Ca²⁺ solution (pCa 4.5) and submaximal Ca²⁺ solutions (pCa 5.7 and 5.1) as described previously (Strang et al. 1994). Isometric tensions measured at submaximal pCa were expressed as a fraction of maximal isometric tension (P₀), i.e. relative tension, P_rel=P/P₀, and were then plotted versus pCa and analysed by least squares regression using the Hill equation: log(P_rel/(1 - P_rel)) =nlog[Ca²⁺]+k where n is the Hill coefficient and k is the intercept of the fitted line with the x-axis in pCa units. Lines were fitted to the tension-pCa curves by inserting constants derived from the above analysis into the following equation: P_rel=[Ca²⁺]ⁿ/(kⁿ+[Ca²⁺]ⁿ) where k denotes the pCa at which relative tension is half-maximal, i.e. pCa₅₀. Data were discarded if maximal tension declined by more than 15 % during the experiment or if mean sarcomere length changed by more than 0.2 μm between isometric relaxed and maximally activating conditions. In experiments investigating the effects of PKA upon the Ca²⁺ sensitivity of tension, after determination of the Ca²⁺ sensitivity of tension under control conditions, the myocyte preparation was incubated at 20°C for 40 min in relaxing solution containing the catalytic subunit of porcine cardiac PKA (3.0 μg ml⁻¹) after which the Ca²⁺ sensitivity of tension was again determined.

Intact myocyte isolation

Hearts were excised as described above and placed in ice-cold perfusion solution (pH 7.4) containing (mM): NaCl, 118; KCl, 4.8; NaH₂PO₄, 1.2; MgCl₂, 1.2; Hepes, 25; pyruvic acid, 5; and glucose, 11. The hearts were then perfused in a retrograde manner using a modified Langendorff apparatus with Ca^{2 +} -free perfusion solution containing 1 mg ml⁻¹ bovine serum albumin and 0.016 mg ml⁻¹ leupeptin at 37°C for 5 min, followed by perfusion for 15-20 min with perfusion solution containing 200 μM Ca²⁺ and 0.45 mg ml⁻¹ collagenase Type D (Boehringer). The hearts were then removed from the Langendorff apparatus and ventricular tissue was minced with scissors and then triturated with a pipette to disaggregate myocytes. After filtering, the myocyte suspension was pelleted by centrifugation and resuspended in fresh perfusion solution containing 100 μM Ca²⁺ three times to ensure removal of collagenase. The cells were stored in perfusion solution containing 100 μM Ca²⁺ at room temperature (22-23°C) until used (within 8 h).

Myocyte shortening and intracellular [Ca²⁺] transients

After isolation, intact myocytes were incubated for 20 min at room temperature in perfusion solution containing 100 μM Ca²⁺ and 3 μM fluo-3 acetoxymethyl ester (fluo-3 AM, Molecular Probes). Residual extracellular dye was removed by centrifugation and replacement of the supernatant with fresh perfusion solution containing no fluo-3 AM three times. Fluorescence and cell shortening of myocytes were determined as described previously (Wolska et al. 1996; Patel et al. 1997) with modification. Myocytes loaded with fluo-3 AM were pipetted into a 250 μl field-stimulation perfusion chamber that was mounted on the stage of an inverted microscope (Olympus). Light from a halogen lamp source was initially passed through a cut-off filter (transmission, > 620 nm) and then through the perfusion chamber, a × 20 epo-fluorescence objective (Olympus), a dichroic mirror and another cut-off filter (transmission, > 520 nm), and was then split by a beam splitter. Filtered light directed towards the eyepiece was split by a second beam splitter, so that one half was directed to a CCD camera and the other half to the eyepiece. The signal from the CCD camera was fed to a video-edge detector enabling on-line myocyte length data acquisition using commercial software (Crystal Biotech) as well as display on a chart recorder (Allen Datagraph) and oscilloscope (Nicolet). Fluo-3 AM within myocytes was excited by light (475 nm) from a fluorimeter (SLM Aminco, SLM Instruments) passed through a band-pass filter (transmission, 400-490 nm) and reflected by a dichroic mirror to the perfusion chamber. Emitted fluorescence (525 nm) followed the same path as the filtered light from the halogen lamp, except that it passed through a band-pass filter placed in front of the photomultiplier tube. The output signal from the photomultiplier tube was recorded and stored using the SLM 8000C software package.

