Journal list menu

Volume 263, Issue 2 p. 495-501
Free Access

Effect of glucose and deoxyglucose on the redistribution of calcium in Ehrlich ascites tumour and Zajdela hepatoma cells and its consequences for mitochondrial energetics

Further arguments for the role of Ca2+ in the mechanism of the Crabtree effect

Lech Wojtczak

Lech Wojtczak

Nencki Institute of Experimental Biology, Warsaw, Poland;

Search for more papers by this author
Vera V. Teplova

Vera V. Teplova

Nencki Institute of Experimental Biology, Warsaw, Poland;

Institute of Theoretical and Experimental Biophysics, Russian Academy of Sciences, Pushchino, Russian Federation

Search for more papers by this author
Krystyna Bogucka

Krystyna Bogucka

Nencki Institute of Experimental Biology, Warsaw, Poland;

Search for more papers by this author
Aneta Czyż

Aneta Czyż

Nencki Institute of Experimental Biology, Warsaw, Poland;

Search for more papers by this author
Agnieszka Makowska

Agnieszka Makowska

Nencki Institute of Experimental Biology, Warsaw, Poland;

Search for more papers by this author
Mariusz R. Więckowski

Mariusz R. Więckowski

Nencki Institute of Experimental Biology, Warsaw, Poland;

Search for more papers by this author
Jerzy Duszyński

Jerzy Duszyński

Nencki Institute of Experimental Biology, Warsaw, Poland;

Search for more papers by this author
Yuri V. Evtodienko

Yuri V. Evtodienko

Institute of Theoretical and Experimental Biophysics, Russian Academy of Sciences, Pushchino, Russian Federation

Search for more papers by this author
First published: 25 December 2001
Citations: 42
L. Wojtczak, Nencki Institute of Experimental Biology, Pasteura 3, 02-093 Warsaw, Poland. Fax: + 48 22 822 5342, Tel.: + 48 22 659 3072, E-mail: [email protected]

Abstract

The distribution of Ca2+ in intact cells was monitored with fluorescent probes: fura-2 for cytosolic [Ca2+] and rhod-2 for mitochondrial [Ca2+]. It was found that in neoplastic cells, such as Ehrlich ascites tumour and Zajdela hepatoma, but not in non-malignant cells, such as fibroblasts, glucose and deoxyglucose elicited release of Ca2+ from endoplasmic reticulum stores and an increase in Ca2+ concentration in the cytosol. Parallel to this, a decrease in the rate of Ca2+ extrusion from the cell and an enhanced uptake of Ca2+ by mitochondria were observed. The increase in mitochondrial [Ca2+] was accompanied by an increase in the mitochondrial membrane potential and the reduction state of nicotinamide nucleotides. F1Fo-ATPase in submitochondrial particles of Zajdela hepatoma was strongly inhibited in the presence of micromolar Ca2+ concentrations, whereas this activity in submitochondrial particles from rat liver appeared to be less sensitive to Ca2+. Indications of glycosylation of Ehrlich ascites tumour cell proteins were also obtained. These data strengthen the proposal [Bogucka, K., Teplova, V.V., Wojtczak, L. and Evtodienko, Y. V. (1995) Biochim. Biophys. Acta1228, 261–266] that the Crabtree effect is produced by mobilization of cell calcium, which is subsequently taken up by mitochondria and inhibits F1Fo-ATP synthase.

