INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE CONTRACTION AND relaxation cycle of muscle fibers is controlled by a sequential rise and fall of the cytosolic Ca²⁺ concentration ([Ca²⁺]_i). In this regard, the sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) isoforms of skeletal (9, 14) and cardiac (5, 10, 16) muscle play an important role by sequestering cytosolic Ca²⁺ in intracellular stores from which it can be subsequently released. The prominent role of SERCA in cardiac muscle is emphasized by its involvement in the inotropic response to sympathetic stimulation through the phospholamban regulatory mechanism (18, 28, 35). Furthermore, selective inhibition of SERCA by thapsigargin is followed by reduction of intracellular Ca²⁺ transients, tension development, and relaxation kinetics in cardiac myocytes, without alterations of plasma membrane electrical parameters (24).

The availability of SERCA1 and SERCA2a cDNA clones, encoding the two ATPase isoforms that are specific for skeletal and cardiac muscle, respectively (4, 22, 27, 30, 32, 37), has rendered possible their expression in COS-1 cells for mutational analysis (17, 29). Most importantly, initial reports indicate that contractile parameters of rat cardiac myocytes may be influenced by overexpression of SERCA2a ATPase by gene transfer in cultured myocytes (13) or by whole mouse transgenic procedures (11, 15). It is noteworthy, in this regard, that various transfection methods differ widely in their ability to affect a significant number of cells in culture. Furthermore, transfection constructs containing viral promoters override specific transcriptional controls and are constitutively effective not only in myocytes, but also in other cell types.

We considered that, in attempts to influence Ca²⁺ homeostasis or other functions by gene transfer into heterogeneous cell populations or whole muscle, it is desirable to achieve effective transfection of the majority of muscle cells and only of muscle cells. Therefore, with the experiments reported here, we have evaluated various methods of gene transfer in cell cultures, using viral and muscle-specific promoters. We investigated whether these promoters retain exclusive transcriptional control of the ATPase gene, independent of intrinsic adenovirus promoters, and compared their efficiency in control of reporter gene and ATPase gene expression. For this purpose, we used LacZ, enhanced green fluorescence protein (EGFP), or avian SERCA1 cDNA, under control of the constitutive cytomegalovirus (CMV) promoter or the cardiac muscle-specific cardiac troponin T (cTnT) promoter (31) for transfection of chick embryo myocytes and fibroblasts in culture. We evaluated the percentage of cells effectively transfected, the extent of preferential expression in myocytes over fibroblasts, the intracellular membrane targeting of the transgenic ATPase, the yield of Ca²⁺ transport activity in cell homogenates, and the effect on [Ca²⁺]_i transients and contractile dynamics in intact myocytes.

	METHODS

Top Abstract Introduction Methods Results Discussion References

DNA constructs and vectors. Chicken fast-twitch muscle SERCA1 (22) cDNA was initially placed in the pUC19 plasmid for amplification and then subcloned into the shuttle plasmid pE1sp1A (Microbix Systems). In the final constructs, the cDNA was preceded by the constitutive CMV promoter or by the cTnT (31) muscle-specific promoter and was followed by a simian virus 40 polyadenylation signal. LacZ (-galactosidase) and EGFP reporter genes, obtained from Clontech (Palo Alto, CA), were also subcloned into the pE1sp1A shuttle plasmid. The shuttle plasmids were either used directly for transfections of myocytes and fibroblasts or alternatively for cotransfection of HEK-293 cells in conjunction with the replication-defective viral plasmid pJM17 (Microbix Systems) to obtain recombinant adenovirus vectors (12). The shuttle vector was constructed such that homologous recombination resulted in antisense direction of the gene of interest with respect to the adenovirus E1 gene promoter. The recombinant products were plaque and band purified, yielding concentrations in the order of 10¹⁰ plaque-forming units (PFU)/ml.