Labelling of mouse myocytes with ³²P

The level of protein phosphorylation in myocyte preparations was determined with the use of a modified version of our previously published protocol (Gupta et al. 1994; Wolska et al. 1996). After isolation, the cells were allowed to settle for 3-5 min. The supernatant fraction was then removed and cells were resuspended in Na-Hepes phosphate-free buffer (mM: KCl, 4.8; MgSO₄, 1.2; NaCl, 132; Hepes, 10; sodium pyruvate, 2.5; and glucose, 10; pH 7.4) with 100 μM Ca²⁺. The procedure was repeated two more times with 200 and 500 μM Ca²⁺. The cells were stored in Na-Hepes phosphate-free buffer with 500 μM Ca²⁺ at room temperature (22-23°C) until used. Myocytes were incubated in 1 mM Ca²⁺ Na-Hepes buffer with 0.5 mCi [³²P]orthophosphate for 30 min at room temperature. Subsequently, cells were washed twice with the Na-Hepes solution with 1 mM Ca²⁺. The myocyte suspensions (150 μl) were mixed with 150 μl of the Na-Hepes phosphate-free buffer with 1 mM Ca²⁺ containing either no added isoprenaline, 70 nM, or 1.0 μM isoprenaline. After 2 min the reaction was stopped by adding 150 μl SDS-stop solution (mM: DTT, 1; Tris-HCl, 30; and EDTA, 3; with 6 % SDS, 15 % glycerol and a trace of Bromophenol Blue). The samples were mixed well and stored at -20°C. Before analysis by SDS-PAGE, the samples were boiled for 10 min to convert the high-molecular weight polymers of phospholamban (PLB) into its low-molecular weight form.

Determination of protein phosphorylation

Gel electrophoresis was performed using a linear 5-20 % polyacrylamide gradient gel as previously described (Rapundalo et al. 1989). An aliquot of cells containing 50 μg protein, as determined using the Lowry method, was applied to each lane. The gels were stained in 30 % methanol, 7 % acetic acid and 0.1 % Coomassie Blue R-250. Initial destaining was done with a solution containing 50 % methanol and 10 % acetic acid. For final destaining 10 % methanol and 10 % acetic acid solution was used. The SDS-PAGE gels were scanned using Personal Densitometer SI (Molecular Dynamics), vacuum dried and placed in Storage Phosphor Screen Cassettes (Kodak) for overnight exposure. The cassettes were scanned in a Molecular Dynamics Storm 840 PhosphorImager. The ImageQuant software from Molecular Dynamics was used for data processing. Phosphorimage and densitometric data were adjusted for background and used for calculations. Data were normalized to the maximum values. Myofilament proteins were identified by co-migration with known standards. PLB was identified using a monoclonal antibody.

Cardiac catheterization

Mice were anaesthetized with 100 mg kg⁻¹ ketamine, 5 mg kg⁻¹ xylazine and 1 mg kg⁻¹ buprenorphine (i.p.), and the left ventricle was catheterized via the right carotid artery using a 1.8 French micromanometer catheter (model SPR-612, Millar, Houston, TX, USA). Analog signals from the pressure transducer and ECG were digitized using an analog-to-digital converter (AD3100, Real Time Devices, State College, PA, USA). Digital files were recorded and analysed with commercially available software (Atlantis and Pegasus software, Lakeshore Technologies, Chicago, IL, USA). Isoprenaline (40 ng kg⁻¹ in 25 μl phosphate-buffered saline) was infused directly into the jugular vein. Recordings were obtained 20, 30 and 40 s after administration of isoprenaline. Five seconds of data collected during maximal stimulation were used to assess the isoprenaline response. Mice were killed at the end of each experiment by i.v. injection of 0.5 ml of a solution containing 25 mg ml⁻¹ sodium pentobarbital and 500 u ml⁻¹ heparin.

Data analysis

Data are presented as means ± standard error of the mean (s.e.m.) or means ± standard deviation (s.d.). The significance of differences between means was evaluated with two-way analysis of variance (for repeated measures when appropriate) and the Student- Neuman-Keuls test for multiple comparisons (Glantz & Slinker, 1990). Values of P < 0.05 were considered statistically significant.

RESULTS

Production of ssTnI transgenic mice

To assess the functional differences between ssTnI and cTnI and to determine the unique roles of cTnI in cardiomyocyte contractile function, we produced transgenic mice that expressed ssTnI in ventricular myocytes under the control of the cardiomyocyte-specific α-MHC promoter. This promoter is expressed in both embryonic and adult atrial cardiomyocytes. However, its expression in ventricular myocytes is restricted to postnatal and adult life. Two different murine ssTnI transgenes were used in these experiments. The first contained an N-terminal FLAG epitope tag that allowed us to definitively identify the transgene by Western blot analysis with an anti-FLAG antibody. The second transgene contained the wild-type murine ssTnI cDNA without the FLAG epitope tag. After injection of these transgenes into fertilized mouse embryos, two independent lines of transgenic mice were obtained that contained approximately ten copies of each transgene as assessed by Southern blot analysis (Fig. 1A). Because identical results were observed in the two transgenic lines, they will not be distinguished below.

