hr. J. Eiochem.Vol. 22. No. 4, pp. 411-417, Printed in Great Britain. All rights reserved

1990

0020-71IX/90 $3.00+ 0.00 Copyright Q 1990Pergamon Press plc

RESPIRATION OF MITOCHONDRIA ISOLATED FROM DIFFERENTIATED AND UNDIFFERENTIATED HT29 COLON CANCER CELLS IN THE PRESENCE OF VARIOUS SUBSTRATES AND ADP GENERATING SYSTEMS T. GAUTHIER, C. DENIS-POUXVIELand J. C. MURAT* Institut de Physiologic, INSERM U.317, UniversitC Paul Sabatier, 2, rue F. Magendie, 31400 Toulouse, France [Tel. 61 55 65 391 (Received 31 August 1989) Abstract-l.

Oxygen consumption was investigated in two cultured subpopulations of either undifferentiated (WC+ cells) or differentiated (Glc cells) HT29 colon cancer cells and in the corresponding isolated mitochondria. In Glc+ cells, a decrease of the respiration is induced by the presence of glucose (Crabtree effect), whereas it is not the case in Glc- cells. 2. The oxidative phosphorylation rate of Glc- mitochondria is found to be much higher than that of Glc+ mitochondria, due to a higher efficiency to oxidize glutamine, glutamate, 2_oxoglutarate, succinate or malate. 3. In both types of mitochondria, respiration can be supported by the ADP formed by adenylate kinase or nucleotide diphosphate kinase, and, although to a lesser extent in Glc- mitochondria, by hexokinase. 4. Glc+ cells are characterized by a low respiration capacity and a high glycolytic flux leading to the Crabtree effect. Glc- cells are characterized by a better correlation between a moderate glycolytic flux and a high respiratory capacity.

INTRODUCTION HT29 cells derived from a human colon adenocarcinema were first established in culture by Fogh and Trempe (1975). HT29 standard cells cultured in

glucose containing medium are undifferentiated (referred as Glc+ cells) and share a common trait with other cancer cells in that the rate of glucose conversion into lactic acid is impressively high (Rousset et al., 1984; Zweibaum et al., 1985; Franklin et al., 1988; Gauthier et al., 1989a; Babia et al., 1989). Long term culture in a glucose free medium results in an “enterocyte-like” differentiation of the cells (referred as differentiated or Glc- cells) which slowly utilize glucose when this sugar is added for a short time in the medium (Zweibaum et al., 1985).

In numerous tumor cells, it has been reported that a wide portion of hexokinase is bound to the mitochondrial outer membrane [for review, see Pedersen (1978)] and that this peculiar location of the enzyme could enhance both the oxidative phosphorylation (Bessman and Geiger, 1980) and the glycolytic flux (Bustamante and Pedersen, 1977). In both Glc+ and Glc- HT29 cells, 75% of the hexokinase is found to be bound to the mitochondria (Gauthier et al., 1989a). From previous ultrastructural studies, we raised the hypothesis that the glycolytic abnormality of the standard HT29 Glc+ cells could be due to a

defect at the level of the mitochondrial membranes, impairing the normal channeling of ATP and ADP between the bound hexokinase and the respiratory chain (Denis-Pouxviel et al., 1987). In a second work (Gauthier et al., 1989b), the rates of glucose phosphorylation by bound hexokinase were investigated in mitochondria isolated from both Glc+ and Glc- cells, either with added ATP (exogenous ATP) or with ATP generated by the mitochondria supplied with ADP and succinate (endogenous ATP). We came to the conclusion that, in mitochondria from Glc’ cells, exogenous ATP was more efficient than endogenous ATP in supporting hexokinase activity. Furthermore, the endogenous ATP used by hexokinase of Glc’ mitochondria was found to be mainly generated by adenylate kinase. The situation was found to be different in mitochondria from Glc- cells, where oxidative phosphorylation appeared to be an efficient supplier of ATP to the bound hexokinase. In the present paper, we report on the efficiency of different substrates to promote respiration in intact cells or in isolated mitochondria, comparing standard Glc+ cells to differentiated Glc- cells. We also report on a functional relationship between mitochondrial kinases (hexokinase, adenylate kinase, nucleotide diphosphate kinase) and respiration in both models. MATERIALS AND METHODS

*To whom all correspondence should be addressed. Abbreviations: CAT, carboxyatractyloside; CCCP, carbonylcyanide m-chlorophenylhydrazone; RCR, respiratory control ratio.

