- Email: [email protected]
Species Dependence of Mitochondrial Calcium Transients during Excitation–Contraction Coupling in Isolated Cardiomyocytes
Biochemical and Biophysical Research Communications 263, 554 –559 (1999) Article ID bbrc.1999.1311, available online at http://www.idealibrary.com on
Species Dependence of Mitochondrial Calcium Transients during Excitation–Contraction Coupling in Isolated Cardiomyocytes Elinor J. Griffiths Bristol Heart Institute, University of Bristol, Bristol Royal Infirmary, United Kingdom
Received July 20, 1999
Whether mitochondrial Ca 21 transport is rapid enough to respond to changes in cytosolic [Ca 21] ([Ca 21] c) which occur during excitation– contraction coupling in the heart is controversial; different results wereobtained with different techniques and different species. In this study mitochondrial [Ca 21] ([Ca 21] m) was measured in indo-1/AM-loaded myocytes from rat and guinea-pig hearts where the cytosolic indo-1 had been removed by extended incubation of cells at 37°C (“heat treatment”). The mitochondrial origin of the remaining fluorescence was confirmed by sensitivity of the indo-1 signal to ruthenium red. In resting rat myocytes, [Ca 21] m was lower than [Ca 21] c, whereas in guinea-pig cells [Ca 21] m was higher than [Ca 21] c. Upon electrical stimulation of cells, no change occurred in [Ca 21] m in rat myocytes. However, in guinea-pig cells mitochondrial Ca 21 transients were clearly visible with a mean indo-1 ratio amplitude of 0.153 6 0.2 (n 5 20), compared with 0.306 6 0.02 (n 5 25), p < 0.001, prior to heat treatment. These observations suggest significant differences in mitochondrial Ca 21 transport in cardiomyocytes from different species. © 1999 Academic Press
Key Words: rat; guinea pig; mitochondria; calcium; indo-1; ruthenium red.
The role of mitochondrial [Ca 21] ([Ca 21] m) in regulation of cytosolic Ca 21 signalling and control of energy production is becoming an area of intense investigation in cardiomyocytes and other cell types (see, for example, 1–5). However, the question of whether [Ca 21] m can respond to the very rapid changes in cytosolic [Ca 21] ([Ca 21] c) which occur during the cardiac cycle of excitation contraction coupling is controversial; some reports found that mitochondrial Ca 21 transients did occur during the contractile cycle (2, 6), whereas others found no change in [Ca 21] m (5, 7, 8). These results are summarised in Table 1, which highlights the fact that two variables occur in these studies: the technique 0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
used and species studied. To date, there have been no reports which directly compare the use of different techniques on the same species, or use of the same technique on different species. The interpretation of results is further complicated by the difficulties in measuring [Ca 21] m in living cardiomyocytes. The first such report used Mn 21 to quench the cytosolic fluorescence signal of indo-1/AM loaded rat myocytes (7), but this was criticised due to the possibility of residual Mn 21 interfering with Ca 21 transport systems. It has recently been confirmed that this technique does give identical results to those obtained when the cytosolic indo-1 signal is removed by another method [8, and see below]. However, the Mn 21quenching method does not appear to work in myocytes from all species (2), an observation which itself suggests that differences in the transport of divalent cations may occur. Studies which use confocal microscopy have the advantage of visualising fluorescence within subcellular mitochondrial compartments, but the possibility of interference by a residual cytosolic signal cannot be entirely ruled out (5). The fluorescent indicator rhod-2 partitions into mitochondria due to their highly negative membrane potential and has been used to measure [Ca 21] m in rabbit myocytes (11). However, under certain loading conditions, this dye has also been used to measure cytosolic [Ca 21] (14) and has the further disadvantage of being a non-ratiometric indicator, thus complicating interpretations from contracting cells. The technique of electron probe microanalysis (EPMA) has been used on rapidly frozen preparations (6, 12), but this can only detect total mitochondrial calcium content. Estimates of the ratio of bound to free calcium within mitochondria vary, but is of the order of several thousand to one (13), raising the possibility that this technique would not detect changes in free matrix [Ca 21] which would be very small compared to the total calcium content. Recently we developed a method for localising indo-1 to the mitochondria of rat myocytes by incubating the
554
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Summary of Data on Mitochondrial Transients Obtained Using Different Techniques, Species, and Myocardial Preparations Authors
Species and preparation
Method
Mitochondrial Ca 21 transients?
Miyata et al. (7) Wendt-Gallitelli and Isenberg (6) Moravec and Bond (9) Chacon et al. (2) Schreur et al. (10) Trollinger et al. (11) Moravec et al. (12) Griffiths et al. (8) Horikawa et al. (13) Zhou et al. (5)
Rat myocytes Guinea-pig myocytes Hamster myocytes Rabbit myocytes Rat heart Rabbit myocytes Hamster heart Rat myocytes Rat papillary muscle Ferret and cat myocytes
Mn-quenching of cytosolic indo-1 EPMA EPMA Confocal imaging fluo-3 and TMRM Mn-quenching of cytosolic indo-1 Confocal imaging rho-2 EPMA Selective removal of cytosolic indo-1 (heat-treatment) EPMA Mn-quenching of cytosolic indo-1
No Yes No Yes No Yes No No No No
Note. Presence of mitochondrial transients defined as a change in [Ca 21] m during a single contraction. EPMA, electron probe microanalysis.