Figure 1. Molecular characterization of the ssTnI mice.

A, Southern blot analysis of tail DNA from two independently derived lines (Tg-FLAG and Tg) of ssTnI transgenic mice and a non-transgenic control littermate (NTg). The probe detects both the endogenous α-MHC gene (αMHC) and the transgene (Tg). B, Western blot analysis of TnI expression in cardiac protein extracts from non-transgenic control mice (NTg), α-MHC ssTnI-FLAG (Tg-FLAG) and α-MHC ssTnI (Tg) transgenic mice. Exclusive expression of cTnI was observed in the non-transgenic heart (first lane), whereas expression of both fsTnI and ssTnI was seen in the soleus (second and fifth lanes). Expression of ssTnI or ssTnI-FLAG transgene-encoded proteins completely replaced endogenous cTnI expression in the transgenic mice (third and fourth lanes). Note the slower electrophoretic mobility of the transgenic ssTnI-FLAG protein relative to wild-type ssTnI due to the addition of the 9 amino acid FLAG epitope tag. This band was unambiguously identified as the FLAG-tagged transgenic protein by immunoblotting with an anti-FLAG antibody (data not shown). C, myofibrillar protein expression in wild-type and transgenic cardiomyocytes. Lanes 1 and 2 show protein content of cardiomyocytes isolated from the hearts of non-transgenic (NTg) and α-MHC ssTnI-FLAG (Tg) transgenic mice, analysed by SDS-PAGE followed by silver staining of the gel. MHC, myosin heavy chain; cTnT, cardiac troponin T; cTnI, cardiac troponin I; ssTnI FLAG, FLAG epitope-tagged slow skeletal muscle troponin I; MLC_1V, myosin light chain 1V; cTnC, cardiac troponin C; MLC_2V, myosin light chain 2V.

Western blot analysis demonstrated that adult hearts from both lines of transgenic mice displayed quantitative replacement of cTnI expression with their respective ssTnI transgene-encoded proteins (Fig. 1B). The transgenic ssTnI proteins could be distinguished from endogenous cTnI both by their different mobilities in SDS-PAGE and by using an anti-FLAG antibody (in the case of the transgenic ssTnI-FLAG mice). Unlike adult wild-type hearts, which expressed cTnI exclusively, the adult transgenic hearts appeared to express only the transgene-encoded slow skeletal isoform of TnI. Moreover, transgene expression appeared to be restricted to the heart. The levels of transgene proteins expressed in the hearts of both lines of transgenic mice were similar to those of endogenous cTnI expressed in the hearts of wild-type littermates (Fig. 1B). This quantitative replacement of cTnI by the transgene-encoded protein may reflect the capacity of cardiomyocytes to regulate the total level of expression of a single contractile protein and its assembly into sarcomeres at the post-translational level.

To rule out the possibility that expression of the ssTnI transgene altered the expression of other contractile proteins in cardiomyocytes, we compared contractile protein expression in transgenic and wild-type cardiomyocytes by SDS-PAGE. Other than the expected alterations in TnI, there were no detectable differences in the expression of contractile proteins in the ssTnI transgenic hearts (Fig. 1C). Moreover, immunoblot analysis with an antibody reactive to all isoforms of TnT did not reveal any differences in the pattern of expression of TnT isoforms in the ssTnI transgenic as compared with wild-type hearts (data not shown). Taken together, these results suggest that the ssTnI transgenic mice should allow an accurate assessment of the effects of stoichiometric replacement of cTnI with ssTnI in cardiomyocytes in vivo.

Phenotype of the ssTnI transgenic mice

The ssTnI mice were fertile and viable. Moreover, we did not observe any increased mortality or gross cardiovascular pathology in these mice up to 18 months of age, which is the longest that any were kept. Heart weight to body weight ratios (expressed as total ventricular mass/body weight and left ventricular mass/body weight) and anaesthetized heart rates (both at baseline and during isoprenaline infusion) were not significantly different from those of age-matched littermate wild-type controls (Fig. 2A and B). The ssTnI mice did not show any detectable histopathological changes, as assessed by both light microscopy (Fig. 3A-D) and electron microscopy (data not shown).

Figure 2. Ventricular mass and heart rate of the ssTnI transgenic hearts.

A, comparison of total ventricular and left ventricular mass to body mass ratios for non-transgenic (NTg) and α-MHC ssTnI transgenic (Tg) mice. Both ventricles were trimmed from freshly isolated hearts from non-transgenic and transgenic mice, and their masses were determined. Ratios are expressed as ventricular weight (in mg) divided by body mass (in g). The free wall of the right ventricle (RV) was then trimmed away to determine left ventricular (LV) mass to body mass ratio. B, comparison of heart rates (beats min⁻¹, bpm) of 10-week-old non-transgenic and α-MHC ssTnI transgenic mice.

