Mitochondria as Targets for Nitric Oxide–Induced Protection During Simulated Ischemia and Reoxygenation in Isolated Neonatal Cardiomyocytes

Isolation and Culture of Rat Ventricular Cardiomyocytes

Neonatal rat ventricular cardiomyocytes were prepared from 1- to 2-day-old Sprague-Dawley rats as described previously.¹² Briefly, neonatal rat ventricles were dispersed in a series of incubations at 37°C in HEPES-buffered salt solution containing 0.6 mg/mL pancreatin (Life Technologies, Inc) and 0.5 mg/mL collagenase (Worthington). To reduce fibroblast contamination to <5%, dispersed cells were preplated for ≥30 minutes, and the unattached cells were replated on gelatin-coated coverslips. Cells were cultured at 37°C in room air with 5% CO₂ in Dulbecco’s modified Eagle’s medium and Medium 199 (M199, at 4:1) supplemented with 10% horse serum, 5% FCS, and 100 U/mL penicillin/streptomycin for the first 24 hours (Life Technologies). Thereafter, cells were maintained in M199 supplemented with 1% FCS. Under these conditions, >80% of cells beat spontaneously for the duration of the experiment. Experiments were performed between days 2 and 3 in culture.

Materials, Solutions, and Dye Loading Protocols

Ischemia buffer was composed of (in mmol/L) NaCl 118, NaHCO₃ 24, KCl 16, KH₂PO₄ 1, CaCl₂ 2.5, MgCl₂ 1.2, Na⁺ EDTA 0.5, and Na⁺ lactate 20, pH 6.2. Control buffer (CB) had the same composition, except KCl was 4 mmol/L, lactate was replaced by 2 mmol/L Na⁺ pyruvate and 10 mmol/L d-glucose, and pH was 7.4. HEPES-buffered saline (HBS) contained (in mmol/L) NaCl 156, KCl 3, MgSO₄ 2, KH₂PO₄ 1.25, CaCl₂ 2, HEPES 10, and d-glucose 10, with pH adjusted to 7.35 with NaOH. PBS contained (in mmol/L) NaCl 155, Na₂HPO₄ 5, and KH₂PO₄ 1.5. Drugs included SNAP (Calbiochem), carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP; Sigma), sodium cyanide (CN⁻; BDH), caffeine, and thapsigargin (Sigma).

Cells were loaded at room temperature with fluorescent indicator dye (Molecular Probes) after pretreatment with SNAP (unless otherwise indicated) as follows: tetramethylrhodamine ethyl ester (TMRE), 3 μmol/L for 15 minutes. Under these conditions, loss of Δψ_m causes a dequench of fluorescence and an increase in signal. JC-1, 10 μmol/L for 10 minutes at 37°C. JC-1 is a dual-emission indicator; monomers fluoresce at 530 nm, and accumulation in mitochondria results in formation of “J” complexes, which emit at 590 nm.^{13 14} Because monomer fluorescence is enhanced in a lipid environment¹³ and is variable between cells, we used the 590-nm signal only to indicate a loss of Δψ_m. Fura-2, 5 μmol/L of the AM ester for 30 minutes. Rhod-2, 10 μmol/L of the AM ester for 30 minutes at 37°C. Loading was carried out before SNAP pretreatment to allow rhod-2/AM to be partitioned and hydrolyzed in mitochondria. Fluo-3, 5 μmol/L of the AM ester for 30 minutes. After dye loading, cells were washed with HBS.

Experimental Setup for Fluorescence Imaging of Cardiomyocytes

Cardiomyocytes cultured on coverslips were placed into a purpose-built chamber on the stage of an inverted epifluorescence microscope (described in detail by Mojet et al¹⁵ ). Cells were perfused at ≈10 mL/min and 34°C to 36°C. Drugs were applied by pressure ejection with glass pipettes with an inner diameter of ≈2 μm, positioned within 100 μm of the cell. Fluorescence was elicited by illumination with a 100-W xenon arc lamp via a spinning filter wheel (Cairn Research) fitted with appropriate 10-nm bandpass filters. During sI/R experiments, cells were illuminated by opening a shutter for 2 seconds every minute to reduce photodynamic damage. Fluorescence signals were measured by use of 2 photomultiplier tubes after passing through bandpass filters appropriate for the fluorescence indicators used.