cells so as to allow selective removal of the cytosolic indo-1 via sarcolemmal anion channels—“heat-treatment” (8). The mitochondrial origin of the remaining fluorescence was confirmed by use of inhibitors of mitochondrial transport mechanisms, and by selective permeabilisation of cells with digitonin to release indo-1 from either the cytosol only or from both cytosol and mitochondria. The present study shows that the heat-treatment technique can be used to determine [Ca 21] m in guinea-pig, aswell as rat, cardiomyocytes, and the results show for the first time that cardiac mitochondrial Ca 21 transport differs greatly between these two species. The implications for cell Ca 21 homeostasis and control of energy production are discussed. MATERIALS AND METHODS Myocyte isolation and measurement of indo-1 fluorescence. Single cardiac myocytes were isolated from rat and guinea-pig ventricles by collagenase digestion (8, 15). Male Wistar rats and male guinea-pigs (220-270g) were killed by cervical dislocation and the heart removed and placed in ice-cold “isolation buffer” plus 0.75 mM CaCl 2. Isolation buffer contained, in mM: 20 sodium N-hydroxyethylpiperazineN9-2-ethanesulphonic acid (Hepes), 130 NaCl, 4.5 KCl, 5 MgCl 2, 1 NaH 2PO 4, 21 glucose, 5 Na-pyruvate, pH 7.25 with NaOH. The heart was perfused via the aorta with isolation buffer plus 0.75 mM CaCl 2 at 37°C for 4 min before switching to Ca 21-free buffer (isolation buffer plus 90 mM EGTA) for 4 min. The perfusate was then switched to “enzyme solution” consisting of 50 ml isolation buffer plus 50 mg collagenase, 5 mg protease and 15 mM CaCl 2. The enzyme solution was continued until the tissue felt soft; approximately 15 min for rat hearts and 5 min for guinea-pig hearts. The heart was then washed with isolation buffer plus 150 mM CaCl 2, ventricles removed, sliced approximately 10 times and shaken for 5 min at 37°C in 20 –25 ml isolation buffer plus 150 mM CaCl 2. After filtration the cells were allowed to sediment in this buffer for 7 min. The supernatant was removed and cells resuspended in 0.5 mM CaCl 2, the process repeated and cells finally resuspended in approximately 15 ml of 1 mM CaCl 2. Loading of cells with indo-1/am. 3 ml cell suspension was incubated with 10 mM indo-1/am for 15 min at 30°C. The cells were centrifuged for 1 min at 500 rpm and resuspended in 5 ml isolation buffer (containing 1 mM CaCl 2). The cells were then either used for measurement of total cell fluorescence, or subjected to “heat-
treatment” for measurement of [Ca 21] m (8): cells were maintained at room temperature (approximately 25°C) for 2.5 h, shaken gently at 37°C for 1.5 h, sedimented by centrifugation, resuspended and stored at room temperature. Measurement of fluorescence and cell length. A small portion of the indo-1 loaded cells was placed in an experimental chamber which was mounted on the stage of an inverted microscope (Nikon Diaphot 300). The normal superfusate contained, in mM: 137 NaCl, 5 KCl, 1.2 MgSO 4, 1.2 NaH 2PO 4, 16 D-glucose, 1 CaCl 2, 20 Hepes pH 7.4 (using NaOH), temperature 37°C. In the “low Na” buffer used to induce cellular Ca 21 loading, the 137 mM NaCl in the normal superfusate was replaced with 137 mM choline chloride (pH 7.4 using KOH). The myocyte to be studied was illuminated with a red light and its image visualized with a TV camera and monitor. Indo-1 was excited at 340 –390 nm and emission detected at 410 6 5 nm and 490 6 5 nm, corresponding to the peak emissions of the Ca 21 bound and Ca 21 free forms of the indicator, respectively. Fluorescent light was detected and collected on-line by a Newcastle Photometric Systems Photon Counting System (Newcastle, UK). Light was collected at a rate of up to one data point/10 ms from a single myocyte following subtraction of background fluorescence. Cell length changes were monitored using a Crescent Electronics Video Edge Motion Detector. Materials. Fluorescent dyes were obtained from Molecular Probes, Inc. (Eugene, OR). Collagenase was Worthington type I and protease Sigma type XIV. Other reagents were obtained from Sigma, BDH or Boehringer Mannheim. Expression of results and statistical analyses. Results are expressed as indo-1 ratios, presented as means 6 S.E. unless raw data tracings are shown. Calibration of the indo-1 signal in terms of absolute values of calcium was not done since this would not have provided any advantage over ratio values in the present experiments, and furthermore may have introduced an additional degree of error. Statistical analyses were performed using Student’s t-test (paired where appropriate) or analysis of variance (ANOVA).
RESULTS Localisation of Indo-1 to Mitochondria in Rat and Guinea-Pig Cardiomyocytes When rat myocytes are loaded with indo-1/AM at 30°C for 15 min the dye partitions approximately equally between cytosolic and mitochondrial compartments (7). We found previously that the cytosolic indo-1 could be completely removed by heat-treatment
555
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Mitochondrial origin of the fluorescence signal remaining following heat-treatment in isolated guinea-pig myocytes. Indo-1/AM loaded, heat-treated cells were exposed to low-Na solution (see Methods) followed by washout. This resulted in a rapid increase in the indo-1 ratio in control cells which was prevented by presence of 20 mM ruthenium red (present for 10 min before the start of the experiment). The result is typical of 3 such experiments.
(8). This removal was prevented by probenecid, which inhibits sarcolemmal anion channels. To determine whether the heat-treatment could be used in guinea-pig myocytes to localise indo-1 to the mitochondria, cells from both species were subjected to heat-treatment 6 1 mM probenecid. In rat myocytes, 44 6 2.3% (n 5 6 –13 cells on each day from 6 hearts) of the original fluorescence remained, whereas for guinea-pig cells the value was lower, 35 6 2.5% (n 5 5–12 cells from 4 hearts). Loss of dye was completely prevented in both cases by probenecid. To confirm that the remaining indo-1 fluorescence was originating from the mitochondria in guinea-pig myocytes, cells were subjected to cellular Ca 21 loading by superfusion with low-Na buffer in absence or presence of ruthenium red, an inhibitor of mitochondrial Ca 21 entry. Figure 1 shows that ruthenium red inhibited the increase in indo-1 ratio observed in control cells. This concentration of ruthenium red had been shown previously to be effective in inhibiting Ca 21 entry into mitochondria in rat myocytes (8). Thus heat treatment was indeed an effective way of measuring [Ca 21] m in guinea-pig, aswell as rat, cardiomyocytes.
requires incubation of cells at 37°C for 1.5 h to remove all the cytosolic indo-1. If cells are incubated for shorter periods, loss of cytosolic indo-1 is incomplete, but if cells are incubated for longer periods no further loss of fluorescence occurred, indicating that all the cytosolic indo-1 had been removed. Figure 3 shows mean values for diastolic and systolic indo-1 ratios and the amplitude of the Ca 21 transient before (total cell [Ca 21] and following ([Ca 21] m only) heat-treatment in myocytes from each species. Note that for rat cells, no values for mitochondrial amplitude are given since no Ca 21 transients were seen in these cells (no difference between diastolic and systolic values). Following heat-treatment, diastolic [Ca 21] increased in guinea-pig myocytes but decreased in rat cells. This indicates that, at rest, [Ca 21] m was higher than [Ca 21] c in guinea-pig cells whereas the reverse was seen in rat cells. The mitochondrial Ca 21 transient of guinea-pig myocytes was smaller than the total cell Ca 21 transient, this was due to both a higher diastolic [Ca 21] and lower systolic [Ca 21] in the mitochondrial compartment. DISCUSSION The results presented in this paper show that [Ca 21] m can be measured in both rat and guinea-pig cardiomyocytes using the technique of heat-treatment to remove cytosolic indo-1. This is the first time that mitochondrial Ca 21 transport in myocytes from different species has been compared directly using the same technique. There was no change in [Ca 21] m during the contractile cycle in rat cells, whereas in guinea-pig cells mitochondrial Ca 21 transients occurred during each contraction on a similar timescale to the cytosolic Ca 21 transients. The amplitude of the mitochondrial transient was smaller than that of the cytosolic Ca 21
Mitochondrial [Ca 21] during the Contractile Cycle When indo-1/AM loaded, untreated, myocytes are electrically stimulated to contract, clear Ca 21 transients are visible during each contraction. However, in rat myocytes following either Mn 21 quenching or heattreatment, transients are no longer visible, suggesting that [Ca 21] m does not change on a beat-to-beat basis (7, 8). Figure 2 shows that following heat-treatment of guinea-pig cells, mitochondrial transients did occur during each contraction. This was unlikely to be due to any interference by cytosolic indo-1 since the results above indicated that ruthenium red could inhibit Ca 21 entry. Furthermore, the procedure of heat treatment
FIG. 2. Mitochondrial [Ca 21] during the contractile cycle in rat and guinea-pig myocytes. Simultaneous recordings of indo-1 fluorescence and cell length are shown.