Figure 3. Histological analysis of the ssTnI hearts.

A-D, light microscopic analysis of hearts from α-MHC ssTnI transgenic mice (Tg) (B and D) and wild-type control littermates (A and C). Hearts were fixed in 4 % paraformaldehyde, sectioned at the level of the papillary muscles, and stained with Haematoxylin and Eosin. A and B show high power photomicrographs of samples from 2.5-month-old mice; scale bar, 25 μm. C and D are samples from 1-year-old mice at the same magnification. No pathological changes were detected in the transgenic hearts at either age.

Ca²⁺ sensitivity of isometric tension and velocities of unloaded shortening in ssTnI transgenic myocytes

The attachment procedure consistently provided low compliance attachments to skinned myocytes, so that sarcomere length could be monitored while relaxed and during maximal activation (Fig. 4). As shown in Table 1 and Fig. 5, compared with wild-type myocytes expressing cTnI, ssTnI-expressing transgenic myocytes had a significantly greater Ca²⁺ sensitivity of tension indicated by a leftward shift in the tension-pCa relationship (wild-type pCa₅₀= 5.61 ± 0.03; transgenic pCa₅₀= 5.98 ± 0.04). After treatment with PKA, there was a significant shift of the pCa-tension relationship of wild-type myocytes to higher [Ca²⁺] (pCa₅₀= 5.42 ± 0.03), indicating that PKA treatment caused a reduction in the Ca²⁺ sensitivity of tension. However, there was no change in the pCa-tension relationship of transgenic myocytes after PKA treatment (pCa₅₀= 5.97 ± 0.03), i.e. PKA had no effect upon the Ca²⁺ sensitivity of tension of transgenic myocytes expressing ssTnI (Fig. 5). There was no difference in apparent co-operativity (Hill coefficient) of tension development in transgenic and wild-type myocytes under control conditions or after PKA treatment (Table 1).

Figure 4. A single ventricular cardiomyocyte during measurements of Ca²⁺ sensitivity of isometric tension.

Photomicrographs of a single murine transgenic cardiac myocyte while relaxed in pCa 9.0 solution (above) and during maximal activation in pCa 4.5 (below). Scale bar is 25 μm.

Table 1.

Ca²⁺ sensitivity of tension and velocity of unloaded shortening of skinned myocytes expressing wild-type cTnI or transgenic ssTnI

	pCa50	Hill coefficient	Vmax (muscle length S−1)
Wild-type
control	5.61 ± 0.03 (10)	3.75 ± 0.32 (10)	2.26 ± 0.14 (10)
PKA	5.42 ± 0.03* (10)	3.25 ± 0.18 (10)	3.64 ± 0.24* (10)
Transgenic
Control	5.98 ± 0.04* (10)	2.67 ± 0.23 (10)	2.56 ± 0.20 (10)
PKA	5.97 ± 0.03† (10)	2.95 ± 0.18 (10)	2.60 ± 0.23 (10)

Data are means ± s.e.m., with the number of cells in parentheses. pCa₅₀, the pCa at which relative tension is half-maximal; Vmax, velocity of unloaded shortening.

^*Significantly different from wild-type control (P < 0.05)

^†significantly different from wild-type PKA (P < 0.05).

Figure 5. pCa-tension relationships of wild-type and transgenic cardiomyocytes.

pCa-tension relationships of skinned cardiac wild-type myocytes expressing cTnI (NTg) and transgenic myocytes expressing ssTnI (Tg) are shown before and after treatment with PKA. •, Tg; ○, Tg + PKA; ▴, NTg; ▵, NTg + PKA.

The slack-test procedure yielded reproducible linear plots of length step amplitude versus duration of unloaded shortening for step sizes between 16 and 22 % of myocyte length. The resulting velocities of unloaded shortening of wild-type and transgenic myocytes did not differ significantly under control conditions (wild-type, 2.26 ± 0.14 muscle lengths s⁻¹; transgenic, 2.56 ± 0.20 muscle lengths s⁻¹; Table 1). After treatment with PKA, there was an increase in the velocity of unloaded shortening of wild-type myocytes (3.64 ± 0.24 muscle lengths s⁻¹), indicating that PKA treatment accelerated the kinetics of cross-bridge turnover. However, PKA did not change the velocity of unloaded shortening of transgenic myocytes (2.60 ± 0.23 muscle lengths s⁻¹).