Chemicals and cell culture Dulbecco’s modified Eagle medium was from Eurobio (Paris, France). Fetal-calf serum was from Boehringer (Mannheim, F.R.G.). All chemicals were purchased from 411

412

T. GAUTHIER et al.

Sigma (St Louis, MO., U.S.A.) or Serva (Heidelberg, F.R.G.) and were of the highest purity grade. The standard HT29 Glc+ cells and the differentiated HT29 Glc subpopulation were obtained from Dr A. Zweibaum (INSERM U. 178, Villejuif, France). The cells were routinely cultured in plastic Petri dishes 60mm dia, with Dulbecco’s modified Eagle medium containing either 25 mM D-glucose (undifferentiated Glc+ cells) or no glucose (differentiated Glc- cells) and supplemented with 10% dialyzed fetal-calf serum. Medium was changed every day for Glc- cells and every two days for Glc+ cells and 24 hr before experiments, Glc- cells were cultured in the presence of 25 mM glucose. Isolation qf mitochondria Cells in the stationary phase, i.e., at day 12 for the standard cells and at day 22 for the Glc- cells, were harvested with 2 ml of Ca’+ and Mg’+ free 0.13 M phosphate buffer (pH 7.4) containing 0.6 mM EDTA and were centrifuged at 500 g for 3 min. The cell pellets collected from 30 dishes (ca 150 mg of cell proteins) were pooled and resuspended in 10 ml of ice-cold isolation medium containing 0.21 M mannitol, 0.07 M sucrose, 2 mM Hepes (pH 7.4) and I mM EGTA. The suspension was homogenized with a glass Potter and mixed with 250 mg of DEAE-cellulose in 5 ml of isolation medium, as preconized by Lawrence and Davies (1986). Nagarse was then added at a final concentration of 5 pg/mg of cell protein (Aubert-Foucher et al., 1985). The mixed solution was stirred on ice for 7 min, then the protease action was stopped by adding 15 ml of isolation medium containing 0.1% bovine serum albumin (medium A). The suspension was centrifuged IOmin at 700g. The supernatant was collected and centrifuged 10 min at 7800g. The sediment was resuspended in IOml of medium A and centrifuged for 4min at 700g. The resulting supernatant was centrifuged 10min at 11.3OOg and the pellet was resuspended in 17 ml of medium A. A 4 ml sample of Percoll was added and the whole mixture was centrifuged for 30 min at I 12,000 g. The mitochondrial fraction was collected and thrice washed with medium A. The final pellet was resuspended in 1 ml of respiration medium (see below). Isolated mitochondria from the two cell subpopulations remained well coupled for at least 2 hr. The integrity of isolated mitochondria was controlled by measuring the activity of soluble enzymes (adenylate kinase and glutamate dehydrogenase) by reference to the activity of membrane integrated succinate dehydrogenase in the cell homogenate and the isolated mitochondrial fraction. In this respect, the method and results were identical to those reported in a previous paper (Denis-Pouxviel ct al., 1987). Measurement of osygen consumption Respiration of intact cells was determined with a Clark type oxygen electrode at 22°C according to Estabrook (1967) in 2ml of a Krebs medium containing 25 mM NaHCO,, 118 mM NaCI, 4.8 mM KCI, 1.2 mM KH,,PO,, 1.2 mM MgSO, and IOmM Hepes (pH 7.5). Respiration of isolated mitochondria was measured in a similar way in 2 ml of an isotonic medium containing 0.21 M mannitol, 70 mM sucrose, 15 mM Hepes (pH 7.4) 0.1% bovine serum albumine, 8 mM MgCl, and 4 mM KH,PO,. Assay of en:~mes actirir) Hexokinase activity (EC 2.7.1.30) was determined according to Biicher et al. (1964) by mixing 20 pl of mitochondrial fraction with 0.98 ml of 50mM triethanolamine buffer (pH 7.6) containing 8 mM MgSO,, 5 mM EDTA, 2 mM glucose. 0.6 mM NAD and 2 U/ml glucose 6-phosphate and dehydrogenase from Leuconostoc mesenteroides. The reaction was started by adding 2 mM ATP and the reduction of NAD was followed at 30°C for 15 min at 340 nm. Adenylate kinase (EC 2.7.4.3) was estimated according to Biicher et al. (1964): 10 nl of a diluted mitochondrial