Confocal images were obtained on an LSM 510 (Carl Zeiss, Inc). Fluo-3 and rhod-2 were excited with the 488-nm argon laser line, and fluorescence was collected at 505 to 550 nm (fluo-3), at >585 nm (rhod-2), or with both channels (JC-1).

Experimental Protocols

Inhibition of Mitochondrial Energetics Using SNAP

Acute (≈10 seconds) administration of NO (Figure 3⇓) was achieved by pressure-ejecting CB containing 1 mmol/L SNAP. Responses to SNAP were compared with those to the uncoupler FCCP (1 μmol/L) and the respiratory inhibitor CN⁻ (2 mmol/L). Alternatively (Figures 1⇓ and 2⇓ and 4 through 7), cells were pretreated with 1 mmol/L SNAP in CB for 90 minutes at 37°C in the incubator, washed with PBS, and dye-loaded for 10 to 30 minutes in HBS before the sI/R protocol was applied. SNAP at 1 mmol/L gives an effective concentration of 3 to 5 μmol/L (NO electrode measurements in incubation medium samples; for method, see Sharpe and Cooper¹⁶ ). As expected, this is only slightly higher than physiological concentrations of NO found in cardiovascular tissue (1.3 μmol/L¹⁷ and ≈1 μmol/L¹⁸ ). Control cells in the cell viability experiment (Figure 1⇓, vehicle) were pretreated for 90 minutes in CB at 37°C without SNAP.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

SNAP pretreatment reduced sI/R-induced cell death. Percentage cell death was assessed by propidium iodide exclusion assay, as outlined in protocol 2. Data are mean±SEM percentage cell death, with n=24 or 32 (8 fields of view in 3 or 4 experiments, each from different cell preparations).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

SNAP pretreatment prevented sI/R-induced changes in [Ca²⁺]_c. Simultaneous recording of changes in [Ca²⁺]_c and Po₂ during ischemia/reoxygenation. Dotted line indicates control cells; solid line, after SNAP pretreatment. Inset shows return of [Ca²⁺]_c oscillations after reoxygenation in control cells.

Statistical Analysis

Data are expressed as mean±SEM and were obtained from 4 to 7 separate experimental preparations. Where appropriate, data are expressed as percentage of baseline (ie, 100% × induced value/value before stimulation). Statistical significance (P<0.05) was determined with a 2-tailed Student’s t test or Mann-Whitney test (Graphpad Instat version 3.0).

Pharmacological Preconditioning by SNAP in a Cellular Model of sI/R Injury Under Epifluorescent Microscopy

The sI/R protocol resulted in 48.5±2.0% cell death in control cells. Ninety minutes of SNAP pretreatment and 30 minutes of incubation before sI/R was significantly protective (Figure 1⇑), producing a reduction in cell death to 34.5±1.3% (P<0.001, Student’s t test with Welch correction). Cell death in untreated cells exposed to 2 hours of perfusion with CB was 29.7±1.4%. This value agrees well with earlier findings,²² whereas Mackay and Mochly-Rosen²³ reported ≈16% cell death in their cultures before perfusion with basic salt solutions and possible fluorescent dye–related photodynamic damage. Therefore, SNAP pretreatment reduced sI/R-induced cell death to 25.4% of the ischemia control.

Measurement of [Ca²⁺]_c During sI/R

In control cells, simulated ischemia induced a progressive rise in [Ca²⁺]_c (Figure 2⇑). On reoxygenation, [Ca²⁺]_c returned to baseline, and [Ca²⁺]_c oscillations that accompanied spontaneous beating returned. Usually a secondary, exponential rise in fura-2 ratio developed, followed eventually by loss of cell membrane integrity. SNAP pretreatment did not affect the resting [Ca²⁺]_c levels but dramatically reduced the magnitude of the [Ca²⁺]_c increase during ischemia, from 516±170% to 226±40% (n=5, P<0.05). In addition, SNAP pretreatment suppressed the secondary [Ca²⁺]_c increase and cellular integrity was maintained to the end of the experiment. The apparent difference in latency until the start of the anoxia-induced increase in [Ca²⁺]_c was not significant.