Intact myocyte cell shortening and intracellular [Ca²⁺] transients

The extent of cell shortening during electrical stimulation of intact wild-type myocytes expressing cTnI was compared with that of transgenic myocytes expressing ssTnI under control conditions and while perfused with 70 nM isoprenaline. As shown in Table 2, under control conditions, free-floating transgenic myocytes shortened to a greater extent than wild-type myocytes (9.4 ± 0.7 % of resting length vs. 5.6 ± 0.6 % of resting length). After perfusion with isoprenaline, the extent of cell shortening increased in both wild-type and transgenic myocytes (2.8-fold and 2.2-fold, respectively), but the extent of cell shortening was still significantly greater in transgenic than in wild-type myocytes (20.3 ± 0.7 % of resting length vs. 15.7 ± 1.1 % of resting length, respectively). In further experiments, cell shortening and a fluorescence signal (fluo-3) reporting intracellular [Ca²⁺] were recorded simultaneously in intact wild-type and transgenic myocytes under control conditions and during perfusion with 70 nM isoprenaline. Similar to the case without fluo-3 loading, isoprenaline caused 3-fold and 2.2-fold increases in the extent of cell shortening in wild-type and transgenic myocytes, respectively. The peak amplitude of the intracellular [Ca²⁺] transient was similar in wild-type and transgenic myocytes under control conditions and increased to similar extents with isoprenaline perfusion, i.e. 2.1-fold and 2.2-fold, respectively. The shortening half-times (time to shorten to one-half of peak) of wild-type and transgenic myocytes were similar under control conditions and after isoprenaline perfusion. On the other hand, the re-lengthening half-time was significantly greater in transgenic myocytes both under control conditions (219 ± 18 vs. 152 ± 15 ms) and after perfusion with isoprenaline (160 ± 12 vs. 93 ± 11 ms). There was no difference in the intracellular [Ca²⁺] rise half-time either under control conditions or after isoprenaline perfusion; however, the intracellular half-time of decay of [Ca²⁺] was significantly greater in transgenic than in wild-type myocytes under control conditions.

Table 2.

Twitch parameters and intracellular Ca²⁺ fluorescence record of intact cardiac myocytes expressing wild-type cTnI or transgenic ssTnI

	% L	Shortening t1/2 (ms)	Re-lengthening t1/2 (ms)	Peak fluorescence (a.u.)	[Ca2+] rise t1/2 (ms)	[Ca2+] decay t1/2 (ms)
Wild-type
Control	5.6 ± 0.6 (11)	58 ± 5 (11)	152 ± 15 (11)	0.050 ± 0.007 (11)	22 ± 1 (11)	228 ± 24 (11)
Iso	15.7 ± 1.1* (11)	46 ± 5* (11)	93 ± 11* (11)	0.109 ± 0.016* (11)	20 ± 1 (11)	126 ± 4* (11)
Transgenic
Control	9.4 ± 0.7* (10)	62 ± 3 (12)	219 ± 18* (12)	0.056 ± 0.007 (12)	20 ± 1 (12)	287 ± 17* (12)
Iso	20.3 ± 0.7†‡ (9)	38 ± 2† (12)	160 ± 12†‡ (12)	0.123 ± 0.017† (12)	18 ± 1 (12)	140 ± 7† (12)

Data are means ± s.e.m., with the numbers of cells in parentheses. %L, extent of shortening as percentage of initial length; Shortening t_1/2, time to one-half shortening; Re-lengthening t_1/2, time to one-half re lengthening; [Ca²⁺] rise t_1/2, intracellular Ca²⁺ rise half-time; [Ca²⁺] decay t_1/2, intracellular Ca²⁺ decay halftime; Iso, 70 nM isoprenaline; a.u., arbitrary units.

^*Significantly different from wild-type control (P < 0.05)

^†significantly different from transgenic control (P < 0.05)

^‡significantly different from wildtype Iso (P < 0.05).

Normalized shortening records from wild-type and transgenic myocytes under control conditions and after perfusion with isoprenaline are compared in Fig. 6. These records demonstrate that the re-lengthening times of transgenic myocytes were longer than in wild-type myocytes under control conditions; and while perfusion with isoprenaline decreased re-lengthening time in both cell types, the re-lengthening time of transgenic myocytes was still significantly greater than for wild-type myocytes. Examination of normalized intracellular Ca²⁺ fluorescence records (Fig. 7) revealed no differences in the time to peak [Ca²⁺] between wild-type and transgenic myocytes under control conditions or after perfusion with isoprenaline. However, the decay time of the intracellular [Ca²⁺] transient was significantly slower in transgenic than in wild-type myocytes under control conditions, a difference that was abolished after perfusion with isoprenaline.

Figure 6. Normalized shortening records of membrane-intact cardiac myocytes.

A wild-type myocyte expressing cTnI (NTg) is compared with a transgenic myocyte expressing ssTnI (Tg) in the absence and presence of 70 nM isoprenaline (Iso).

Figure 7. Normalized intracellular Ca²⁺ fluorescence records.