Abbreviations

  • [Ca2+]c
  • cytosolic Ca2+ concentration
  • [Ca2+]m
  • intramitochondrial Ca2+ concentration
  • TMRM
  • tetramethylrhodamine methyl ester
  • CCCP
  • carbonyl cyanide m-chlorophenylhydrazone.
  • Many kinds of tumour exhibit an unusual reaction towards glucose: addition of millimolar concentrations of glucose to the culture medium results in a decrease in the cell respiration rate. This behaviour, known by the name of its discoverer as the Crabtree effect [1], has also been observed in some non-malignant cells and tissues, such as spermatozoa [2,3], proliferating thymocytes [4], intestinal mucosa [5] and mammalian embryos at their very early developmental stage [6]. Common features of all these tissues, both malignant and normal, are a high proliferation rate and/or a high glycolytic rate. Although the Crabtree effect has been known for seven decades (for an early review see [7]), its molecular mechanism is not fully understood. First attempts to explain it pointed to the competition between respiration and glycolysis for precursors of ATP production, i.e. ADP [8,9] and inorganic phosphate [10–12]. However, this explanation is not satisfactory because deoxyglucose, which is phosphorylated in the cytoplasm but not further metabolized and therefore does not generate ATP, produced even stronger inhibition of respiration than did glucose [13] and phosphate transport into the cell has been shown to be fast enough [13]. Other possible explanations of the mechanisms of the Crabtree effect were a shift in intracellular pH [7], a change in permeability properties of the inner mitochondrial membrane [14], a specific isoenzyme pattern of the glycolytic pathway and regulatory behaviour of key enzymes of this pathway [15], and specific topography of enzymes in rapidly growing tumours, in particular membrane-bound hexokinase [16]. However, none of these suggestions provides a complete explanation of how glucose inhibits oxygen consumption in these tissues.

    We have shown previously [17,18] that glucose and 2-deoxyglucose produce an increase in the concentration of free Ca2+ ions in the cytoplasm of Ehrlich ascites tumour cells and that this happens mainly at the expense of Ca2+ stores in the endoplasmic reticulum. Subsequently, it was shown [19] that, in digitonin-permeabilized Ehrlich tumour cells, elevated Ca2+ concentrations resulted in a complete block of ATP synthesis by mitochondria without the loss of their membrane potential. Isolated Ehrlich ascites tumour mitochondria, when preloaded with Ca2+, exhibited a decrease in the rates of both ATP synthesis and hydrolysis [20], apparently because of association of the natural inhibitory subunit with the F1Fo-ATPase complex. These results prompted us to propose the following sequence of events eventually leading to inhibition of respiration as a result of glucose supplementation: (a) glucose elevates the cytoplasmic concentration of Ca2+; (b) this elicits an increased accumulation of Ca2+ in the mitochondria; (c) loading of mitochondria with Ca2+ leads to an enhanced association of the inhibitory subunit with the F1Fo complex, which results in (d) inhibition of coupled respiration [20] (reviewed [21,22]).

    This scheme, although logical, has several unsolved points. First, it has not been experimentally proven that a glucose-induced increase in cytosolic Ca2+ concentration ([Ca2+]c) results in accumulation of Ca2+ in the mitochondria. The participation of mitochondria in the regulation of [Ca2+]c after the addition of deoxyglucose has been indicated only indirectly [23]. Secondly and most importantly, the mechanism by which glucose and deoxyglucose increase [Ca2+]c remains unclear. And thirdly, increased [Ca2+]c as a result of glucose supplementation has been documented for Ehrlich ascites tumour cells only and therefore cannot be proposed so far as a general phenomenon responsible for the Crabtree effect. The aim of the present investigation was to clarify these points.

    Materials and methods

    Ehrlich ascites tumour was cultivated in Swiss albino mice and harvested as described previously [19,20]. Zajdela hepatoma was cultivated in the peritoneal cavity of Wistar rats and harvested 5–6 days after inoculation as described for Ehrlich ascites cells. Human skin fibroblasts were grown for three to six passages in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum.

    Ehrlich ascites tumour mitochondria were isolated as described previously [24] and rat liver mitochondria by the standard procedure [25]. Submitochondrial particles were obtained by sonication as described previously [20].

    The standard incubation medium contained 140 mm NaCl, 5 mm KCl, 1 mm MgCl2, 1.5 mm CaCl2, 10 mm Hepes/NaOH (pH 7.6) and 3 mm l-glutamine. Unless indicated otherwise, the experiments were performed at 38 °C.

