The Na+/Ca2+ exchanger (NCX) is proposed to be an important protein in the regulation of Ca2+ movements in the heart. This Ca2+ regulatory action is thought to modulate contractile activity in the heart under normal physiological conditions and may contribute to the Ca2+ overload that occurs during ischemic reperfusion challenge. To evaluate these hypotheses, adult rat cardiomyocytes were exposed to an adenovirus that codes for short hairpin RNA (shRNA) targeting NCX gene expression through RNA interference. An adenovirus transcribing a short RNA with a scrambled nucleotide sequence was compared with the NCX-shRNA nucleotide sequence and used as a control. Freshly isolated rat cardiomyocytes were infected with virus for 48 h before examination. Cardiomyocytes maintained their characteristic morphological appearance during this short time period after isolation. NCX expression was inhibited by up to ∼60% by the shRNA treatment as determined by Western blot analysis. The depletion in NCX protein was accompanied by a significant depression of NCX activity in shRNA-treated cells. Ca2+ homeostasis was unaltered in the shRNA-treated cells upon electrical stimulation compared with control cells. However, when cardiomyocytes were exposed to a simulated ischemic solution, NCX-depleted cells were significantly protected from the rise in cytoplasmic Ca2+ and damage that was detected in control cells during ischemia and reperfusion. Our data support the role for NCX in ischemic injury to the heart and demonstrate the usefulness of altering gene expression with an adenoviral-delivery system of shRNA in adult cardiomyocytes.
- short hairpin ribonucleic acid
the na+/ca2+ exchanger (NCX) is an important protein in the regulation of intracellular Ca2+ concentration and excitation-contraction coupling in the heart (2, 3, 17). The forward mode of the exchanger induces cardiac relaxation by removing Ca2+ from the cell in exchange for extracellular Na+ (2, 3, 17). In the reverse mode, the NCX contributes to cardiac contraction by bringing Ca2+ into the cell in exchange for intracellular Na+ (2, 3, 17). The NCX has also been associated with pathophysiological processes like ischemia-reperfusion injury, glycoside toxicity, cardiac hypertrophy, and heart failure (2, 4, 5, 8, 10, 11, 17, 20). The NCX has been identified, therefore, as a potentially important therapeutic target for the treatment of heart disease (12).
The role of the NCX in these processes has been determined mostly through the use of inhibitory peptides and NCX-selective drugs to block its activity (12, 14). More recently, a variety of molecular tools have been used to alter NCX gene expression and thereby alter NCX function. These include the use of transgenic mice, antisense cDNA, RNA interference (RNAi), and adenoviruses to increase or decrease the expression of NCX protein in the heart cell (1, 4, 7–11, 18, 19).
Our laboratory has employed the technique of RNAi to decrease the amount of expressed NCX protein in rat neonatal cardiomyocytes (1, 7–9). RNAi is one of the most effective methods available to inhibit NCX expression and activity (7, 9). Despite ∼95% depletion of NCX protein in the neonatal cardiomyocytes, only a few alterations in their contractile activity were observed (7). Similar conclusions had been reached earlier from data obtained from knockout mice depleted of cardiac NCX by ∼80% (18). This has brought into question the critical role of NCX in excitation-contraction coupling (7). However, the role of the NCX in ischemic-reperfusion injury does not appear to be in question. Overexpression of NCX in cardiomyocytes has resulted in augmented damage when cells were subjected to an ischemic insult (4, 8), and, conversely, cells survived better when the NCX expression was depressed (8, 10, 11, 18).
Unfortunately, some of these experiments (7, 8) have been conducted in neonatal cardiomyocytes where it is easier to manipulate gene expression than in adult cells. Adult and neonatal cardiomyocytes differ significantly in their excitation-contraction coupling process, activity, and expression levels of NCX (21–23) and in their response to ischemic reperfusion challenge (16). Other than the mouse transgenic data, there are no data on contractile performance of adult cardiomyocytes when NCX expression is inhibited. Although genetically modified mice provide excellent models to examine the effects of changes in NCX expression, the use of single cells can avoid potential influences from extracardiac factors (hemodynamic, hormonal, etc.). There are also no data presently available on the ability of these NCX-depleted adult cardiomyocytes to survive an ischemic insult. In the present study, we have applied an adenovirus to deliver short hairpin RNA (shRNA) to adult rat cardiomyocytes to decrease the amount of NCX expression. The objectives of the present study, therefore, were to: 1) determine if the shRNA methodology could be successfully employed to alter NCX expression in adult cardiomyocytes, 2) characterize the effect of inhibition of NCX expression by shRNA on cardiomyocyte contractile function, and 3) examine characteristics of ischemia-reperfusion injury in adult rat cardiomyocytes depleted of NCX protein by RNAi.