556
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Changes in total and mitochondrial [Ca 21] upon electrical stimulation. Mean values for diastolic, systolic and amplitude (systolic-diastolic) indo-1 ratio are presented. Indo-1 ratios were determined prior to (total) and following (mitochondrial) heattreatment (see text for details). Note that amplitude values are not given for rat cells since no mitochondrial Ca 21 transients are visible. Statistical significance: **p , 0.001, *p , 0.01 vs total (paired t-test); for rat vs guinea-pig: $$p , 0.001, $p , 0.05 for corresponding indo-1 ratios (unpaired t-test).
transient due to both a higher diastolic and lower systolic [Ca 21] in the mitochondria. The lack of response of [Ca 21] m in rat cells during the contractile cycle agrees with previous work of ours and others (7, 8, 10). Such a response was predicted from early studies of Ca 21 transport on suspensions of isolated mitochondria [reviewed in 16, 17]. These studies revealed that Ca 21 influx occurs via a uniporter, inhibited by ruthenium red and dependent on membrane potential, and efflux occurs via Na 1/Ca 21 exchange, inhibited by diltiazem, clonazepam and CGP37157 (16 –18). A third non-specific pathway can be induced under pathological conditions and high Ca 21 loading, namely the mitochondrial permeability transition pore, but the role of this, if any, under physiological conditions is not known (16, 19, 20). At low external [Ca 21] ([Ca 21] e), the activity of the exchanger exceeds that of the uniporter, thus keeping [Ca 21] m below that of the [Ca 21] e. However, as [Ca 21] e rises above approximately 0.5–1 mM, the efflux pathway saturates whereas the sigmoidal kinetics of the uniporter allow much faster rates of Ca 21 entry resulting in higher [Ca 21] m compared with [Ca 21] e. Such predictions were
confirmed more recently by Leisey at al. (21) using isolated rat heart mitochondria exposed to very rapid changes in [Ca 21] e. These predictions also fit with the results presented here and previously (7, 8): in resting rat myocytes, [Ca 21] m remained below [Ca 21] c. However, if a sustained elevation of [Ca 21] c occurred, such as during rapid stimulation in presence of an adrenergic agonist, then [Ca 21] m increased, reaching a peak in 1-2 min (7, 8). The results of the present study and that of Chacon et al. (2) suggest that in intact rabbit and guinea-pig myocytes the kinetics of mitochondrial Ca 21 transport are very different from those of rat myocytes. For [Ca 21] m to respond to the very rapid changes in [Ca 21] c would require a much faster uptake and release of Ca 21 by mitochondria in these species. At present it cannot be determined whether this is due to intrinsic properties of the transporters, and would therefore persist through isolation, or to differences in the intracellular environments between different species. With regard to the latter, both Mg 21 and spermine can greatly affect Ca 21 transport in isolated mitochondria (16, 22); for example, inclusion of 1 mM Mg 21 in the incubation medium changes the kinetics of Ca 21 uptake from hyperbolic to sigmoidal. There have also been reports of adrenergic modulation of transport; an increased rate of Ca 21 uptake into mitochondria isolated from a 1adrenergically stimulated hearts (23) and a preliminary report of a b-adrenergic inhibition of Ca 21 release in isolated myocytes (24). Direct evidence does, of course, require studies of the isolated systems, but this has yet to be attempted. Species Differences in Excitation–Contraction Coupling The evidence presented here that the mitochondrial Ca 21 transporters have different activities in different species has a precedent in the known properties of other transport systems; the contribution of sarcolemmal (SL) and sarcoplasmic reticular (SR) Ca 21 transport systems in controlling contraction and relaxation varies considerably between species (23). For example, the contribution of the SR Ca-ATPase is estimated to account for 92% of the total Ca-extrusion mechanisms during relaxation in rat myocytes, but is lower in other species, approximately 70% in rabbit and guinea-pig myocytes (26). The latter species have correspondingly higher rates of SL-Na/Ca activity, approximately 30% compared to only 7% in the rat. The contribution of the so-called “slow” systems i.e. the mitochondria and SL Ca-ATPase, to Ca 21 extrusion was estimated to contribute less than 2% in all species (26). The results presented in this paper suggest that this is not surprising; the function of the mitochondrial transporters in guinea-pig cells appears to be to almost mirror the changes in [Ca 21] c, rather than remove the Ca 21 to an
557
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
inactive pool, whereas in rat myocytes the changes in [Ca 21] c are detected more slowly by the mitochondria. Our results also suggest a possible role for the mitochondrial Ca 21 transporters in modulating resting [Ca 21] c levels; the higher activity of the uniporter in guinea-pig myocytes may account for the observed lower resting [Ca 21] c compared with [Ca 21] m in these cells whereas the reverse is seen in rat cells due to the relative inactivity of the uniporter under resting conditions. Subcellular Organisation The subcellular spatial organisation of mitochondria is another important factor in considering the Ca 21 transporting properties of these organelles. According to our predictions, [Ca 21] m falls rapidly upon relaxation of guinea-pig cells, but declines more slowly in rat cells once the average [Ca 21] c has decreased. However, whether the Ca 21 is released into the cytosol and then removed by either SR or SL pathways, or whether the mitochondrial transporters are located in close enough proximity to one or other of these transporters to allow direct Ca 21 removal, is not known. Recent work in HeLa cells revealed numerous close contact sites between the ER and mitochondria; Rizzuto et al. (27), using two differently coloured green fluorescent proteins targeted to different sites, found that upon opening of the IP 3 gated channel of the endoplasmic reticulum, the mitochondrial surface was exposed to a much higher [Ca 21] than the bulk phase of the cytosol. Another study using a neuronal cell line found that there was “privileged” access of Ca 21 into mitochondria upon activation of N-methyl-D-aspartate (NMDA) receptors: upon cell stimulation with NMDA, mitochondrial and cytosolic [Ca 21] changed with similar timecourses, whereas if [Ca 21] was elevated by a different means (using ionomycin or KCl-induced depolarisation), [Ca 21] m lagged behind changes in [Ca 21] c (28). Thus in non-myocardial cells there is evidence for association of mitochondrial Ca 21 channels with transporters of other membrane systems, but it remains to be seen whether this is also the case in heart cells. [Ca 21] m and Energy Production An important question remains as to the role of mitochondrial Ca 21 transport in regulating ATP synthesis. In isolated mitochondria, ATP synthesis is stimulated as the ADP/ATP ratio rises, however, in vivo NMR studies on whole hearts found no change in this ratio upon increased workloads, suggesting another mechanism was responsible (29). Work in the early 1980s showed that mitochondrial dehydrogenases could be activated by Ca 21 at submicromolar concentrations [reviewed in 30, 31). This, together with the NMR data, gave rise to the now commonly accepted theory that the role of [Ca 21] m in heart muscle is to
co-ordinate ATP supply by the mitochondria with the ATP demand by myocyte contractile and metabolic activities (30, 31). Thus increases in [Ca 21] c, for example during hormonal stimulation or upon increased cardiac workload, would be relayed to the mitochondria resulting in stimulation of mitochondrial dehydrogenase activity and increased NADH supply for ATP. It is, however, becoming increasingly clear that the story is much more complicated; NADH levels have been reported to increase, decrease or remain unchanged in myocardial preparations (15, 32–34), a stimulation of pyruvate dehydrogenase activity upon inotropic stimulation of isolated hearts has been reported with no corresponding change in [Ca 21] m (12) and Ca 21 may exert control at other levels of mitochondrial function, for example by regulating the ADP/ATP translocase and ATP synthase (31, 35). In summary, this paper has shown that the contradictory work obtained on mitochondrial Ca 21 transport in myocytes can be explained by the different species used. However, this result itself opens many important, unresolved, areas: the subcellular distribution of mitochondria and possible areas of contact with the SR or SL, the intracellular factors that control mitochondrial Ca 21 transport, and the implications for the differences in transport in the control of cellular Ca 21 homeostasis and energy production. ACKNOWLEDGMENT This work was supported by the British Heart Foundation.