fraction was added to 1 ml of 50 mM triethanolamine buffer (pH 7.6) containing 5mM EDTA, 75 mM CKl, 8 mM MgSO,, I mM ATP, 0.5 mM PEP, 0.3 mM NADH, 2 PM rotenone, 0. I ngglml oligomycin, 15 U/ml pyruvate kinase and 50 U/ml lactate dehyhrogenase. The- reaction was started by adding 1.5 mM AMP. The reaction was run at 30°C for 10 min and change in optical density was recorded at 340nm. Glutamate dehydrogenase (EC 1.4.1.3) was assayed as proposed by Brdiczka et al. (1985): lOpI of a diluted mitochondrial fraction was added to I ml of 50mM triethanolamine buffer (pH 7.6) containing 5 mM EDTA, and 40 mM (NH& SO,, 1.6 mM ADP, 0.3 mM NADH 2 mM 2-oxoglutarate. The oxidation of NADH was followed at 30°C for 10 min. at 340 nm. Succinate dehydrogenase (EC I .3.99.19) was measured according to Brdiczka et ul. (1985): 20 ~1 of mitochondrial solution was mixed with I ml of 100 mM phosphate buffer (pH 7.6) containing 0.1 mM cytochrome C. I mM KCN and 0.1 mM phenazin metasulfate. The reaction was started by adding 25 mM succinate. Change in optical density at 550 nm was read for I5 min at 30’ C. in a double beam spectrophotometer, with a reference cuvette which did not contain succinate. Nucleotide diphosphate kinase (EC 2.7.4.6.) activity was measured spectrophotometrically at 340 nm by mixing 20 ~1 of mitochondria suspension with 960~1 of 50mM triethanolamine buffer -(pH 7.6) containing 5 mM EDTA, 8 mM MgSO,, 20mM glucose, 0.5mM ADP, 7.5 mM AMP, 0.6 mM NAD. I U/ml of hexokinase and 1 U/ml of glucose 6-P dehydrogenase; the reaction was initiated by adding 20 ul of a 25 mM GTP solution and run at 30°C. One milh-unit (mu) is defined as the amount of enzyme converting one nanomol of substrate per minute at 30°C. Protein content was determined by the method of Bradford (1976) using bovine serum albumine as standard. All data were analyzed with an Apple He computer using the Biodata Handling uroarams written bv Barlow (1983). Paired t test was usedfor comparing the values, signi&ande being admitted for P < 0.05. RESULTS

Respiration

of intact cells

Figure 1 shows the oxygen consumption of freshly harvested HT29 cells suspended in Krebs medium (see Material and Methods). It was observed that addition of 200 ng/ml antimycin A, which is a specific inhibitor of mitochondrial electron transport, totally inhibited the cellular respiration and that the basal oxygen uptake was greatly stimulated by addition of the protonophoretic uncoupling agent carbonylcyanide m -chlorophenyl (CCCP) at the optimal concentration of 1 ,uM. This ascertains that the respiration observed with the intact HT29 cells is entirely due to the activity of the mitochondrial respiratory chain controlled by the ApH+ potential. The respiration rate of the undifferentiated Glc+ cells, measured in the absence of added substrate, is not significantly modified by the addition of 2.5 mM glutamine (Fig. 2). By contrast, as seen in Figs 1 and 2, the addition of 10 mM glucose induces a 30% decrease in the rate of oxygen consumption, which is known as Crabtree effect. Simultaneous addition of glucose and glutamine leads to an inhibition of the respiration similar to that observed with glucose alone. Identical results were obtained when Glc- cells were deprived of glucose for 8 hr before the measurement (not shown).

Respiration of HT29 colon cancer cells

25 nmolo2

I imin crab. 9&l.