MATERIALS AND METHODS
This work was approved by the Animal Care Protocol Committee from the University of Manitoba. Guidelines for the care and treatment of animals from the Canadian Council on Animal Care were strictly followed.
Isolating and maintaining the cardiomyocytes.
Single cardiomyocytes were isolated from the hearts of adult Sprague-Dawley rats (weighing ∼150–250 g) by enzymatic digestion of cardiac tissue as described previously (13). Briefly, the heart was removed from the rat, cannulated, and perfused in a Langendorff perfusion mode for 2 min with a Tyrode buffer (TB) containing (in mM) 140 NaCl, 5.4 KCl, 1.0 Ca2+, 1.0 MgCl2, 0.33 NaH2PO4, 5.0 HEPES, and 10 d-glucose and 0.05% BSA Fraction V at pH 7.4 and bubbled with 100% O2 at 37°C. The buffer was then switched to a Ca2+-free TB for 2 min followed by ∼10 min with TB containing 0.5 mg/ml collagenase type 2 (Worthington). The heart was placed in cold KB buffer containing (in mM) 70 KOH, 40 KCl, 50 glutamate, 20 KH2PO4, 3 MgCl2, 10 d-glucose, 1.0 EGTA, and 10 HEPES (pH 7.4) and was scissor minced. Cells were separated with a disposable pipette and filtered through a single layer of gauze in a sterile centrifuge tube. Cells were allowed to settle for 15 min, after which the supernatant was poured off, and the cells were once again resuspended in fresh KB buffer.
Isolated cells were plated on laminin-coated 25-mm cover slips in a 35-mm culture dish. Unattached cells were washed off, and the remaining cells were incubated overnight at 37°C with 5% CO2 in medium 199 (Invitrogen) containing penicillin (50 U/ml), streptomycin (50 U/ml), and 1% of antibiotic/antimycotic solution (Invitrogen).
Construction of adenovirus vectors and infection of cardiomyocytes.
Three different adenoviruses to deliver shRNAs targeting different regions of the NCX sequence and one scrambled control were generated previously in our laboratory (7–9). The adenovirus with the RNAi sequence that inhibited NCX1 expression most effectively (7) was chosen for the experiments described in this study. The characteristics of the adenoviruses and the vectors are found in detail elsewhere (7–9). An adenovirus carrying the Enhanced Green Fluorescent Protein (EGFP) was used to assess the degree of infection. The cardiomyocytes were infected the morning after isolation with adenoviruses that carried the shRNA sequence for cardiac NCX1 (or a scrambled sequence for control) and with an EGFP-expressing adenovirus. Cardiomyocytes were infected with 1 × 107 and 5 × 107 plaque-forming units (PFU) /ml of adenovirus and incubated for 48 h in culture medium at 370C, 5% CO2 before the experiment. After 48 h of incubation, cover slips with infected and control cardiomyocytes were mounted in a Leiden chamber with a Medical Systems PDMI-2 Open Perfusion Micro-Incubator (Greenvale, NY) at 37°C. Cells were perfused with a HEPES-buffered solution bubbled with 100% oxygen. The control perfusion buffer contained (in mM) 140 NaCl, 6 KCl, 1 MgCl2, 1.25 CaCl2, 10 d-d-glucose, and 6 HEPES (pH 7.4). Cells were perfused and paced at 0.5 Hz with a duration of 200 ms using platinum electrodes for all experimental procedures.
Measurement of cellular Ca2+.
The Ca2+-sensitive dye fura 2-AM (Molecular Probes, Eugene, OR) was used as an intracellular indicator of Ca2+. Myocytes adherent to laminin-coated glass cover slips were loaded for 15 min at 37°C and washed before experiments. After loading, the cells were allowed to de-esterify for 30 min at room temperature in the dark. These cells were then placed in a Leiden chamber that was mounted on a Nikon Diaphot microscope with a ×40 epifluorescent objective. Cells were excited at 340 and 380 nm with an emission of 505 nm. The fluorescence was collected on a PTI spectrofluorometer and quantitated ratiometrically, as described in detail previously (7, 13). Calcium changes occurring were measured as the difference of the 340/380 ratio from basal levels. All data were obtained from individual cells, and the means ± SE were determined for each experimental group.