REFERENCES 1. Hajnoczky, G., Robb-Gaspers, L. D., Seitz, M. B., and Thomas, A. P. (1995) Cell 82, 415– 424. 2. Chacon, E., Ohata, H., Harper, I. S., Trollinger, D. R., Herman, B., and Lemasters, J. J. (1996) FEBS Lett. 382, 31–36. 3. Rutter, G. A., Burnett, P., Rizzuto, R., et al. (1996) Proc. Natl. Acad. Sci. USA 93, 5489 –5494. 4. Babcock, D. F., Herrington, J., Goodwin, P. C., Park, Y. B., and Hille, B. (1997) J. Cell Biol. 136, 833– 844. 5. Zhou, Z., Matlib, M. A., and Bers, D. M. (1998) J. Physiol. 507, 379 – 403. 6. Wendt-Gallitelli, M.-F., and Isenberg, G. (1991) J. Physiol. 435, 349 –372. 7. Miyata, H., Silverman, H. S., Sollot, S. J., Lakatta, E. G., Stern, M. D., and Hansford, R. G. (1991) Am. J. Physiol. 261, H1123– H1134. 8. Griffiths, E. J., Stern, M. D., and Silverman, H. S. (1997) Am. J. Physiol. 273, C37–C44. 9. Moravec, C. S., and Bond, M. (1991) Am. J. Physiol. 260, H989 – H997. 10. Schreur, J. H. M., Figueredo, V. M., Miyamae, M., Shames, D. M., Baker, A. J., and Camacho, S. A. (1996) Biophys. J. 70, 2571–2580. 11. Trollinger, D. R., Cascio, W. E., and Lemasters, J. J. (1997) Biochem. Biophys. Res. Commun. 236, 738 –742. 12. Moravec, C. S., Desnoyer, R. W., Milovanovic, M., Schluchter, M. D., and Bond, M. (1997) Am. J. Physiol. 273, H1432–H1439.
558
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
13. Horikawa, Y., Goel, A., Somlyo, A. P., and Somlyo, A. V. (1998) Biophys. J. 74, 1579 –1590. 14. Del Nido, P. J., Glynn, P., Buenavrntura, P., Salama, G., and Koretsky, A. P. (1998) Am. J. Physiol. 274, H728 –H741. 15. Griffiths, E. J., Lin, H., and Suleiman, M.-S. (1998) Biochem. Pharmacol. 56, 173–179. 16. Crompton, M. (1990) in Calcium and the Heart (Langer, G. A., Ed.), pp. 167–198. Raven Press, New York. 17. Gunter, T. E., and Pfeiffer, D. R. (1990) Am. J. Physiol. 258, C304 –310. 18. Cox, D. A., and Matlib, M. A. (1993) J. Biol. Chem. 268, 938 – 947. 19. Bernardi, P., Broekemeier, K. M., and Pfeiffer, D. R. (1994) J. Bioenerg. Biomembr. 26, 509 –517. 20. Halestrap, A. P., Griffiths, E. J., Connern, C. P., and Kerr, P. M. (1997) Mol. Cell Biochem. 174, 167–172. 21. Leisey, J. R., Grotyohann, L. W., Scott, D. A., and Scaduto, R. C. (1993) Am. J. Physiol. 255, C162–C168. 22. Ventura, C., Ferroni, C., Flamigni, F., Stefanelli,C., and Caprogrossi, M. C. (1994) Am. J. Physiol. 267, H587–H592. 23. Kessar, P., and Crompton, M. (1981) Biochem. J. 200, 379 –388.
24. Griffiths, E. J., Stern, M. D., and Silverman, H. S. (1994) J. Mol. Cell Cardiol. 26, CLIV. 25. Bers, D. M. (1997) Bas. Res. Cardiol. 92, S1, 1–10. 26. Bassani, J. W. M., Bassani, R. A., and Bers, D. M. (1994) J. Physiol. 476, 279 –293. 27. Rizzuto, R., Pinton, P., Carrington, W., et al. (1998) Science 280, 1763–1766. 28. Peng, T. I., and Greenan, J. T. (1998) Mol. Pharmacol. 53, 974 –980. 29. Balaban, R. S., Kantor, H. L., Katz, L. A., and Briggs, R. W. (1986) Science 232, 1121–1123. 30. McCormack, J. G., Halestrap, A. P., and Denton, R. M. (1990) Physiol. Rev. 70, 391– 425. 31. Hansford, R. G. (1994) J. Bioenerg. Biomemb. 26, 495–508. 32. Heineman, F. W., and Balaban, R. S. (1993) Am. J. Physiol. 264, H433–H440. 33. Brandes, R., and Bers, D. M. (1997) Circ. Res. 80, 82– 87. 34. Griffiths, E. J., Wei, S.-K., Haigney, M. C. P., Ocampo, C. J., Sern, M. D., and Silverman, H. S. (1997) Cell Calcium 21, 321–329. 35. Das, A. M., and Harris, D. A. (1990) Cardiovasc. Res. 24, 411– 417.