QLC- CELLS

26 nmdo,

413

L mill

Fig. 1. Recording of oxygen consumption by differentiated and undifferentiated HT29 cells. Respiration was started by addition of cell suspension (5-7 mg prot. in 1ml) to I ml of Krebs medium. Glucose (lOmM), CCCP (1 MM) and antimycin A (200 ng!ml) were sequentially added at the indicated times. One representative experiment is shown. Numbers in parentheses express oxygen consumption in nmof Orjminjmg prot.

Differentiated Glc- HT29 cells, cultured for 24 hr in a glucose containing medium can be considered as the closest to normal cells. In this model, the respiration rate measured in the absence of added substrate is as high as the respiration rate measured with Glc” cells. However, the oxygen consumption is only slightly affected by addition of either glucose of gtutamine. It is noteworthy that the uncoupled respiration rate, i.e. measured in the presence of 1 JAM CCCP, is signi~cantly higher in the Glc- cells than in the Glc+ cells suggesting a higher efficiency of mitochondrial respiratory chain in the differentiated Glc- cells. Respiration of isolated mitochondria

We have studied the oxygen consumption of mitochondria isolated from Glc’ cells or from Glc- cells

6xcP

Fig. 2. Respiration rates of differentiated and undifferentiated harvested cells. Oxygen consumption was measured in Glc+ (m) or Glc cells (a), as described in Materials and Methods. Concentrations of glucose and glutamine added were respectively 10 and 2.5 mM. A I PM sample of CCCP was used to uncouple the oxygen uptake. Reported values are the mean * SE of 15 separate experiments. A significant difference at P c: 0.05 is indicated by an asterisk when comparing the 2 cell subpopufations and by a cross when comparing values with the basal respiration (endo. subs.+ndogcnous substrates. i.e. no substrate added). glut. t- gluta.-glucose + glutamine.

cultured for 24 hr in the presence of glucose. In both cases. the oxygen uptake was found to be greatly inhibited by adding 1 .LIM carboxyatractyloside (CAT) (which irreversibly inhibits the ADP/ATP translocator) or 10 pg/ml oligomycin (which blocks ATP synthase) and to be greatly increased by addition of 5 PM CCCP, indicating that the isolated mit~hond~a were originally weIl~~up1~ (Fig. 3). The affinity of the respiratory chain for ADP was determined in the presence of 5 mM succinate and increasing ADP concentrations from IO to 120 PM: as reported in Table 1. the K,,, for ADP, the ADP/O ratio and the respiratory control ratio (RCR) were found to be similar in the two types of mitochondria. The ability of various substrates to support respiration was investigated using either phosphorylating mitochondria (i.e. in the presence of 5OOpM ADP; Fig. 4A) or non-phosphorylating mitochondria (i.e. in the presence of 5 PM CCCP; Fig. 4B). In both experimental conditions and with each substrate tested (glutamine, glutamate, 2-oxogiutarate, succinate, malate), mit~hond~a from Glc- cells display higher respiration rates than mit~hondria from Glc* cells. The most pronounced difference is found with glutamine. These results suggest that the difference in respiration rates between the two types of mitochondria is due to a difference in the ability to oxidize metabolites rather than in the ability to phosphorylate ADP. This is also supported by the decrease in succinate dehydrogenase and glutamate dehydrogenase activities observed in the mitochondria from Glc- cells as compared to mitochondria from Glcx cells (137.1 + 17.8 vs 203.1 + 10.5 mU/mg protein for succinate dehydrogenase and 129.4 i 12.8 vs 261.3 I 43.7 mU/mg protein for glutamate dehydrogenase).

T. GAUTHIER et al.

414 SUCCLNATE ADP Mltochondrla from Ok+ call8

127.51

25 nmd0,

L

J

ANTlMYClN

A

i

lmin

1154.21

L J

25 nmolo, lmin

ANTIMYCIN

A

Fig. 3. Recording of oxygen consumption by mitochondria isolated from differentiated and undifferentiated HT29 cells. Respiration was started by addition of mitochondrial suspension (0.5 mg prot. in 0.1 mi) to 1.9 ml of isotonic medium containing 2 PM rotenone (see Materials and Methods). Succinate (5 mM), ADP (0.5 mM), CAT (1 PM), CCCP (5 PM) and antimycin A (5 pg/ml) were sequentially added at the indicated times. One representative experiment is shown. Numbers in brackets express oxygen consumption in nmoles O,/min/mg prot, Under phosphorylating conditions (Fig. 4A), succinate and 2-oxoglutarate appear to be the most potent metabolites for promoting oxygen uptake whereas the precursors of these metabolites, gIutamate and glutamine, are less efficient, the difference being more pronounced in mitochondria from Glc+ cells. It is

noteworthy that 2-oxoglutarate can enter the mitochondria despite the absence of any added malate, as also observed by Moreadith and Lehninger (1984) in Table I. Respiratory indices Glc’ mitochondria