Measurement of NCX activity with LiCl method.
Cardiomyocytes were perfused at 1.0 ml/min with a control TB containing (in mM) 140 NaCl, 6.0 KCl, 1.25 Ca2+, 1.0 MgCl2, 5.0 HEPES, and 10 d-glucose bubbled with 100% O2 at 37°C. Using a perfusion head allowing for the rapid change of perfusion buffer (<1 s), cells were then perfused with an identical TB substituting LiCl for NaCl (pH 7.4) at 37°C until a maximal increase in Ca2+ levels occurred (∼1 min). After reaching a plateau in Ca2+, the buffer was rapidly switched to a normal Tyrode exchanging Na+ for Li+ (8).
Perfusion of cardiomyocytes with ouabain.
Cardiomyocytes were allowed to equilibrate in control TB before experimentation. Cells were perfused at 37°C and stimulated at 0.5 Hz. Cardiomyocytes were perfused with buffer containing 1 or 5 mM ouabain for 10 min followed by a washout period of 10 min (8). Ca2+ was measured as the change in the 340/380 ratio over time.
Cardiomyocytes were perfused with a control TB at 37°C and stimulated at 0.5 Hz. Ischemia was initiated as previously described in detail (13) for a period of 45 min followed by a reperfusion of 60 min. Briefly, a single electrically stimulated beating cell was randomly chosen, perfused at 1 ml/min, and equilibrated in control HEPES-buffered solution containing (in mM) 140 NaCl, 6 KCl, 1 MgCl2, 1.25 CaCl2, 6 HEPES (pH 7.4), and 10 d-glucose at 37°C and bubbled with 100% oxygen. The ischemic buffer was similar in composition except KCl was 8 mM, pH was 6.0, d-glucose was absent, and it was bubbled with 100% nitrogen. Cell viability was obtained by counting cells, after the ischemic event, in various random fields and representing the data as a percentage of total counted cells, as previously described (13). This simple method of evaluating cell viability closely correlates with enzyme release and fluorescent indicators of cell damage and death (13).
Western blot analysis.
Attached cardiomyocytes were lysed with RIPA buffer (50 mM Tris, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 1 mM EDTA, and 1 mM EGTA, Protease Inhibitor Cocktail with 1 mM phenylmethylsulfonyl fluoride, and 1 mM benzamidine, pH 7.5) and collected for Western blot analysis. Samples (50 μg of total protein/lane) in Laemmli sample buffer were resolved on 7% SDS-PAGE. Proteins were transferred to nitrocellulose membrane in a wet transfer apparatus overnight at 4°C at 35 volts. Anti-NCX (R3F1 monoclonal, 1:1,000 dilution; Swant) primary antibody was used to detect NCX level, and anti-actin A2066 (1:2,000 dilution; Sigma) primary antibody was used to detect total actin as a loading control. The signal from horseradish peroxidase-conjugated secondary antibody was developed using West Pico chemiluminescence substrate (Pierce). The signal was collected in a Bio-Rad detection system and quantified by densitometry analysis using Quantity One software (Bio-Rad) (7).
Cells on cover slips were fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and blocked with 2% milk-0.1% Triton X-100 in PBS. R3F1 primary antibody to NCX (1:150 dilution) was followed by Alexa 488-conjugated goat anti-mouse secondary antibody (1:700 dilution; Molecular Probes). Cells were mounted on glass slides using mounting medium with DAPI (Vectashield; Vector Laboratories). All images were collected with a Nikon fluorescent microscope (Nikon Eclipse TE2000S) (7), and the value of fluorescence was obtained from the total cellular area by converting the number of pixels following a scale of fluorescence into numeric values. Analysis of fluorescence was performed on an SGI workstation using GE Healthcare Image Space 3.2.1 software.
Data are expressed as means ± SE. The data were analyzed by a Student's t-test or a one-way ANOVA. A Student Newman-Keul's post hoc test was used to determine statistical difference after the ANOVA. Statistical significance was set at a P < 0.05.