Cell cultures. Primary cultures of cardiac myocytes were obtained from pooled hearts of day 8 chicken embryos, which were first placed in cold heart medium [500 ml medium 199 (M199) plus Earle's balanced salts, 25 ml fetal bovine serum (FBS), 5 ml penicillin-streptomycin, and 5 ml Fungizone]. After we had removed atria and pericardial membranes with the aid of a dissecting microscope and gently teased the muscle tissue apart, the fragments obtained from 20 to 40 hearts were washed in Hanks' buffered salt solution and then subjected to digestion in 5.0 ml of trypsin solution (0.05 g trypsin, 0.2 g EDTA, 1 g glucose, 0.58 g NaHCO₃, and 4.5 mg/l phenol red) stirred with a magnetic bar at room temperature. After 10 min of digestion, the medium was discarded and the muscle fragments were then subjected to six consecutive trypsinizations of 10 min each. At the end of each trypsinization, the supernatant was collected and added to an equal volume of cold heart medium to prevent further trypsinization of the collected cells. The pooled supernatants were then centrifuged for 5 min at 2,500 g in a refrigerated centrifuge. The sedimented cells were resuspended in 10 ml of heart medium and preplated for 1 h in a 100-mm culture dish at 37°C in 5% CO₂. One hour after preplating on uncoated dishes (for selected attachment of fibroblasts), the detached myocytes were collected in 10 ml heart medium. This suspension was diluted again to plate ~2 × 10⁶ cells/35-mm dish on collagen-coated dishes. Twenty-four hours after the initial plating, detached myocytes were removed by changing the medium. The remaining 30-40% confluent cultures (mostly myocytes and a few remaining fibroblasts) were used for transfections. Sterile conditions were maintained as much as possible throughout these procedures.

Transfections. Transfections were carried out on cell cultures (30-40% confluent fibroblasts or nearly confluent myocytes) by calcium phosphate (6) or liposome (PerFect transfection kit, Invitrogen) methods. Alternatively, the adenovirus-polylysine component method (7, 36), based on physical aggregation of adenovirus, polylysine, and transfection plasmid, was used as described by Kohout et al. (26). For this purpose we used replication-defective adenovirus type 5 mutant Ad5dl312, kindly supplied by Dr. Thomas Shenk (19, 20), propagated in HEK-293 cells, and purified by CsCl density gradient centrifugation before mixing with polylysine and transfection plasmid.

Immunostaining. The lawns of cultured cells were first washed with PBS and then fixed with 4% formaldehyde for 10 min. After repeated washings with PBS, blocking was produced by 10 min of incubation with 1% serum albumin and 0.5% lysine in PBS, followed by a 45-min incubation with the primary antibody at a concentration of 5-10 µg/ml of PBS containing 1% albumin, 0.5% lysine, and 0.25% saponin (permeabilization medium). After being washed with PBS, the cells were incubated for 45 min with biotinylated anti-mouse secondary antibody (Vector Laboratory, Burlingame, CA) at a concentration of 5 µg/ml permeabilization medium. The cells were then washed with PBS and incubated for 20 min with fluorescein streptavidin (Amersham) at a concentration of 5 µg/ml permeabilization medium. The sample was then washed again with PBS, 70% ethanol, and 90% ethanol, allowed to dry, and processed for fluorescence microscopy using a Zeiss Axiophot microscope equipped with a mercury vapor lamp, excitation filters, and digital video acquisition.

Cell homogenates, protein determinations, and Western blots. The myocytes on a 100-mm culture dish were rinsed with 10 ml PBS and collected by scraping with a spatula in 10 ml of a cold medium containing 50 mM 3-(N-morpholino)propanesulfonic acid (MOPS; pH 7.0), 10 mM NaF, 1 mM EDTA, 0.3 M sucrose, 0.4 mM Pefabloc, 10 µg/ml aprotinin, 2 µg/ml leupeptin, and 1 µg/ml pepstatin A. The cells were then sedimented by centrifugation at 2,500 g for 2 min, resuspended in 1 ml of the same medium, frozen in liquid nitrogen, and stored at 70°C. Within 2 wk of storage, the frozen cells were thawed and homogenized with 80 strokes of a hand-held homogenizer immediately before their use for Ca²⁺ uptake measurements. The total protein concentration was determined by measurements of ultraviolet absorption (280 nm) in 0.1% sodium dodecyl sulfate (SDS), using bovine serum albumin as a standard. Samples were also subjected to SDS gel electrophoresis for determination of ATPase by Western blots. For these experiments, myocytes were collected using PBS containing 1 mM EDTA and the protease inhibitors indicated above. The cells were then pelleted, resuspended in the same solution, and homogenized by sonication. The protein concentration of the homogenates was determined by the bicinchoninic acid assay method (Pierce kit), and SDS was added (1%). The ATPase was then resolved in SDS gels, transferred to nitrocellulose membranes, and probed with monoclonal antibodies specifically reactive to the chicken SERCA1 (CaF3-5C3; Ref. 22) or to the chicken SERCA2a ATPase (CaS-3H2; Ref. 21). The secondary antibody was goat anti-mouse horseradish peroxidase-conjugated (Bio-Rad), and the reactive bands were detected using the enhanced chemiluminescence Western blotting kit (Amersham).