559
Species Dependence of Mitochondrial Calcium Transients during Excitation–Contraction Coupling in Isolated Cardiomyocytes Elinor J. Griffiths Bristol Heart Institute, University of Bristol, Bristol Royal Infirmary, United Kingdom
Received July 20, 1999
Whether mitochondrial Ca 21 transport is rapid enough to respond to changes in cytosolic [Ca 21] ([Ca 21] c) which occur during excitation– contraction coupling in the heart is controversial; different results wereobtained with different techniques and different species. In this study mitochondrial [Ca 21] ([Ca 21] m) was measured in indo-1/AM-loaded myocytes from rat and guinea-pig hearts where the cytosolic indo-1 had been removed by extended incubation of cells at 37°C (“heat treatment”). The mitochondrial origin of the remaining fluorescence was confirmed by sensitivity of the indo-1 signal to ruthenium red. In resting rat myocytes, [Ca 21] m was lower than [Ca 21] c, whereas in guinea-pig cells [Ca 21] m was higher than [Ca 21] c. Upon electrical stimulation of cells, no change occurred in [Ca 21] m in rat myocytes. However, in guinea-pig cells mitochondrial Ca 21 transients were clearly visible with a mean indo-1 ratio amplitude of 0.153 6 0.2 (n 5 20), compared with 0.306 6 0.02 (n 5 25), p < 0.001, prior to heat treatment. These observations suggest significant differences in mitochondrial Ca 21 transport in cardiomyocytes from different species. © 1999 Academic Press
Key Words: rat; guinea pig; mitochondria; calcium; indo-1; ruthenium red.
The role of mitochondrial [Ca 21] ([Ca 21] m) in regulation of cytosolic Ca 21 signalling and control of energy production is becoming an area of intense investigation in cardiomyocytes and other cell types (see, for example, 1–5). However, the question of whether [Ca 21] m can respond to the very rapid changes in cytosolic [Ca 21] ([Ca 21] c) which occur during the cardiac cycle of excitation contraction coupling is controversial; some reports found that mitochondrial Ca 21 transients did occur during the contractile cycle (2, 6), whereas others found no change in [Ca 21] m (5, 7, 8). These results are summarised in Table 1, which highlights the fact that two variables occur in these studies: the technique 0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
used and species studied. To date, there have been no reports which directly compare the use of different techniques on the same species, or use of the same technique on different species. The interpretation of results is further complicated by the difficulties in measuring [Ca 21] m in living cardiomyocytes. The first such report used Mn 21 to quench the cytosolic fluorescence signal of indo-1/AM loaded rat myocytes (7), but this was criticised due to the possibility of residual Mn 21 interfering with Ca 21 transport systems. It has recently been confirmed that this technique does give identical results to those obtained when the cytosolic indo-1 signal is removed by another method [8, and see below]. However, the Mn 21quenching method does not appear to work in myocytes from all species (2), an observation which itself suggests that differences in the transport of divalent cations may occur. Studies which use confocal microscopy have the advantage of visualising fluorescence within subcellular mitochondrial compartments, but the possibility of interference by a residual cytosolic signal cannot be entirely ruled out (5). The fluorescent indicator rhod-2 partitions into mitochondria due to their highly negative membrane potential and has been used to measure [Ca 21] m in rabbit myocytes (11). However, under certain loading conditions, this dye has also been used to measure cytosolic [Ca 21] (14) and has the further disadvantage of being a non-ratiometric indicator, thus complicating interpretations from contracting cells. The technique of electron probe microanalysis (EPMA) has been used on rapidly frozen preparations (6, 12), but this can only detect total mitochondrial calcium content. Estimates of the ratio of bound to free calcium within mitochondria vary, but is of the order of several thousand to one (13), raising the possibility that this technique would not detect changes in free matrix [Ca 21] which would be very small compared to the total calcium content. Recently we developed a method for localising indo-1 to the mitochondria of rat myocytes by incubating the
554
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Summary of Data on Mitochondrial Transients Obtained Using Different Techniques, Species, and Myocardial Preparations Authors
Species and preparation
Method
Mitochondrial Ca 21 transients?
Miyata et al. (7) Wendt-Gallitelli and Isenberg (6) Moravec and Bond (9) Chacon et al. (2) Schreur et al. (10) Trollinger et al. (11) Moravec et al. (12) Griffiths et al. (8) Horikawa et al. (13) Zhou et al. (5)
Rat myocytes Guinea-pig myocytes Hamster myocytes Rabbit myocytes Rat heart Rabbit myocytes Hamster heart Rat myocytes Rat papillary muscle Ferret and cat myocytes
Mn-quenching of cytosolic indo-1 EPMA EPMA Confocal imaging fluo-3 and TMRM Mn-quenching of cytosolic indo-1 Confocal imaging rho-2 EPMA Selective removal of cytosolic indo-1 (heat-treatment) EPMA Mn-quenching of cytosolic indo-1
No Yes No Yes No Yes No No No No
Note. Presence of mitochondrial transients defined as a change in [Ca 21] m during a single contraction. EPMA, electron probe microanalysis.