As shown in Fig. 1, the morphology of the cardiomyocytes infected with adenoviruses carrying a scrambled sequence (AdScr) or shRNA for NCX (AdRNAi) was similar to uninfected control cells after 48 h of infection. This duration of infection was chosen to achieve maximal inhibition of NCX expression without affecting the morphological integrity of the cardiomyocytes when infected at an adenovirus concentration 5 × 107 PFU/ml. Longer incubation times were found to elicit undesirable morphological changes (flattening of the cells), and higher adenovirus titers induced nonspecific toxicity within the cells. Thus all subsequent experiments were carried out under these specific conditions.
NCX expression was measured by cytoimmunofluorescent techniques that provided a complimentary and semiquantitative measurement of NCX expression. As shown in representative images in Fig. 1, the fluorescent signal intensity (% of control) for NCX expression was significantly lower in the RNAi-treated cardiomyocytes compared with uninfected controls and scrambled infected controls (nonbackground subtracted). This depression was quantitated and confirmed in a number of cells (Fig. 1). The immunocytochemistry information provides valuable qualitative data of the changes present, but the Western data are more accurate quantitatively than the immunocytochemistry.
NCX1 expression was more carefully quantitated by Western blots (Fig. 2). The RNAi treatment successfully inhibited NCX expression. Expression of NCX in AdRNAi-infected cells represented only 47.4 ± 10.3% for 1 × 107 PFU/ml and 37.4 ± 3.8% for 5 × 107 PFU/ml compared with the level expressed in AdScr-infected cells.
These experiments demonstrate a successful inhibition of NCX expression was achieved through RNAi treatment but gives no indication of NCX function. To determine NCX function, the LiCl rapid washout technique (1, 9) was used to establish a rapid Na+ gradient and stimulate reverse-mode exchange. As shown in Fig. 3, LiCl treatment of the cells initiated a large increase in diastolic Ca2+ in those cells treated with the scrambled adenovirus. In addition, the time-to-peak Ca2+ was significantly increased by LiCl in the AdScr-infected cells. The response to LiCl in both of these parameters was significantly blunted in the cells treated with the shRNA adenovirus for the NCX.
Inhibition of the Na+ pump with ouabain is another biochemical mechanism to create a transsarcolemmal Na+ gradient and stimulate NCX activity. In control cells, there was a significant increase in diastolic Ca2+ (Fig. 4). This was significantly depressed in cardiomyocytes that had been depleted of NCX protein through RNAi treatment. Both the ouabain and LiCl experiments, therefore, demonstrate that the NCX is active in control cells but RNAi treatment has attenuated its function in addition to depressing NCX expression.
The previous experiments created defined conditions to allow us to measure the function of the exchanger. However, it was also important to examine the capacity of the NCX-depleted cells to regulate intracellular Ca2+ under more physiological conditions. Electrical stimulation of cellular contraction is a useful intervention to identify the importance of the exchanger in regulating Ca2+ homeostasis. Basal Ca2+ measurements were collected in cardiomyocytes paced at 0.5 Hz to enable us to study the role of the exchanger on a beat-to-beat basis. Data from control cells (AdScr) and AdRNAi (1 × 107 PFU/ml and 5 × 107 PFU/ml)-infected cells were compared and summarized in Table 1. There were no significant differences in basal diastolic and systolic Ca2+ values (represented as the 340/380 ratio) among any of the study groups. The amplitude of the transient, measured as the systolic minus the diastolic Ca2+ ratios, was also not statistically different among any of the groups. The time to peak contraction and the time to half relaxation of the transient also exhibited no significant differences.
The response of the cells that were depleted in NCX was examined during exposure to an ischemia-mimetic solution followed by reperfusion. Diastolic Ca2+ and Ca2+ transients were monitored during ischemia and reperfusion. Diastolic Ca2+ increased in response to ischemia and reperfusion. Ca2+ transients were attenuated during ischemia and did not recover completely during reperfusion. After the ischemic challenge, cardiomyocytes that had reduced NCX expression did not exhibit as large a rise in diastolic Ca2+ as AdScr-infected control cells did during the ischemic period or during reperfusion (Fig. 5). Ca2+ transients were not significantly different between the two groups during ischemia or reperfusion.
Cell viability was also monitored following the ischemia-reperfusion insult (Fig. 6). There was a significant increase in the percentage of cells that displayed a normal, healthy morphology and a decrease in the percentage of cells that exhibited a balled up, rounded morphology indicative of cell damage and death in the NCX reduced group of cardiomyocytes compared with the scrambled control group of cells.