    For the determination of [Ca2+]c the cells were loaded with fura-2 [17,26] and examined in the ratio mode [27] at 340/380 nm excitation wavelength and 510 nm emission wavelength in a Shimadzu RF5000 spectrofluorimeter (Tokyo, Japan). For estimation of intramitochondrial [Ca2+] ([Ca2+]m), the cells were loaded with dihydro-rhod-2 [28], and the fluorescence was recorded at 568 nm and 590 nm excitation and emission wavelengths, respectively. For measurement of the mitochondrial membrane potential (ΔΨ), the cells were loaded with tetramethylrhodamine methyl ester (TMRM) and examined at 556 nm and 576 nm excitation and emission wavelength, respectively [29]. Reduction of nicotinamide nucleotides was estimated fluorimetrically at 340 and 450 nm excitation and emission wavelengths, respectively.

    Cell respiration was measured with a Clark oxygen electrode. ATPase was determined spectrophotometrically in the ATP-regenerating system, essentially by the method of Pullman et al. [30] as described previously [20]. ATP and ADP contents of the cells were measured fluorimetrically by enzymatic assays [31] in neutralized perchloric acid extracts.

    Binding of deoxyglucose to cell proteins (glycosylation) was measured using deoxy[14C]glucose. Ehrlich ascites tumour cells corresponding to 12 mg protein·mL−1 were suspended in the incubation medium in which glutamine was replaced by 10 mm pyruvate and CaCl2 was made 1 mm. Deoxy[14C]glucose was added to a concentration of 10 μm (corresponding to 1.1 × 106 dpm·mL−1), and, after various incubation periods at room temperature, 1.0 mL aliquots were taken and quenched with 0.7 m perchloric acid. Precipitated protein was washed several times with 0.05 m perchloric acid, solubilized and counted for radioactivity.

    Chemicals of the following origin were used: fura-2 acetoxymethyl ester, rhod-2 acetoxymethyl ester and TMRM from Molecular Probes (Eugene, OR, USA); thapsigargin from Sigma (St Louis, MO, USA); ionomycin from Calbiochem (La Jolla, CA, USA); hexokinase, glucose-6-phosphate dehydrogenase, pyruvate kinase, lactate dehydrogenase and phosphoenolpyruvate (for ADP and ATP measurements) from Boehringer (Mannheim, Germany). 2-Deoxy[1-14C]glucose was from Amersham (UK).

    The figures show representative experiments out of at least three of each type.

    Results

    Effect of glucose and deoxyglucose on intracellular redistribution of Ca2+

    In order to ascertain whether changes in [Ca2+]c induced by glucose and deoxyglucose [17,18] are accompanied by changes in [Ca2+]m, Ehrlich ascites tumour cells were loaded with either fura-2 or rhod-2. The latter probe is taken up by energized mitochondria (because of its positive charge) and therefore primarily monitors Ca2+ concentration in the mitochondrial compartment of intact cells. Mitochondrial selectivity of this probe is enhanced when it is applied to the cells in the non-fluorescent reduced form (dihydro-rhod-2), which is then oxidized to the fluorescent dye inside mitochondria [28]. Using these procedures, we could follow changes in both [Ca2+]c and [Ca2+]m in parallel samples.

    In agreement with previous observations [17,18], glucose and deoxyglucose produced in Ehrlich ascites tumour cells a pronounced increase in [Ca2+]c, followed by a decrease a few minutes later (Fig. 1, traces A and B). Also as expected, the uncoupler of oxidative phosphorylation, carbonyl cyanide m-chlorophenylhydrazone (CCCP), increased [Ca2+]c as the result of Ca2+ release from the mitochondria. The addition of oligomycin, an inhibitor of mitochondrial F1Fo-ATPase, protected cytosolic ATP against hydrolysis by uncoupled mitochondria and therefore enabled the undisturbed functioning of the Ca2+ pump in the endoplasmic reticulum. Assays with cells loaded with dihydro-rhod-2 showed that, under identical conditions, mitochondria accumulated large amounts of Ca2+, this accumulation being continued for several minutes, also during the phase of decreasing [Ca2+]c. Mitochondrial Ca2+ was rapidly released by CCCP (Fig. 1, traces C and D).