Ca²⁺ transport in cell homogenates. The Ca²⁺ uptake medium contained 40.0 mM MOPS (pH 7.0), 100.0 mM KCl, 5.0 mM MgCl₂, 5.0 mM NaN₃, 0.2 mM ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid, 1 µM ruthenium red, 0.2 mM [⁴⁵Ca]CaCl₂, and 100-150 µg/ml cell homogenate protein. The reaction was started by the addition of 5.0 mM potassium oxalate and, after 2 min, 5.0 mM ATP at 37°C. Samples were collected before and, at serial times, after the addition of ATP. The samples (1 ml each) were passed through 0.45-µm Millipore filters, which were washed with 10.0 ml of 2.0 mM LaCl₃ in 10 mM MOPS (pH 7.0), blotted, and placed in scintillation vials for determination of radioactivity.

[Ca²⁺]_i transients and contractility of intact myocytes. Myocytes cultured on glass coverslips were loaded with fluo 3 (Molecular Probes) by incubation for 15 min with 5 µM fluo 3-acetoxymethyl ester from a 442 mM stock in dimethyl sulfoxide and 20% (wt/wt) Pluronic F-127. The cells were then placed in a superfusion bath on the stage of a fluorescence microscope and were superfused (1 ml/min) with buffer containing (in mM) 20 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid, 125 NaCl, 5 KCl, 20 glucose, 0.8 MgSO₄, 1 Na₂PO₄, and 1.8 CaCl₂ (pH 7.4) at 30°C. Cells were field stimulated at 2 Hz using 5-ms pulses with a magnitude of 1.5 times threshold. Fluorescence was measured on a Nikon diaphot microscope using a commercially available fluorescence detection system [Photon Technology International (PTI), South Brunswick, NJ]. A 75-W xenon lamp was used as the excitation source, and the excitation wavelength (488 nm) was selected with a monochromator and a 510-nm dichroic long band-pass (DCLP) mirror. Emission (510-610 nm) was collected with a 610-nm DCLP mirror mounted at a 45° angle to the photomultiplier tube. Fluo 3 emission was digitized and collected at 200 Hz using OSCAR software (PTI). The resulting [Ca²⁺]_i transients were reported as fluorescence emission following stimulation, relative to fluorescence emission at rest. In parallel experiments, the myocyte shortening dynamics were recorded by video microscopy.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Vectors and transfection efficiency. In preliminary experiments, we evaluated various procedures for gene transfer, including physical aggregation with calcium phosphate precipitates, liposomes, or adenovirus-polylysine aggregates, and compared these methods with recombinant adenovirus vectors. The number of cells expressing the transfected gene was evaluated by direct microscopic visualization of intrinsic fluorescence or following incubation with chromogenic substrates or immunofluorescent staining. Consistent with previous reports (23, 25), we found that the recombinant adenovirus is a highly efficient vector. An example of our quantitative evaluation of transfection efficiency is shown in Table 1, where fluorescent cell counts as well as total fluorescence levels following infection with recombinant CMV-EGFP-adenovirus are reported. It is apparent that the percentage of effectively transfected cells increases as the adenovirus PFU level is raised. In fact, the percentage increases steeply as the PFU level is raised from 0.08 to 0.8 PFU/cell, and then reaches an asymptotic level near 100% as the PFU level is raised further. On the other hand, the total fluorescence continues to increase in proportion to the PFU level, likely due to a higher number of gene copies introduced in each cell.

Expression of transgenic SERCA1 ATPase. For the experiments on transgenic ATPase expression, we made two recombinant adenovirus constructs containing SERCA1 cDNA inserts that were placed under the control of either the muscle-specific cTnT promoter or the constitutive CMV promoter and were followed by an identical polyadenylation signal. An advantage of transfections with SERCA1 cDNA is that expression of the skeletal ATPase isoform in cardiac myocytes can be monitored with the monoclonal antibody CaF3-5C3 (22), which does not react with the endogenous SERCA2a enzyme. We then found that the recombinant SERCA1-adenovirus vector is highly efficient and yields ATPase expression, under control of either promoter, in the great majority of cardiac myocytes in culture (Fig. 1). Expression of SERCA1 ATPase was also demonstrated by Western blots obtained with the SERCA1-specific CaF3-5C3 monoclonal antibody (Fig. 2).