cells so as to allow selective removal of the cytosolic indo-1 via sarcolemmal anion channels—“heat-treatment” (8). The mitochondrial origin of the remaining fluorescence was confirmed by use of inhibitors of mitochondrial transport mechanisms, and by selective permeabilisation of cells with digitonin to release indo-1 from either the cytosol only or from both cytosol and mitochondria. The present study shows that the heat-treatment technique can be used to determine [Ca 21] m in guinea-pig, aswell as rat, cardiomyocytes, and the results show for the first time that cardiac mitochondrial Ca 21 transport differs greatly between these two species. The implications for cell Ca 21 homeostasis and control of energy production are discussed. MATERIALS AND METHODS Myocyte isolation and measurement of indo-1 fluorescence. Single cardiac myocytes were isolated from rat and guinea-pig ventricles by collagenase digestion (8, 15). Male Wistar rats and male guinea-pigs (220-270g) were killed by cervical dislocation and the heart removed and placed in ice-cold “isolation buffer” plus 0.75 mM CaCl 2. Isolation buffer contained, in mM: 20 sodium N-hydroxyethylpiperazineN9-2-ethanesulphonic acid (Hepes), 130 NaCl, 4.5 KCl, 5 MgCl 2, 1 NaH 2PO 4, 21 glucose, 5 Na-pyruvate, pH 7.25 with NaOH. The heart was perfused via the aorta with isolation buffer plus 0.75 mM CaCl 2 at 37°C for 4 min before switching to Ca 21-free buffer (isolation buffer plus 90 mM EGTA) for 4 min. The perfusate was then switched to “enzyme solution” consisting of 50 ml isolation buffer plus 50 mg collagenase, 5 mg protease and 15 mM CaCl 2. The enzyme solution was continued until the tissue felt soft; approximately 15 min for rat hearts and 5 min for guinea-pig hearts. The heart was then washed with isolation buffer plus 150 mM CaCl 2, ventricles removed, sliced approximately 10 times and shaken for 5 min at 37°C in 20 –25 ml isolation buffer plus 150 mM CaCl 2. After filtration the cells were allowed to sediment in this buffer for 7 min. The supernatant was removed and cells resuspended in 0.5 mM CaCl 2, the process repeated and cells finally resuspended in approximately 15 ml of 1 mM CaCl 2. Loading of cells with indo-1/am. 3 ml cell suspension was incubated with 10 mM indo-1/am for 15 min at 30°C. The cells were centrifuged for 1 min at 500 rpm and resuspended in 5 ml isolation buffer (containing 1 mM CaCl 2). The cells were then either used for measurement of total cell fluorescence, or subjected to “heat-
treatment” for measurement of [Ca 21] m (8): cells were maintained at room temperature (approximately 25°C) for 2.5 h, shaken gently at 37°C for 1.5 h, sedimented by centrifugation, resuspended and stored at room temperature. Measurement of fluorescence and cell length. A small portion of the indo-1 loaded cells was placed in an experimental chamber which was mounted on the stage of an inverted microscope (Nikon Diaphot 300). The normal superfusate contained, in mM: 137 NaCl, 5 KCl, 1.2 MgSO 4, 1.2 NaH 2PO 4, 16 D-glucose, 1 CaCl 2, 20 Hepes pH 7.4 (using NaOH), temperature 37°C. In the “low Na” buffer used to induce cellular Ca 21 loading, the 137 mM NaCl in the normal superfusate was replaced with 137 mM choline chloride (pH 7.4 using KOH). The myocyte to be studied was illuminated with a red light and its image visualized with a TV camera and monitor. Indo-1 was excited at 340 –390 nm and emission detected at 410 6 5 nm and 490 6 5 nm, corresponding to the peak emissions of the Ca 21 bound and Ca 21 free forms of the indicator, respectively. Fluorescent light was detected and collected on-line by a Newcastle Photometric Systems Photon Counting System (Newcastle, UK). Light was collected at a rate of up to one data point/10 ms from a single myocyte following subtraction of background fluorescence. Cell length changes were monitored using a Crescent Electronics Video Edge Motion Detector. Materials. Fluorescent dyes were obtained from Molecular Probes, Inc. (Eugene, OR). Collagenase was Worthington type I and protease Sigma type XIV. Other reagents were obtained from Sigma, BDH or Boehringer Mannheim. Expression of results and statistical analyses. Results are expressed as indo-1 ratios, presented as means 6 S.E. unless raw data tracings are shown. Calibration of the indo-1 signal in terms of absolute values of calcium was not done since this would not have provided any advantage over ratio values in the present experiments, and furthermore may have introduced an additional degree of error. Statistical analyses were performed using Student’s t-test (paired where appropriate) or analysis of variance (ANOVA).
RESULTS Localisation of Indo-1 to Mitochondria in Rat and Guinea-Pig Cardiomyocytes When rat myocytes are loaded with indo-1/AM at 30°C for 15 min the dye partitions approximately equally between cytosolic and mitochondrial compartments (7). We found previously that the cytosolic indo-1 could be completely removed by heat-treatment
555
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Mitochondrial origin of the fluorescence signal remaining following heat-treatment in isolated guinea-pig myocytes. Indo-1/AM loaded, heat-treated cells were exposed to low-Na solution (see Methods) followed by washout. This resulted in a rapid increase in the indo-1 ratio in control cells which was prevented by presence of 20 mM ruthenium red (present for 10 min before the start of the experiment). The result is typical of 3 such experiments.
(8). This removal was prevented by probenecid, which inhibits sarcolemmal anion channels. To determine whether the heat-treatment could be used in guinea-pig myocytes to localise indo-1 to the mitochondria, cells from both species were subjected to heat-treatment 6 1 mM probenecid. In rat myocytes, 44 6 2.3% (n 5 6 –13 cells on each day from 6 hearts) of the original fluorescence remained, whereas for guinea-pig cells the value was lower, 35 6 2.5% (n 5 5–12 cells from 4 hearts). Loss of dye was completely prevented in both cases by probenecid. To confirm that the remaining indo-1 fluorescence was originating from the mitochondria in guinea-pig myocytes, cells were subjected to cellular Ca 21 loading by superfusion with low-Na buffer in absence or presence of ruthenium red, an inhibitor of mitochondrial Ca 21 entry. Figure 1 shows that ruthenium red inhibited the increase in indo-1 ratio observed in control cells. This concentration of ruthenium red had been shown previously to be effective in inhibiting Ca 21 entry into mitochondria in rat myocytes (8). Thus heat treatment was indeed an effective way of measuring [Ca 21] m in guinea-pig, aswell as rat, cardiomyocytes.
requires incubation of cells at 37°C for 1.5 h to remove all the cytosolic indo-1. If cells are incubated for shorter periods, loss of cytosolic indo-1 is incomplete, but if cells are incubated for longer periods no further loss of fluorescence occurred, indicating that all the cytosolic indo-1 had been removed. Figure 3 shows mean values for diastolic and systolic indo-1 ratios and the amplitude of the Ca 21 transient before (total cell [Ca 21] and following ([Ca 21] m only) heat-treatment in myocytes from each species. Note that for rat cells, no values for mitochondrial amplitude are given since no Ca 21 transients were seen in these cells (no difference between diastolic and systolic values). Following heat-treatment, diastolic [Ca 21] increased in guinea-pig myocytes but decreased in rat cells. This indicates that, at rest, [Ca 21] m was higher than [Ca 21] c in guinea-pig cells whereas the reverse was seen in rat cells. The mitochondrial Ca 21 transient of guinea-pig myocytes was smaller than the total cell Ca 21 transient, this was due to both a higher diastolic [Ca 21] and lower systolic [Ca 21] in the mitochondrial compartment. DISCUSSION The results presented in this paper show that [Ca 21] m can be measured in both rat and guinea-pig cardiomyocytes using the technique of heat-treatment to remove cytosolic indo-1. This is the first time that mitochondrial Ca 21 transport in myocytes from different species has been compared directly using the same technique. There was no change in [Ca 21] m during the contractile cycle in rat cells, whereas in guinea-pig cells mitochondrial Ca 21 transients occurred during each contraction on a similar timescale to the cytosolic Ca 21 transients. The amplitude of the mitochondrial transient was smaller than that of the cytosolic Ca 21
Mitochondrial [Ca 21] during the Contractile Cycle When indo-1/AM loaded, untreated, myocytes are electrically stimulated to contract, clear Ca 21 transients are visible during each contraction. However, in rat myocytes following either Mn 21 quenching or heattreatment, transients are no longer visible, suggesting that [Ca 21] m does not change on a beat-to-beat basis (7, 8). Figure 2 shows that following heat-treatment of guinea-pig cells, mitochondrial transients did occur during each contraction. This was unlikely to be due to any interference by cytosolic indo-1 since the results above indicated that ruthenium red could inhibit Ca 21 entry. Furthermore, the procedure of heat treatment
FIG. 2. Mitochondrial [Ca 21] during the contractile cycle in rat and guinea-pig myocytes. Simultaneous recordings of indo-1 fluorescence and cell length are shown.