A membrane-intact wild-type cardiac myocyte expressing cTnI (NTg) is compared with a transgenic myocyte expressing ssTnI (Tg) in the absence and presence of 70 nM isoprenaline (Iso).

We also investigated whether the differences in Ca²⁺ and contraction dynamics between wild-type and transgenic myocytes could be due to differences in the phosphorylation of phospholamban (PLB) or myosin binding protein C (MyBP-C). Isolated myocytes were loaded metabolically with ³²P to label the nucleotide pool. Figure 8 illustrates levels of protein phosphorylation in myocytes isolated from wild-type and transgenic hearts before and after treatment with isoprenaline. We could not detect phosphorylation of cTnI in the transgenic myocytes, whereas the wild-type myocytes showed the expected increases in both cTnI and PLB phosphorylation after isoprenaline stimulation (Kranias & Solaro, 1982). The wild-type and transgenic myocytes demonstrated similar basal and isoprenaline-stimulated levels of phosphorylation of MyBP-C. Levels of phosphorylation of myosin light chain 2 (MLC₂) were the same in the wild-type and transgenic myocytes.

Figure 8. Phosphorylation of proteins under basal conditions and after isoprenaline stimulation.

Myocytes were isolated and processed as described in Methods. A, wild-type myocytes. B, ssTnI-expressing transgenic myocytes. Lane 1, control conditions; lane 2, 70 nM isoprenaline; lane 3, 1 μM isoprenaline. Note the lack of phosphorylation of TnI in transgenic myocytes. Values (n= 4) of relative phosphorylation of myosin light chain 2 (MLC₂) in basal conditions and in the presence of either 70 nM or 1.0 μm isoprenaline were not significantly different. In wild-type myocytes there were significant increases in the relative phosphorylation of cTnI after treatment with isoprenaline. In both wild-type and transgenic myocytes there were significant increases in levels of phosphorylation of PLB in the presence of isoprenaline. MyBP-C, myosin binding protein C.

Impaired diastolic function in the ssTnI mice

The alterations in Ca²⁺ sensitivity and in responsiveness to PKA and isoprenaline treatment observed in single cardiomyocytes from the ssTnI transgenic mice suggested that these mice might display alterations in cardiac contractile function in vivo. Thus, we compared left ventricular pressure tracings from wild-type and ssTnI transgenic mice at baseline and after isoprenaline infusion (Fig. 9A). Systolic contractile function, as assessed by recording instantaneous left ventricular pressure (P) and calculating maximum dP/dt (dP/dt_max), was comparable at baseline in wild-type and ssTnI transgenic mice (Fig. 9B). Moreover, the ssTnI transgenic mice responded with normal increases in dP/dt_max after isoprenaline infusion. In contrast, ssTnI transgenic mice exhibited significantly greater minimum dP/dt (dP/dt_min) than wild-type mice at baseline and after isoprenaline infusion (Fig. 9B). Thus, consistent with the increased re-lengthening time of isolated cardiomyocytes in vitro, hearts of ssTnI mice exhibited significantly impaired diastolic relaxation under baseline conditions and blunted augmentation of relaxation in response to isoprenaline in vivo.

Figure 9. Haemodynamic analysis of the ssTnI mice.

A, high fidelity left ventricular pressure tracings from wild-type (NTg) and α-MHC ssTnI transgenic (Tg) mice. The continuous lines denote LV pressure; the dotted lines represent the first derivative of pressure (dP/dt). Note the relative decrease in the absolute value of dP/dt_min in the ssTnI mouse, whereas dP/dt_max and peak systolic pressure remain normal. B and C, mean maximum and minimum dP/dt in ssTnI transgenic (•) (n= 10) and non-transgenic control littermates (○) (n= 10) were measured both at baseline (B) and after treatment with a single intravenous bolus of 40 ng kg⁻¹ isoprenaline (I). The data are shown as means ±s.d.* Significant difference (P < 0.05) between the non-transgenic and transgenic groups. No significant differences were noted in heart rate or end-diastolic LV pressure at baseline or during stimulation with isoprenaline.

DISCUSSION

We used transgenic mice expressing ssTnI in adult cardiac myocytes to study the role of cTnI in determining contractile properties of cardiac muscle. The use of a transgenic mouse model allowed us to relate changes in the contractile function of isolated cardiomyocytes to alterations in cardiac function at the organ level. Moreover, the use of transgenic mice avoided potential non-specific effects on myocyte structure and function that can accompany virus-mediated gene transfer into isolated myocytes or isoform-exchange experiments into permeabilized myocytes, both of which have been used previously to study the roles of individual contractile proteins in cardiomyocyte function (Ball et al. 1994; Wattanapermpool et al. 1995; Westfall et al. 1997).