    Effect of glucose and deoxyglucose on the cytosolic and mitochondrial concentrations of Ca 2+ in Ehrlich ascites tumour cells.[Ca2+]c was monitored with fura-2 (traces A and B) and [Ca2+]m with rhod-2 (traces C and D). The following additions were made where indicated: glucose (Gluc), 10 mm; deoxyglucose (DOG), 10 mm; oligomycin (Oligo), 10 μg·mL−1; CCCP, 2.5 μm.

    The Ca2+ ionophore ionomycin, added to Ehrlich ascites cells at very low, submicromolar, concentration, produced an abrupt increase in [Ca2+]c as the result of the release of Ca2+ from intracellular stores, followed by a rapid decrease. It could be assumed that ionomycin at this low concentration increased the calcium permeability mainly of endoplasmic reticulum membranes (the surface of which is much higher than that of the plasma membrane), still enabling Ca2+ extrusion by the plasma-membrane-located calcium pump so that a new steady-state level of [Ca2+]c was established. The addition of glucose resulted in a lowering of the maximum [Ca2+]c produced by ionomycin, slowing down of Ca2+ extrusion and establishment of a new steady-state at a somewhat higher level (Fig. 2). A lower maximum [Ca2+]c produced by ionomycin in the presence of glucose is in agreement with the observation (Fig. 1A) that glucose causes redistribution of cellular Ca2+, presumably lowering the endoplasmic reticulum Ca2+ content. The rate of Ca2+ extrusion after the addition of ionomycin differed considerably from preparation to preparation. However, the slowing-down effect of glucose was reproducible and amounted to 34–46%.

    Effect of glucose on the rate of Ca 2+ extrusion from the cytosol of Ehrlich ascites tumour cells. The cells were incubated in the standard medium the pH of which was brought to 8.0, and Ca2+ concentration was increased to 3 mm. [Ca2+]c was monitored with fura-2. Glucose (Gluc) was added to a final concentration of 10 mm and ionomycin (IM) to 0.13 μm.

    Effect of [Ca2+]m on mitochondrial energy-coupling parameters

    Parallel to the rise in [Ca2+]m was an increase in mitochondrial membrane potential (ΔΨ) in Ehrlich ascites cells, as measured with TMRM (Fig. 3, traces A and B). Interestingly, the effect of deoxyglucose was usually greater than that of glucose. Oligomycin produced a further increase in ΔΨ, corresponding to the resting-state (State 4) level. Glucose and deoxyglucose also increased the reduction of nicotinamide nucleotides (Fig. 3, traces C and D). This was a result of changes in the redox state of both cytosolic and mitochondrial NAD(P). However, as it responded very rapidly to the addition of CCCP, it presumably corresponded mainly to the mitochondrial pool. For comparison, Fig. 3 also shows the well-known effects of glucose and deoxyglucose on cell respiration (the Crabtree effect, traces E and F).

    Effect of glucose and deoxyglucose on the mitochondrial membrane potential, redox state of nicotinamide nucleotides and respiration of Ehrlich ascites tumour cells.ΔΨ was monitored with TMRM, reduction of NAD(P) was followed fluorimetrically, and O2 uptake was measured with an oxygen electrode. All additions were as indicated in Fig. 1.

    To confirm that the observed increase in ΔΨ and the reduction state of nicotinamide nucleotides can be ascribed to elevated [Ca2+]c and to the subsequent uptake of Ca2+ by mitochondria, the increase in [Ca2+]c was also induced by thapsigargin, which is known to inhibit Ca2+ accumulation in the endoplasmic reticulum stores [32], and external ATP, which triggers Ca2+ release from these stores by the inositol trisphosphate-dependent mechanism [26]. It appeared that, in both cases, the increase in [Ca2+]c was accompanied or followed by an increase in ΔΨ and the reduction state of nicotinamide nucleotides (Fig. 4). The increase in ΔΨ and the reduction of NAD(P) produced by thapsigargin could be interpreted as being the result of inhibition of endoplasmic reticulum Ca2+-ATPase and thus an increase in the cytosolic ATP/ADP ratio. However, no such explanation is valid for cells doped with external ATP. Hence, an increase in the energy state of the mitochondria in this case must be interpreted as resulting from increased accumulation of Ca2+ by these organelles.