Effect of SNAP on Mitochondrial Energetics

Acute Administration

NO-induced inhibition of mitochondrial respiration was assessed in single cells by NADH autofluorescence. A brief (10-second) application of SNAP caused a reproducible and rapidly reversible rise in NADH autofluorescence (106.9±1.9% of baseline, P=0.012, n=7; see Figure 3⇓). The response to SNAP was small compared with the responses to CN⁻ or FCCP, which indicate the full range of changes in NADH redox state (123.8±4.2%, n=6, and 80.1±3.4%, n=9, respectively). Consistent with its modest inhibitory effect on redox state, SNAP did not affect Δψ_m (assessed with TMRE; data not shown), possibly because the change in Δψ_m was too small to detect.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Short application of SNAP induced a small reversible inhibition of mitochondrial respiration. Records showing changes in NADH autofluorescence as percentage of prestimulus baseline in response to inhibition of respiration by 2 mmol/L CN⁻ or 1 mmol/L SNAP, or to uncoupling by 1 μmol/L FCCP, which increases oxygen consumption and hence oxidation of NADH. Each agent was applied by pressure ejection for 10 to 20 seconds, and differences in time course of responses primarily reflect differences in time for drug to leave cell. Traces are displaced for clarity, with dashed lines indicating prestimulus levels.

Prolonged SNAP Pretreatment

Mitochondria of control cells loaded with JC-1 showed red fluorescence, attributable to Δψ_m-dependent JC-1 accumulation and subsequent J-complex formation. In cells pretreated with SNAP for 90 minutes, the mean red fluorescence was reduced to 54.9% of control (P<0.0001, n=16 and 15, respectively; see Figure 4b⇓). The SNAP-induced changes in JC-1 fluorescence were further characterized both morphologically and functionally. Confocal microscopy clearly showed that the red fluorescence originated from mitochondria (Figure 4c⇓), whereas SNAP pretreatment reduced the overall red fluorescence by 51.9% (P<0.001, n=7 fields). To assess Δψ_m-dependent mitochondrial staining, the change in signal was examined with photomultipliers. We found that the red fluorescence gradually decreased with time (Figure 4d⇓). Application of FCCP, however, caused a greater change in red fluorescence in control cells than in SNAP-pretreated cells. To quantify this effect, the FCCP-induced change was expressed relative to the extrapolation of preapplication decay. SNAP pretreatment reduced the FCCP response from 32.1±1.8% to 22.1±3.5% (n=24 and 27, respectively, P=0.016, Student’s t test with Welch correction). These data strongly suggest that the mitochondria were already partially depolarized in SNAP-pretreated cells and that this effect persisted for ≥1 hour after the donor was removed. The partial inhibition of mitochondrial respiration was not associated with any compromise in cell viability up to 24 hours later (MTT bioreduction assay).⁶

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Incubation with SNAP for 90 minutes induced a modest, sustained mitochondrial depolarization. A, Linear protocol; B, chart showing total cellular red fluorescence in JC-1–stained control and SNAP-pretreated cells. To characterize JC-1 fluorescence in control and SNAP-pretreated cells under otherwise identical conditions, C shows example confocal images, and D shows time-dependent changes, integrated with photomultiplier tubes. FCCP 1 μmol/L was applied by pressure ejection to eliminate all Δψ_m-dependent fluorescence. At 3 minutes after FCCP application, mean Δψ_m-dependent fluorescence in SNAP-pretreated cells was compared with that in control treated cells, both expressed relative to fluorescence derived by extrapolation (dotted lines) of a single exponential fit to data before FCCP application. arb.u. indicates arbitrary units.