The ssTnI transgenic mice were viable and fertile and did not display obvious cardiovascular pathology. Thus, ssTnI can subserve most of the functions of cTnI in cardiomyocytes; however, significant differences in contractile function were evident. First, compared with ssTnI, cTnI confers a decreased Ca²⁺ sensitivity of tension on cardiac myofibrils. Second, cTnI appears to be required for the PKA-induced decrease in Ca²⁺ sensitivity seen in normal cardiomyocytes. Third, compared with ssTnI, cTnI appears to facilitate Ca²⁺ release from the thin filament, thereby accelerating re-lengthening of isolated cardiomyocytes and enhancing diastolic relaxation in the intact heart. Because both independently derived lines of transgenic mice displayed quantitative replacement of cTnI expression with transgene-encoded ssTnI, and because the two lines displayed identical changes in contractile function, we conclude that the observed phenotype was specific and reflected a complete isoform switch in these mice.

Our results demonstrating that transgenic expression of ssTnI in the adult heart increased the Ca²⁺ sensitivity of tension in skinned cardiomyocytes is consistent with previous observations that myofilaments of neonatal heart (which expresses ssTnI) are more sensitive to Ca²⁺ than those of the adult heart (which expresses cTnI) (Solaro et al. 1988; Dieckman & Solaro, 1990), and with a previous report demonstrating that adenovirus-mediated expression of ssTnI in freshly isolated rodent cardiomyocytes results in an increased Ca²⁺ sensitivity of tension (Westfall et al. 1997). Taken together, these results suggest that the dynamic expression of different isoforms of TnI plays an important role in determining the Ca²⁺ sensitivity of the thin filament. Previous studies have suggested that binding of TnI to TnC increases the affinity of the TnC regulatory site for Ca²⁺ (Robertson et al. 1982). Thus, the relatively higher affinity of ssTnI for TnC may explain in part the increased Ca²⁺ sensitivity of ssTnI-expressing myocytes. The C-terminal region of TnI (amino acids 136-209) is thought to interact directly with the N-terminal region of TnC which contains the regulatory Ca²⁺ binding sites (Robertson et al. 1982; Kleerekoper et al. 1995). Therefore, the structural differences between the C-terminal regions of ssTnI and cTnI may influence the affinity of binding of these two TnI isoforms to TnC and thereby determine the different Ca²⁺ sensitivities of cardiac and skeletal myofibrils. This model is testable by making transgenic mice that express chimeric TnI molecules under the control of the α-MHC promoter.

Unlike skeletal muscle fibres, normal cardiomyocytes exhibit a significant reduction in the Ca²⁺ sensitivity of tension after exposure to PKA (Herzig et al. 1981). This change in Ca²⁺ sensitivity is attributed to phosphorylation of Ser22 and Ser23, which are present in the unique N-terminal extension of cTnI (Wattanapermpool et al. 1995). Phosphorylation of cTnI is thought to reduce the Ca²⁺ sensitivity of tension by decreasing the affinity of cTnI for TnC with a resultant decrease in the affinity of TnC for Ca²⁺ (Zhang et al. 1995). In the present study, ssTnI-expressing transgenic cardiomyocytes did not exhibit a reduction in the Ca²⁺ sensitivity of tension after exposure to PKA. Moreover, in contrast to wild-type skinned cardiomyocytes in which the velocity of unloaded shortening increased after exposure to PKA, ssTnI-expressing transgenic skinned cardiomyocytes exhibited no increase in shortening velocity after PKA treatment. This observation is consistent with previous findings that cTnI phosphorylation speeds cross-bridge cycling kinetics (Strang et al. 1994). Although a role for PKA-dependent phosphorylation of MyBP-C in enhancing relaxation associated with β-adrenergic receptor stimulation has been suggested (Hartzell, 1984), our results indicate that cTnI phosphorylation is sufficient to alter the dynamics of relaxation in hearts in response to β-adrenergic receptor stimulation.

β-Adrenergic receptor stimulation of intact cardiac myocytes results in increased open probability of L-type Ca²⁺ channels (Reuter, 1983; Brum et al. 1984) which causes an increased influx of Ca²⁺ into the cytosol during the action potential. After β-adrenergic receptor stimulation, increased Ca²⁺ influx may directly activate cardiac myofilaments and may induce increased release of Ca²⁺ from the sarcoplasmic reticulum (SR) due to the graded nature of calcium-induced calcium release (Niggli & Lederer, 1990), increased Ca²⁺ content of the SR, or increased open probability of the phosphorylated SR ryanodine-sensitive Ca²⁺ release channel (Takasago et al. 1989; Yoshida et al. 1992; Strand et al. 1993; Valdivia et al. 1995). β-Adrenergic receptor-mediated phosphorylation of PLB stimulates the activity of the SR Ca²⁺ pump, resulting in rapid translocation of Ca²⁺ from the cytosol to the SR (Tada et al. 1982; Lindemann et al. 1983). Thus, after perfusion with isoprenaline, the increase in amplitude of the Ca²⁺ transient and decreases in relaxation time and decay time of the intracellular Ca²⁺ transients in wild-type and transgenic myocytes are predominantly attributable to phosphorylation of L-type Ca²⁺ channels, ryanodine-sensitive Ca²⁺ release channels and PLB. However, expression of ssTnI in ventricular myocytes resulted in a greater extent of shortening and an increased re-lengthening half-time of transgenic compared with wild-type myocytes under control conditions and after perfusion with isoprenaline. This result is consistent with the idea that ssTnI isoform expression increases myofilament responsiveness to Ca²⁺.