    Effect of thapsigargin and external ATP on cytosolic Ca 2+ concentration, mitochondrial membrane potential and redox state of nicotinamide nucleotides in Ehrlich ascites tumour cells. Measurements were made as described in the legend to Fig. 3, and the additions were as follows: thapsigargin (TG), 2 μm; ATP, 80 μm; CCCP, 2.5 μm.

    Similar pictures of the redistribution of Ca2+, the increase in ΔΨ and the reduction state of nicotinamide nucleotides on addition of deoxyglucose or glucose were obtained with Zajdela hepatoma cells (Fig. 5). It should be noted that mitochondria in isolated Zajdela hepatoma cells are almost completely deenergised and become energised only after the addition of deoxyglucose (trace B) or glucose (not shown), ΔΨ being further potentiated by oligomycin. Trace D shows, in addition, that NAD(P) reduction also occurred in the presence of 10 mm iodoacetic acid, i.e. under conditions of almost complete inhibition of glycolysis. It should be remembered that, in all experiments, the incubation medium contained glutamine (or pyruvate, as in Fig. 5D), so that mitochondria were supplied with sufficient substrate to generate ΔΨ and to reduce nicotinamide nucleotides independently of glycolysis.

    Effect of deoxyglucose or glucose on cytosolic Ca 2+ concentration (trace A), mitochondrial membrane potential (trace B) and redox state of nicotinamide nucleotides (traces C and D) in Zajdela hepatoma cells. Conditions were as in Figs 1 and 3. Cytosolic [Ca2+] (trace A) is expressed as the fluorescence ratio of fura-2 at 340 vs 380 nm excitation wavelength rather than in absolute terms. In trace D, the dashed line shows changes in NAD(P)H fluorescence in the presence of 10 mm iodoacetate.

    In non-malignant cells, i.e. human fibroblasts, glucose did not produce an increase in [Ca2+]c (not shown).

    ATP content of Ehrlich ascites tumour cells incubated in the medium containing pyruvate amounted to 22.8 ± 2.0 nmol·(mg protein)−1 (n = 4). The addition of 10 mm glucose to the medium did not change this level for up to 10 min. In contrast, the addition of 10 mm deoxyglucose resulted in a dramatic decrease in cellular ATP to 3 nmol·(mg protein)−1 after 1 min and practically to zero after 5 min.

    Effect of Ca2+ on mitochondrial ATPase

    It has been shown previously [20] that Ca2+ exerts a much stronger inhibitory effect on F1Fo-ATPase of Ehrlich ascites tumour cells than on that of rat liver. To check whether this high sensitivity to Ca2+ is a specific property of this particular tumour line or a more general phenomenon, the Ca2+ effect was next compared for mitochondrial ATPase of Zajdela hepatoma and normal rat liver. As shown in Fig. 6, 10 μm Ca2+ inhibited ATPase activity of submitochondrial particles from Zajdela hepatoma by 50%, which is essentially the same as for the enzyme in Ehrlich ascites tumour [20]. The same Ca2+ concentration decreased F1Fo-ATPase of rat liver submitochondrial particles by 15%.

    Effect of Ca 2+ concentration on ATPase activity in submitochondrial particles of Zajdela hepatoma. (●) Zajdela hepatoma; (○) rat liver.