View larger version (116K): [in this window] [in a new window]   Fig. 1.   ATPase expression in cardiac myocytes transfected with sarco(endo) plasmic reticulum Ca2+-ATPase (SERCA) 1 cDNA. Chicken embryo cardiac myocytes were transfected with SERCA1 cDNA under control of cardiac troponin T (cTnT) promoter by means of recombinant adenovirus (rec-Adv) vector. Myocytes were processed for immunofluorescent staining 48 h after transfection, using SERCA1-specific antibody CaF3-5C3 (22). Left: control cells that were stained following a transfection procedure without SERCA1 cDNA. Right: note high percentage of transfected myocytes showing SERCA1 expression. Magnification: ×100.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Western blots showing cTnT-SERCA1 and cytomegalovirus promoter (CMV)-SERCA1 expression in chick embryo fibroblasts (A) and cardiac myocytes (B). Cultures of chicken embryo skin fibroblasts or cardiac myocytes were infected with either cTnT-SERCA1 [lanes 3 and 4; 50 plaque-forming units (PFU)/cell] or with CMV-SERCA1 (lanes 5 and 6; 10 PFU/cell) recombinant adenovirus or were subjected to a sham transfection procedure (lanes 1 and 2; no cDNA). Cells were harvested, and samples from whole homogenates of fibroblasts or myocytes were processed for Western blotting using a SERCA1-specific monoclonal antibody. Note occurrence of CMV-SERCA1 expression in both fibroblasts and cardiac myocytes. Note also occurrence of cTnT-SERCA1 expression only in cardiac myocytes.

Cell specificity and efficiency of the cTnT promoter. The advantage of the cTnT promoter is its cell specificity (2, 31). In comparative experiments with chick embryo cardiac myocytes and chick embryo skin fibroblasts, we detected, by Western blots or microscopy, no transgenic expression in fibroblasts under control of the muscle-specific cTnT promoter, while obtaining high expression in cardiac myocytes with the same promoter (Figs. 2 and 3). In addition to demonstrating the cell specificity of the cTnT promoter, the lack of expression in fibroblasts indicates that the transfected gene is not influenced by intrinsic promoters of the recombinant adenovirus. It should be pointed out that ATPase expression is obtained in both myocytes and fibroblasts when the transfected gene is placed under control of the constitutive CMV promoter (Figs. 2 and 3). We found that SERCA1 expression under control of the CMV promoter in fibroblasts was only 65 ± 5% of that in myocytes, when the same adenovirus vector was used at equal PFU titer.

View larger version (93K): [in this window] [in a new window]   Fig. 3.   Lack of SERCA1 expression in fibroblasts following gene transfer under control of cTnT promoter. Cultures of chicken embryo fibroblasts were infected either with CMV-LacZ (left) or cTnT-SERCA1 (right) recombinant adenovirus. Cells were stained 48 h after gene transfer, either for -galactosidase activity (left) or for reactivity to SERCA1 monoclonal antibodies (middle and right). Left: number of fibroblasts in culture and their sensitivity to recombinant adenovirus infection are shown. Right: no SERCA1 expression occurs in fibroblasts under control of cTnT promoter. In these experiments, specificity of promoters was identical whether LacZ or SERCA1 gene was used. Magnification: ×100.

Intracellular targeting of transgenic ATPase expression. Immunofluorescent micrographs of myocytes expressing SERCA1 following transfection by the liposome or recombinant adenovirus methods (Fig. 4) are consistent with transgenic SERCA1 targeting to intracellular membranes, independent of the transfection procedure. We extended our experimentation to clarify whether the membrane targeting of transgenic SERCA1 isoform expression is the same as that of the endogenous SERCA2a ATPase. To this aim, we subjected transfected cells to homogenization and differential centrifugation and then obtained Western blots of the subcellular fractions, staining the same samples in parallel with the monoclonal antibody CaF3-5C3, which is specific for chicken SERCA1 ATPase (22), and with the monoclonal antibody CaS-3H2, which is specific for the chicken endogenous SERCA2a ATPase (4). It is shown in Fig. 5 that the distribution of immunofluorescent label among various subcellular fractions is identical for the transgenic and endogenous ATPases following infection with cTnT-SERCA1 adenovirus. Both endogenous and transgenic ATPases are prevalently associated with the microsomal fraction (i.e., sarcoendoplasmic reticulum).