556
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Changes in total and mitochondrial [Ca 21] upon electrical stimulation. Mean values for diastolic, systolic and amplitude (systolic-diastolic) indo-1 ratio are presented. Indo-1 ratios were determined prior to (total) and following (mitochondrial) heattreatment (see text for details). Note that amplitude values are not given for rat cells since no mitochondrial Ca 21 transients are visible. Statistical significance: **p , 0.001, *p , 0.01 vs total (paired t-test); for rat vs guinea-pig: $$p , 0.001, $p , 0.05 for corresponding indo-1 ratios (unpaired t-test).
transient due to both a higher diastolic and lower systolic [Ca 21] in the mitochondria. The lack of response of [Ca 21] m in rat cells during the contractile cycle agrees with previous work of ours and others (7, 8, 10). Such a response was predicted from early studies of Ca 21 transport on suspensions of isolated mitochondria [reviewed in 16, 17]. These studies revealed that Ca 21 influx occurs via a uniporter, inhibited by ruthenium red and dependent on membrane potential, and efflux occurs via Na 1/Ca 21 exchange, inhibited by diltiazem, clonazepam and CGP37157 (16 –18). A third non-specific pathway can be induced under pathological conditions and high Ca 21 loading, namely the mitochondrial permeability transition pore, but the role of this, if any, under physiological conditions is not known (16, 19, 20). At low external [Ca 21] ([Ca 21] e), the activity of the exchanger exceeds that of the uniporter, thus keeping [Ca 21] m below that of the [Ca 21] e. However, as [Ca 21] e rises above approximately 0.5–1 mM, the efflux pathway saturates whereas the sigmoidal kinetics of the uniporter allow much faster rates of Ca 21 entry resulting in higher [Ca 21] m compared with [Ca 21] e. Such predictions were
confirmed more recently by Leisey at al. (21) using isolated rat heart mitochondria exposed to very rapid changes in [Ca 21] e. These predictions also fit with the results presented here and previously (7, 8): in resting rat myocytes, [Ca 21] m remained below [Ca 21] c. However, if a sustained elevation of [Ca 21] c occurred, such as during rapid stimulation in presence of an adrenergic agonist, then [Ca 21] m increased, reaching a peak in 1-2 min (7, 8). The results of the present study and that of Chacon et al. (2) suggest that in intact rabbit and guinea-pig myocytes the kinetics of mitochondrial Ca 21 transport are very different from those of rat myocytes. For [Ca 21] m to respond to the very rapid changes in [Ca 21] c would require a much faster uptake and release of Ca 21 by mitochondria in these species. At present it cannot be determined whether this is due to intrinsic properties of the transporters, and would therefore persist through isolation, or to differences in the intracellular environments between different species. With regard to the latter, both Mg 21 and spermine can greatly affect Ca 21 transport in isolated mitochondria (16, 22); for example, inclusion of 1 mM Mg 21 in the incubation medium changes the kinetics of Ca 21 uptake from hyperbolic to sigmoidal. There have also been reports of adrenergic modulation of transport; an increased rate of Ca 21 uptake into mitochondria isolated from a 1adrenergically stimulated hearts (23) and a preliminary report of a b-adrenergic inhibition of Ca 21 release in isolated myocytes (24). Direct evidence does, of course, require studies of the isolated systems, but this has yet to be attempted. Species Differences in Excitation–Contraction Coupling The evidence presented here that the mitochondrial Ca 21 transporters have different activities in different species has a precedent in the known properties of other transport systems; the contribution of sarcolemmal (SL) and sarcoplasmic reticular (SR) Ca 21 transport systems in controlling contraction and relaxation varies considerably between species (23). For example, the contribution of the SR Ca-ATPase is estimated to account for 92% of the total Ca-extrusion mechanisms during relaxation in rat myocytes, but is lower in other species, approximately 70% in rabbit and guinea-pig myocytes (26). The latter species have correspondingly higher rates of SL-Na/Ca activity, approximately 30% compared to only 7% in the rat. The contribution of the so-called “slow” systems i.e. the mitochondria and SL Ca-ATPase, to Ca 21 extrusion was estimated to contribute less than 2% in all species (26). The results presented in this paper suggest that this is not surprising; the function of the mitochondrial transporters in guinea-pig cells appears to be to almost mirror the changes in [Ca 21] c, rather than remove the Ca 21 to an
557
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
inactive pool, whereas in rat myocytes the changes in [Ca 21] c are detected more slowly by the mitochondria. Our results also suggest a possible role for the mitochondrial Ca 21 transporters in modulating resting [Ca 21] c levels; the higher activity of the uniporter in guinea-pig myocytes may account for the observed lower resting [Ca 21] c compared with [Ca 21] m in these cells whereas the reverse is seen in rat cells due to the relative inactivity of the uniporter under resting conditions. Subcellular Organisation The subcellular spatial organisation of mitochondria is another important factor in considering the Ca 21 transporting properties of these organelles. According to our predictions, [Ca 21] m falls rapidly upon relaxation of guinea-pig cells, but declines more slowly in rat cells once the average [Ca 21] c has decreased. However, whether the Ca 21 is released into the cytosol and then removed by either SR or SL pathways, or whether the mitochondrial transporters are located in close enough proximity to one or other of these transporters to allow direct Ca 21 removal, is not known. Recent work in HeLa cells revealed numerous close contact sites between the ER and mitochondria; Rizzuto et al. (27), using two differently coloured green fluorescent proteins targeted to different sites, found that upon opening of the IP 3 gated channel of the endoplasmic reticulum, the mitochondrial surface was exposed to a much higher [Ca 21] than the bulk phase of the cytosol. Another study using a neuronal cell line found that there was “privileged” access of Ca 21 into mitochondria upon activation of N-methyl-D-aspartate (NMDA) receptors: upon cell stimulation with NMDA, mitochondrial and cytosolic [Ca 21] changed with similar timecourses, whereas if [Ca 21] was elevated by a different means (using ionomycin or KCl-induced depolarisation), [Ca 21] m lagged behind changes in [Ca 21] c (28). Thus in non-myocardial cells there is evidence for association of mitochondrial Ca 21 channels with transporters of other membrane systems, but it remains to be seen whether this is also the case in heart cells. [Ca 21] m and Energy Production An important question remains as to the role of mitochondrial Ca 21 transport in regulating ATP synthesis. In isolated mitochondria, ATP synthesis is stimulated as the ADP/ATP ratio rises, however, in vivo NMR studies on whole hearts found no change in this ratio upon increased workloads, suggesting another mechanism was responsible (29). Work in the early 1980s showed that mitochondrial dehydrogenases could be activated by Ca 21 at submicromolar concentrations [reviewed in 30, 31). This, together with the NMR data, gave rise to the now commonly accepted theory that the role of [Ca 21] m in heart muscle is to
co-ordinate ATP supply by the mitochondria with the ATP demand by myocyte contractile and metabolic activities (30, 31). Thus increases in [Ca 21] c, for example during hormonal stimulation or upon increased cardiac workload, would be relayed to the mitochondria resulting in stimulation of mitochondrial dehydrogenase activity and increased NADH supply for ATP. It is, however, becoming increasingly clear that the story is much more complicated; NADH levels have been reported to increase, decrease or remain unchanged in myocardial preparations (15, 32–34), a stimulation of pyruvate dehydrogenase activity upon inotropic stimulation of isolated hearts has been reported with no corresponding change in [Ca 21] m (12) and Ca 21 may exert control at other levels of mitochondrial function, for example by regulating the ADP/ATP translocase and ATP synthase (31, 35). In summary, this paper has shown that the contradictory work obtained on mitochondrial Ca 21 transport in myocytes can be explained by the different species used. However, this result itself opens many important, unresolved, areas: the subcellular distribution of mitochondria and possible areas of contact with the SR or SL, the intracellular factors that control mitochondrial Ca 21 transport, and the implications for the differences in transport in the control of cellular Ca 21 homeostasis and energy production. ACKNOWLEDGMENT This work was supported by the British Heart Foundation.