Effect of SNAP on Mitochondrial Ca²⁺ Uptake

A moderate but sustained reduction of Δψ_m might be expected to reduce mitochondrial Ca²⁺ uptake, which is driven by the electrochemical potential for Ca²⁺ across the mitochondrial inner membrane. To assess the capacity for mitochondria to take up Ca²⁺ independently of other cellular Ca²⁺ fluxes, we developed protocol 4 (see also Figure 5a⇓). Cells were coloaded with rhod-2 and fluo-3 to measure changes in [Ca²⁺]_m and [Ca²⁺]_c, respectively, from the same volume of the cell by confocal microscopy.²⁴ Figure 5b⇓ shows that the rhod-2 fluorescence was clearly localized to mitochondria, whereas the fluo-3 fluorescence was diffusely distributed throughout the cell. Changes in fluo-3 and rhod-2 fluorescence also differed in time course. Brief depolarization of the sarcolemma induced a transient increase in [Ca²⁺]_c and a slower and sustained increase in [Ca²⁺]_m (Figure 5c⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Simultaneous measurement of [Ca²⁺]_c and [Ca²⁺]_m. A, Diagram showing protocol 4, to isolate mitochondrial Ca²⁺ uptake from other Ca²⁺ fluxes. B, Confocal images of neonatal rat ventricular cardiomyocytes coloaded with rhod-2 and fluo-3. C, Effect of a defined Ca²⁺ load induced by application of 50 mmol/L K⁺ on fluo-3 and rhod-2 fluorescence measured over a small cell volume containing a single identifiable mitochondrion. Squares indicate maximum change.

Having established the methodology, we then studied the effect of SNAP pretreatment on the mitochondrial Ca²⁺ uptake (Figure 6⇓). In SNAP-pretreated cells, the mitochondrial signal was harder to identify clearly, and the increase in [Ca²⁺]_m after depolarization was markedly reduced (compare Figure 6b⇓ with 6a), suggesting impaired mitochondrial Ca²⁺ uptake. Mitochondrial Ca²⁺ uptake, however, depends on local [Ca²⁺]_c, and it is therefore necessary to evaluate the peak increase in [Ca²⁺]_m as a function of the peak local [Ca²⁺]_c (Figure 7⇓). Figure 7a⇓ shows that mitochondria in SNAP-pretreated cells accumulated less Ca²⁺ than mitochondria in control cells when exposed to an equivalent [Ca²⁺]_c load. As expected from evidence that NO inhibits sarcolemmal voltage-dependent Ca²⁺ channels,²⁵ the average [Ca²⁺]_c rose by 268±12% in SNAP-pretreated cells compared with 343±10% in control cells, a modest but significant decrease (P<0.0001, n=58; Mann-Whitney; Figure 7b⇓). Excluding all cells with a low increase in [Ca²⁺]_c (threshold 250% of baseline, average [Ca²⁺]_c 349±17% versus 357±9%, P=0.39, n=25 and 52, respectively, Mann-Whitney), however, it was clear that the rise in [Ca²⁺]_m was significantly reduced from 166±7% in control cells to 133±4% in SNAP-pretreated cells (P=0.0004, n=52 and 25, respectively, Mann-Whitney).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Example of SNAP-induced reduction in mitochondrial Ca²⁺ uptake. Confocal images of rhod-2 fluorescence in control (A) and SNAP-pretreated (B) cells before (i) and after (ii) a cytosolic Ca²⁺ load was induced, using protocol in Figure 5a⇑. To demonstrate spatial and temporal change in rhod-2 fluorescence, surface plots (iii) were obtained by plotting time-dependent changes in intensity profile along a line across cell. Note that application of 50 mmol/L K⁺ (arrow) induced very little change in rhod-2 fluorescence over nucleus (*). Scale bars=10 μm. arb. u. indicates arbitrary units.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

SNAP pretreatment reduces capacity of mitochondria to take up Ca²⁺. A, Maximum change in [Ca²⁺]_m as a function of maximum increase in [Ca²⁺]_c (see squares in Figure 5c⇑) obtained using protocol 4 (Figure 5a⇑). B, SNAP pretreatment reduced mean increase in [Ca²⁺]_c. C, At similar [Ca²⁺]_c, SNAP pretreatment reduced load-induced increase in [Ca²⁺]_m.

These data strongly suggest that the capacity of mitochondria to accumulate Ca²⁺ is attenuated in SNAP-pretreated cells.