    The Crabtree effect and the cellular Ca2+ pool

    To confirm further that the Crabtree effect is related to the release of Ca2+ from the endoplasmic reticulum stores into the cytosol, the effect of glucose on cell respiration was measured in Ehrlich ascites cells with different degrees of Ca2+ loading. Cells were isolated and loaded with fura-2 in Ca2+-free medium. Subsequently, these cells were incubated with 2 mm Ca2+ for various periods of time and the rate of Ca2+ release into the cytoplasm after addition of glucose was followed. In parallel samples, the total amount of Ca2+ accumulated in the stores was ascertained from the magnitude of its release after addition of ionomycin. The same preparations of the cells, not loaded with fura-2 but preincubated in Ca2+-free medium and subsequently with 2 mm Ca2+ for the same period of time, were used for respiration measurements. As shown in Fig. 7A, there was a correlation between the extent of inhibition of respiration (the Crabtree effect) and the increase in [Ca2+]c after the addition of glucose. Figure 7B shows that the longer the cells were incubated with 2 mm Ca2+, the more Ca2+ was accumulated in ionomycin-sensitive stores.

    Dependence of the magnitude of the Crabtree effect on the content of Ca 2+ stores in Ehrlich ascites tumour cells. (A) Inhibition of respiration (the Crabtree effect, ○) and increase in cytosolic [Ca2+] (●) induced by glucose after preincubation of the cells for various periods of time in the presence of Ca2+. (B) The magnitude of the Ca2+ stores depending on the preincubation time in the presence of Ca2+. The release of Ca2+ from the stores was induced by ionomycin (IM), 1.3 μm.

    A similar correspondence between the amount of Ca2+ released into the cytoplasm by glucose [17], the degree of store loading and the magnitude of the Crabtree effect in Ehrlich ascites tumour cells was also observed on increasing the pH of the incubation medium from 7 to 8 (not shown).

    Binding of deoxy[14C]glucose to cell proteins

    Figure 8 shows covalent binding of deoxy[14C]glucose to cell proteins. It is evident that this binding increases with time, is potentiated by borohydride and is almost completely abolished if the addition of deoxy[14C]glucose is preceded by the addition of 10 mm glucose.

    Binding of deoxy[ 14 C]glucose to proteins of Ehrlich ascites tumour cells. The experimental procedure was as described in Materials and methods. (○) Without reduction; (●) after reduction with borohydride; (▿) addition of deoxy[14C]glucose was preceded by the addition of 10 mm glucose.

    Discussion

    The present investigation shows that the increase in cytosolic Ca2+ concentration induced by glucose and deoxyglucose can be observed not only in Ehrlich ascites tumour cells [17,18] but also in Zajdela hepatoma, another malignant tissue of quite different origin, but not in non-malignant human fibroblasts. This result strengthens our previous assumption [21] that the increase in [Ca2+]c caused by glucose and deoxyglucose may be a general property of cancer cells, responsible for the characteristic decrease in malignant cell respiration in the presence of hexoses, known as the Crabtree effect [1]. We have also demonstrated a high sensitivity of Zajdela hepatoma F1Fo-ATPase to micromolar concentrations of Ca2+, similar to that of Ehrlich ascites tumour F1Fo-ATPase [20]. Finally, it has been experimentally documented that the increase in [Ca2+]c is, in fact, accompanied, or, rather, followed, by a pronounced increase in [Ca2+]m.

    This provides further strong support for the hypothesis [20,21] that the Crabtree effect is the result of the following sequence of events: (a) glucose or deoxyglucose elicits an increase in [Ca2+]c; (b) this is followed by an increased electrophoretic accumulation of Ca2+ in mitochondria; (c) elevated [Ca2+]m results in the inhibition of F1Fo-ATPase, presumably as the result of an increased association of the inhibitory subunit with the enzyme complex [20]; (d) the decrease in ATP synthesis results in an increase in ΔΨ, reduction of mitochondrial nicotinamide nucleotides and, eventually, a decrease in mitochondrial respiration. Increased ΔΨ and reduction of total cellular NAD(P) after the addition of glucose could be the effect of glycolysis and the competition for ADP between oxidative phosphorylation and glycolytic ATP synthesis. However, this explanation does not hold for the non-metabolizable glucose analogue, deoxyglucose, which causes production of ADP rather than its phosphorylation. It therefore seems likely that the changes produced by glucose are, at least partly, also due to the inhibition of mitochondrial ATP synthase (the F1Fo complex), thus forming the so called ‘oligomycin-like’ effect.