REFERENCES 1. Hajnoczky, G., Robb-Gaspers, L. D., Seitz, M. B., and Thomas, A. P. (1995) Cell 82, 415– 424. 2. Chacon, E., Ohata, H., Harper, I. S., Trollinger, D. R., Herman, B., and Lemasters, J. J. (1996) FEBS Lett. 382, 31–36. 3. Rutter, G. A., Burnett, P., Rizzuto, R., et al. (1996) Proc. Natl. Acad. Sci. USA 93, 5489 –5494. 4. Babcock, D. F., Herrington, J., Goodwin, P. C., Park, Y. B., and Hille, B. (1997) J. Cell Biol. 136, 833– 844. 5. Zhou, Z., Matlib, M. A., and Bers, D. M. (1998) J. Physiol. 507, 379 – 403. 6. Wendt-Gallitelli, M.-F., and Isenberg, G. (1991) J. Physiol. 435, 349 –372. 7. Miyata, H., Silverman, H. S., Sollot, S. J., Lakatta, E. G., Stern, M. D., and Hansford, R. G. (1991) Am. J. Physiol. 261, H1123– H1134. 8. Griffiths, E. J., Stern, M. D., and Silverman, H. S. (1997) Am. J. Physiol. 273, C37–C44. 9. Moravec, C. S., and Bond, M. (1991) Am. J. Physiol. 260, H989 – H997. 10. Schreur, J. H. M., Figueredo, V. M., Miyamae, M., Shames, D. M., Baker, A. J., and Camacho, S. A. (1996) Biophys. J. 70, 2571–2580. 11. Trollinger, D. R., Cascio, W. E., and Lemasters, J. J. (1997) Biochem. Biophys. Res. Commun. 236, 738 –742. 12. Moravec, C. S., Desnoyer, R. W., Milovanovic, M., Schluchter, M. D., and Bond, M. (1997) Am. J. Physiol. 273, H1432–H1439.
558
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
13. Horikawa, Y., Goel, A., Somlyo, A. P., and Somlyo, A. V. (1998) Biophys. J. 74, 1579 –1590. 14. Del Nido, P. J., Glynn, P., Buenavrntura, P., Salama, G., and Koretsky, A. P. (1998) Am. J. Physiol. 274, H728 –H741. 15. Griffiths, E. J., Lin, H., and Suleiman, M.-S. (1998) Biochem. Pharmacol. 56, 173–179. 16. Crompton, M. (1990) in Calcium and the Heart (Langer, G. A., Ed.), pp. 167–198. Raven Press, New York. 17. Gunter, T. E., and Pfeiffer, D. R. (1990) Am. J. Physiol. 258, C304 –310. 18. Cox, D. A., and Matlib, M. A. (1993) J. Biol. Chem. 268, 938 – 947. 19. Bernardi, P., Broekemeier, K. M., and Pfeiffer, D. R. (1994) J. Bioenerg. Biomembr. 26, 509 –517. 20. Halestrap, A. P., Griffiths, E. J., Connern, C. P., and Kerr, P. M. (1997) Mol. Cell Biochem. 174, 167–172. 21. Leisey, J. R., Grotyohann, L. W., Scott, D. A., and Scaduto, R. C. (1993) Am. J. Physiol. 255, C162–C168. 22. Ventura, C., Ferroni, C., Flamigni, F., Stefanelli,C., and Caprogrossi, M. C. (1994) Am. J. Physiol. 267, H587–H592. 23. Kessar, P., and Crompton, M. (1981) Biochem. J. 200, 379 –388.
24. Griffiths, E. J., Stern, M. D., and Silverman, H. S. (1994) J. Mol. Cell Cardiol. 26, CLIV. 25. Bers, D. M. (1997) Bas. Res. Cardiol. 92, S1, 1–10. 26. Bassani, J. W. M., Bassani, R. A., and Bers, D. M. (1994) J. Physiol. 476, 279 –293. 27. Rizzuto, R., Pinton, P., Carrington, W., et al. (1998) Science 280, 1763–1766. 28. Peng, T. I., and Greenan, J. T. (1998) Mol. Pharmacol. 53, 974 –980. 29. Balaban, R. S., Kantor, H. L., Katz, L. A., and Briggs, R. W. (1986) Science 232, 1121–1123. 30. McCormack, J. G., Halestrap, A. P., and Denton, R. M. (1990) Physiol. Rev. 70, 391– 425. 31. Hansford, R. G. (1994) J. Bioenerg. Biomemb. 26, 495–508. 32. Heineman, F. W., and Balaban, R. S. (1993) Am. J. Physiol. 264, H433–H440. 33. Brandes, R., and Bers, D. M. (1997) Circ. Res. 80, 82– 87. 34. Griffiths, E. J., Wei, S.-K., Haigney, M. C. P., Ocampo, C. J., Sern, M. D., and Silverman, H. S. (1997) Cell Calcium 21, 321–329. 35. Das, A. M., and Harris, D. A. (1990) Cardiovasc. Res. 24, 411– 417.
559