View larger version (108K): [in this window] [in a new window]   Fig. 4.   Intracellular membrane targeting of ATPase expression in cardiac myocytes transfected with SERCA1 cDNA. Chicken embryo cardiac myocytes were transfected with SERCA1 cDNA under control of cTnT promoter, either by means of liposome aggregates (left) or recombinant adenovirus vector (right). Cells were processed for immunofluorescent staining 48 h after transfection, using CaF3-5C3 SERCA1-specific antibody (22). Magnification: ×400.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Western blots showing localization of transgenic cTnT-SERCA1 and endogenous SERCA2a (endoSERCA2a) in subcellular fractions of transfected cardiac myocytes. Cultures of chicken embryo cardiac myocytes were infected with cTnT-SERCA1 recombinant adenovirus. Two days after infection, cells were harvested, homogenized, and subjected to differential centrifugations, and subcellular fractions were collected. Preparations were processed for Western blotting, and same samples were probed in parallel with monoclonal antibodies specific for SERCA1 (CaF3-5C3) and SERCA2a (CaS-3H2) isoform. Lanes 1 and 5: whole homogenate; lanes 2 and 6: first fraction (cell membranes and nuclei); lanes 3 and 7: second fraction (mitochondria); and lanes 4 and 8: third fraction (microsomes). Lanes 1, 2, 3, and 4 were probed with CaF3-5C3, whereas lanes 5, 6, 7, and 8 were tested with CaS-3H2. Note that transgenic ATPase is prevalently associated with microsomal fraction (i.e., sarcoendoplasmic reticulum), colocalizing with endogenous SERCA2a.

ATP-dependent Ca²⁺ uptake. Active transport of Ca²⁺ by SERCA can be assessed by the use of cardiac muscle homogenates in a reaction mixture containing radioactive calcium isotope and ATP. We found that homogenates of chick embryo cardiac culture sustain ATP-dependent Ca²⁺ uptake with an average initial velocity of 3.7 nmol Ca²⁺ · mg protein¹ · min¹. The activity is totally inhibited by 1 micromolar thapsigargin (Fig. 6), which is a highly specific inhibitor of SERCA (33).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Ca2+ uptake by homogenates of control and SERCA1 transgenic myocytes. Cultures of cardiac myocytes were infected with cTnT-SERCA1 (+) or CMV-SERCA1 (, ) recombinant adenovirus, with PFU levels of 100 (+), 3 (), or 10 () PFU/seeded cell. Myocytes were collected for measurements of ATP-dependent Ca2+ uptake 48 h after infection. Control cultures () were subjected to gene transfer procedure without cDNA. Note that Ca2+ uptake was totally inhibited by 1 µM thapsigargin (), which is a specific inhibitor of SERCA enzymes. Uptake shown was not corrected to exclude contribution of fibroblasts. Western blots (inset) show levels of transgenic SERCA1 and endogenous SERCA2a expression in control myocytes (lane 1), in cells infected with cTnT-SERCA1 adenovirus (lanes 2 and 3), and in cells infected with CMV-SERCA1 adenovirus at lower (lane 4) and higher (lane 5) PFU titer.