    The mechanism by which glucose and deoxyglucose increase [Ca2+]c remains unclear. However, the present work provides some indications. First, it was shown that a substantial amount of deoxy[14C]glucose covalently binds to cell proteins and that this is potentiated by borohydride (Fig. 8), suggesting a Schiff base formation and subsequent stable glycosylation. Binding of deoxy[14C]glucose was almost completely prevented if glucose was added first, indicating that glucose and deoxyglucose bind to the same sites. Secondly, extrusion of Ca2+ from the cytoplasm occurred much more slowly if glucose was present in the medium (Fig. 2). It has been demonstrated [33,34] that incubation of erythrocytes or erythrocyte membranes with glucose decreased the activity of membrane Ca2+-ATPase. González Flecha et al. [35] have subsequently shown that this is due to non-enzymatic glycation of the erythrocyte membrane Ca2+ pump. It is therefore tempting to speculate that Ca2+ pumps of malignant cells can also be inhibited by non-enzymatic and/or enzymatic glycosylation, thus changing the distribution of Ca2+ between the cytosol and endoplasmic reticulum stores as well as decreasing the extrusion of Ca2+ from the cell.

    Deoxyglucose (but not glucose) may also decrease removal of Ca2+ from the cytosol in a more trivial way, namely by slowing down the operation of the Ca2+ pumps in both the endoplasmic reticulum and the plasma membrane as the result of depletion of cytosolic ATP. This mechanism presumably operates in addition to that described above, so that the effects of deoxyglucose on Ca2+ increase in the cytosol, and its accumulation in mitochondria and changes in mitochondrial energy-coupling parameters are higher than those of glucose (Figs 1 and 3).

    Our preliminary experiments (M. R. Więckowski and L. Wojtczak, unpublished work) have shown that a decrease in cell respiration and an increase in mitochondrial membrane potential, i.e. effects similar to those observed after glucose addition, can also be produced in some malignant cells, including Ehrlich ascites tumour, by cyclosporin A, a potent blocker of the mitochondrial permeability transition pore [36,37]. On the other hand, recent results obtained by Brdiczka et al. [38] indicate that a membrane pore exhibiting all the properties of the mitochondrial permeability transition pore can be formed in phospholipid membranes by reconstituting the adenine nucleotide carrier, porin, cyclophilin and hexokinase. Glucose caused partial closure of this pore. In many kinds of malignant cells, including Ehrlich ascites tumour, a large proportion of the cell’s hexokinase is associated with the mitochondria [39], where it is bound within or close to the contact sites between the outer and the inner membranes [40,41]. It can therefore be speculated that the observed recoupling action of glucose and its analogue, deoxyglucose, in Ehrlich ascites tumour and Zajdela hepatoma may partly be due to closure of the mitochondrial permeability transition pore of these cells. This may be an additional mechanism of the Crabtree effect.

    The role of Ca2+ ions as mitochondria-targeted second messengers is now well recognized [42]. In most tissues, such as muscle and liver, Ca2+ increases cell aerobic metabolism by activating several mitochondrial primary dehydrogenases [43]. From the present work as well as from our previous investigations on the Crabtree effect [17–20,23] (reviewed in [21,22]), a different strategy can be proposed for malignant and highly glycolytic non-malignant cells. Here, in addition to activation of mitochondrial dehydrogenases, Ca2+ also produces inhibition of F1Fo-ATP synthase, the latter effect presumably being more pronounced than the former. The concerted action of the two mechanisms results in a considerable increase in mitochondrial membrane potential and redox state of nicotinamide nucleotides, which are very important for biosynthetic processes in these highly proliferating cells.

    Acknowledgement

    The invaluable technical assistance of Maria Buszkowska is greatly appreciated.