Our results have shown that a reduction in NCX expression can be successfully achieved in adult cardiomyocytes through the use of RNAi delivered by an adenoviral approach. The use of adenoviruses to deliver shRNA to inhibit NCX expression has been shown previously by our laboratory to be a very successful method to inhibit NCX expression (1, 7–9). However, this work was achieved using neonatal cardiomyocytes that are easier to maintain in culture. Adult cardiomyocytes are susceptible to morphological alterations when maintained for extended periods in cell culture medium. The effects of the adenovirus-delivered shRNA were limited, therefore, because of the restricted amount of time that the cells could be left in culture medium. We found that 48 h was the maximum length of time that the cells could maintain the obvious block-like, striated morphological appearance that is typical of an adult cardiomyocyte. With a half-time for NCX turnover of 33 h (19), it was clear from the outset that complete inhibition of NCX expression was unlikely to be achieved. However, NCX expression was inhibited ∼60% compared with the scrambled control infection. This degree of depletion of NCX protein over this 48-h time period is consistent with the decay curve plotted previously for NCX (7). Although this was not as efficient as the >95% inhibition that was achieved in neonatal cardiomyocytes (7), it was sufficient to induce a significant depression in NCX function. This provided us with an important opportunity to determine if this inhibition in NCX expression and function would have an impact on the function of the adult cardiomyocytes during physiological stimulation and during ischemic challenge.
The success in inhibiting NCX expression in the adult cardiomyocytes allowed us to specifically assess two responses in the present study: 1) the contribution of NCX to Ca2+ homeostasis in the adult cardiomyocytes when electrically stimulated and 2) the role of the NCX in ischemic injury. With regard to the former objective, our results demonstrate that adult cardiomyocytes that have a depressed NCX expression and activity maintain normal Ca2+ homeostasis upon electrical stimulation of contraction. This may be somewhat surprising in view of the critical role that has been suggested for the NCX in excitation-contraction coupling in the heart (2, 3). However, it is consistent with several recent studies in neonatal cardiomyocytes and transgenic mouse models where severe NCX depletion did not result in a dramatic loss of contractile activity and Ca2+ homeostasis (7, 18). These studies did show changes to the characteristics of cardiac contractile activity and Ca2+ homeostasis (i.e., slowing, etc.) (7, 18), whereas the present results detected no changes at all in any of the parameters monitored. The amount of NCX expression and function remaining in the shRNA-treated adult cardiomyocytes in the present study appeared to be sufficient to maintain normal Ca2+ homeostasis.
With regard to the second objective of the present study, our results demonstrate that adult cardiomyocytes that have a depressed NCX expression and activity resist the Ca2+ overload that occurs during ischemic challenge and reperfusion challenge. This is consistent with a great number of studies that have shown that increased expression leads to greater damage (4) and, conversely, inhibition of NCX by drugs or reduced NCX expression protects the heart/cardiomyocytes from ischemic-reperfusion injury (8, 10–12, 18). In our study, NCX-reduced cells were significantly protected from the rise in cytoplasmic Ca2+ and subsequent damage recorded at the level that was detected in control cells, during ischemia and reperfusion. NCX inhibition during ischemia would theoretically leave intracellular Na+ levels high. Blocking the Na+/H+ exchanger (NHX) or the Na+ channels has been shown to lower intracellular Na+ levels and protects the heart from ischemic-reperfusion injury (6, 15). The advantage of blocking the NCX instead of these two alternative Na+ transport pathways is twofold. First, it likely allows the transarcolemmal H+ gradient to dissipate more quickly because the NHX should still be very active. This would allow the Na+ pump to activate faster as well and reduce the intracellular Na+ concentrations. Second, NCX inhibition is directly blocking the primary cause of ischemic damage, intracellular Ca2+ overload. Our study contributes, therefore, to increasingly persuasive evidence that NCX is extremely important in the Ca2+ overload and damage that accompanies the ischemic-reperfusion insult in the heart. We may now conclude that this protection is afforded to both neonatal cells and adult preparations as well.
This study was supported through operating funds provided by the Heart and Stroke Foundation of Manitoba. Infrastructural support was provided by St. Boniface Hospital and Research Foundation. C. Hurtado was a Trainee of the Heart and Stroke Foundation of Canada.
No conflicts of interest are declared by the authors.
- Copyright © 2010 the American Physiological Society