1

1

Effects of SERCA1 transfection on contractile behavior and [Ca²⁺]_i transients of intact myocytes. A series of experiments was performed to determine whether the increase in Ca²⁺ transport in vesicles isolated from SERCA1-transfected myocytes is reflected in changes of contractile dynamics and/or Ca²⁺-handling properties of intact cells. As shown in Fig. 7A, transfected cells display dramatically shortened twitches, due to a reduction of both tension development and relaxation times. In fact, waveform analysis demonstrated a reduction of the width at half height from 223 ± 10 ms (n = 24) for control cells to 160 ± 13 ms (n = 15) for cTnT-SERCA1 transfected cells. Similar effects were noted on the [Ca²⁺]_i transients (Fig. 7B), as the first order time constant of the decay phase was decreased by 40%, from 190 ± 18 ms (n = 27) in control to 113 ± 13 ms (n = 15) in cTnT-SERCA1 transfected cells. Similar results were obtained with transgenic expression of SERCA1 under control of the CMV promoter. These effects, originally observed by Hajjar et al. (13) following transfection of neonatal rat myocytes with heterologous SERCA2a under control of a viral promoter, cannot be related quantitatively to transgene expression as accurately as the transport measurements described above. Nevertheless, the results shown in Fig. 7 demonstrate that expression of transgenic ATPase under control of the cell-specific cTnT promoter has a strong influence on the [Ca²⁺]_i transients in situ, just as transgenic expression does when under control of a viral promoter.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Contractility (A) and cytosolic Ca2+ ([Ca2+]i) transients (B) in control and SERCA1-transfected cultured myocytes. A: cultured myocytes were field stimulated at 2 Hz, and contractile behavior was recorded by video microscopy. B: fluo 3-loaded cultured myocytes were field stimulated, and resulting [Ca2+]i transients were recorded and analyzed as described in METHODS. Normalized signal averages from 3 cells are shown for each curve. See text for statistical analysis.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Consistent with previous reports (1, 23, 25, 34), our experiments demonstrate unambiguously that recombinant adenovirus is a very efficient vector for gene transfer into myocytes and fibroblasts, yielding transgene expression in nearly all cells exposed when the multiplicity of infection is optimized. The efficiency of recombinant adenovirus is much higher than that obtained with methods based on aggregation of transfection plasmids with calcium phosphate, liposome, or adenovirus-polylysine complex.

Independent of the transfection vector, our findings contribute to characterization of the cTnT promoter. We used the 268-base pair (bp) segment (Fig. 8) of the chicken cTnT promoter (31), which includes tandem M-CAT, "CarG," "MEF-1," and SP1 motifs in the proximal region (129 to 49 bp), and a cardiac element in the distal region (269 to 201 bp). Similar motifs are also present in a proximal (284 to 72 bp) and a distal (1810 to 1110 bp) segment of the SERCA2 promoter (3). We used the cTnT promoter for its compact size and very strong specificity. In fact, in our experiments, the 268-bp segment of the cTnT promoter proved to be highly specific for myocytes, with no appreciable leak in fibroblasts. Although the 268-bp segment of the cTnT promoter is significantly weaker than the CMV promoter, we were able to obtain similar levels of functional ATPase expression by adjusting the PFU levels of recombinant adenovirus vectors containing the two different promoters.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Functional elements of 268-bp segment of cardiac troponin T (cTnT) promoter. Elements involved in transcription regulation are underlined (31).

It is noteworthy that previous studies of this cell-specific promoter were performed with reporter genes by transfection methods involving a small percentage of cells in heterogeneous cultures. In our experiments, we have extended the characterization of a short segment (268 bp) of the cTnT promoter by the use of an isomorphic endogenous gene that requires specific intracellular targeting for its function. We have also used a recombinant adenovirus vector that mediates gene transfer into the majority of cells in culture and demonstrated that in the recombinant virus the gene remains under exclusive control of the cell-specific promoter and is not influenced by intrinsic viral promoters.

Functionally, we obtained a fourfold increase in ATPdependent calcium uptake over endogenous SERCA ATPase levels in chick cardiac embryonic myocytes, after SERCA1 gene transfer using recombinant adenovirus vectors under control of the cell-specific promoter. This is quite a bit higher than that obtained previously by means of transgenic expression of heterologous SERCA2a under control of constitutive viral promoters in cultured myocytes of neonatal rats (11, 13) and in transgenic mouse hearts (15). Most importantly, our experiments suggest that there is an upper limit for the ability of the myocytes to express functional SERCA protein, a limit that can be reached either under control of the cell-specific or the constitutive viral promoter. Finally, we find that transgenic SERCA1 expression under control of the cTnT promoter affects contractile dynamics and [Ca²⁺]_i transients in situ just as much as transgenic expression under control of the viral promoter.

Our findings raise the possibility of manipulating Ca²⁺ homeostasis and Ca²⁺-dependent functions specifically in cardiac myocytes within heterogeneous cell populations by means of gene transfer under control of cell-specific promoters. Furthermore, our observations may be helpful in designing suitable constructs for cell-specific transgenic targeting in whole cardiac muscle by means of recombinant adenovirus vectors (8).