Our results provide strong evidence that phosphorylation of cTnI plays a significant role in the relaxant effect of β-adrenergic receptor stimulation on the heart. However, initial studies with the PLB-deficient mouse (PLB-KO) concluded that phosphorylation of PLB was fully responsible for the relaxant effect (Luo et al. 1994). Further studies of the PLB-KO hearts together with data reported here have made it clear that both cTnI and PLB are important in enhancing relaxation during β-adrenergic receptor stimulation. The role of cTnI was inferred from earlier experiments comparing mechanical activity and Ca²⁺ transients of ventricular myocytes isolated from wild-type and PLB-KO hearts in control conditions and during β-adrenergic receptor stimulation (Wolska et al. 1996). Even though relaxation was substantially increased in the PLB-KO ventricular myocytes compared with the controls, stimulation of the PLB-KO cells with isoprenaline further increased the rate of relaxation. The isoprenaline-induced increase in maximum re-lengthening velocity was increased by about 170 % in PLB-deficient cells compared with 445 % in wild-type myocytes. In agreement with this conclusion, DeSantiago et al. (1999) reported that a relaxant effect of isoprenaline stimulation was retained in isometrically contracting papillary muscles isolated from PLB-KO mouse heart. In these experiments, care was taken to match the peak force developed by muscles from the wild-type and PLB-KO hearts. In wild-type muscles, τ (the rate constant of relaxation) was reduced by about 50 % after treatment with isoprenaline. In PLB-KO muscles, τ was reduced by about 20 % after treatment with isoprenaline. Results of these studies on isolated cells and muscles from the PLB-KO mice agree well with echocardiographic data demonstrating that isoprenaline increased the velocity of circumferential shortening and left ventricular fractional shortening in PLB-KO hearts (Hoit et al. 1995). It thus seems likely that the relaxant effect of β-adrenergic receptor stimulation reflects an integrated regulatory mechanism involving phosphorylation of PLB as well as cTnI.

The results of the present study can be interpreted to suggest that TnI isoforms influence TnC Ca²⁺ binding. The greater Ca²⁺ sensitivity of tension, the greater extent of shortening and the greater re-lengthening time of transgenic myocytes suggests that the K_d for Ca²⁺ binding to the low affinity binding site of TnC is reduced in the presence of ssTnI. Since K_d=K_off/K_on, either a decrease in K_off or an increase in K_on could account for these observed results. Because no detectable differences in the shortening half-time or intracellular [Ca²⁺] rise half-time were observed between wild-type and transgenic myocytes, the greater half-time of re-lengthening observed in transgenic myocytes is most consistent with a ssTnI-induced decrease in the rate of Ca²⁺ dissociation from, but not an increased rate of association with, the low affinity binding site of TnC. Alternatively, the data can be interpreted in terms of a two-state kinetic model of cross-bridge interaction (Brenner, 1988) in which steady-state force (P) is described by P=nF(f_app/(f_app+g_app)), where n is the number of cycling cross-bridges, F is the average force per cross-bridge, f_app is the rate constant for the transition from the non-force-generating state to the force-generating state, and g_app is the rate constant for the reverse. An increase in the Ca²⁺ sensitivity of tension as observed in the present study may be due to an increase in the number of cycling cross-bridges, the force per cross-bridge, or the proportion of cross-bridges in the force-generating state as a result of an increase in f_app, a decrease in g_app, or both.

Diastolic dysfunction is important in the pathophysiology of many forms of human heart failure, particularly those associated with cardiac hypertrophy. Our finding that alterations in TnI isoform expression can lead to slowed cardiomyocyte relaxation in vitro and impaired diastolic function in vivo suggests that mutations in the cTnI gene might account for some forms of inherited cardiomyopathy and that re-expression of ssTnI in adult cardiac myocytes might contribute to the diastolic dysfunction seen in some patients with acquired cardiac dysfunction. Of potential interest in this regard, six mutations of the cTnI gene have recently been described among patients with familial hypertrophic cardiomyopathy (Kimura et al. 1997).