Glc- mitochondria

ADPi ratio I .47 * 0.09 NS 1.57 + 0.14 RCR 3.53 IO.31 NS 4.28 ri: 1.22 16.8 + 8.8 NS 12.9 + 4.6 K, ADP (PM) The oxygen consumption was measured with isolated mitochondria and in presence of 5 mM succinate and 2pM rotenone, as described in Materials and Methods. ADPi is the ratio of nmoles ADP added in the medium over the corresponding nmoles of oxygen consumed by mitochondria. Respiratory Control Ratio (RCR) is obtained by dividing the rate of 0, consumption in state 3 by the rate of O? consumption in state 4. Results are the mean I SE from 4 separate experiments (NS = not significantly different).

EhrIich tumor mitochondria. Endogenous malate might be sufficient for allowing the early entry of 2-oxoglutarate, which, in turn, could generate more malate through the Krebs cycle; this is in agreement with the 2-3 min lag-period observed before reaching the maximum velocity. In both types of mitochondria, 0.5 mM pyruvate was found to poorly stimulate the respiration. From the respiration rates and the related ADP/O ratio obtained in the presence of succinate, it can be calculated that mitochondria from Glc- cells might produce ATP 1.6-times faster than mitochondria from Glc+ cells (213.5 k21.3 vs 131.4f 14.7nmol ATP~min/mg prot.). When mitochondria from Glc+ cells are uncoupled by the addition of CCCP (Fig. 4B), respiration supported by glutamine is twice lower than respiration supported by glutamate. On the contrary, in mitochondria from Glc- cells, no significant difference is found between the uncoupled respirations supported by either glutamine or glutamate. This result suggests that, unlike what occurs in the Gic- cells, the transformation of glutamine in an

Respiration of HT29 colon cancer cells

415

Role of mitochondrial kinases in supplying ADP for respiration *

80

The ADP necessary to the oxidative phosphorylation can be generated in the inte~embrane space either by adenylate kinase (supplied with ATP and AMP) or by nucleotide diphosphate kinase (supplied with ATP and UDP) or at the mitochondrial surface by bound hexokinase (supplied with glucose and ATP). As seen in Table 2, adenylate kinase and nucleotide diphosphate kinase are both more active in mito~hondria from differentiated Glc- ceils than in mitochondria from Glc+ cells. By contrast, as already reported (Gauthier et al., 1989b), the hexokinase activity is twice more elevated in mitochondria from Glc+ cells. Table 2 shows that when ADP is provided by adenylate kinase or nucleotide diphosphate kinase, higher rates of respiration are observed than when ADP is directly added in the medium; this is valid with both types of mitochondria. When ADP is provided by hexokinase, oxygen consumption reaches 94% of that observed with added ADP when using Glc+ mitochondria; this percentage drops to 49% when using Glc- mitochondria. No stimulation of respiration was obtained when adding ATP and glycerol, which is probably due to the absence of glycerol kinase bound to the mitochondria (not shown).

+

60

DISCUSSION 0 RIIE

#mi?tgiiime~~~

nwie

!a-=“.

Fig. 4. Respiration rate of mitochondria isolated from Glc’ and Glc- cells. Oxygen consumption was measured in phosphorylating mitochondria (A) or in uncoupled mitochondria (B) isolated from Glc+ (m) or Glc- cells (tZ). The concentration of CCCP added was 5 FM. When succinate was used, 2 FM rotenone was added in the medium. The concentration of tested metabolites was 5 mM except for pyruvate (0.5 mM). Reported values are the mean f SE of 5 separate experiments. *Indicates a significant difference at P

Respiration of mitochondria isolated from differentiated and undifferentiated HT29 colon cancer cells in the presence of various substrates and ADP generating systems.

1. Oxygen consumption was investigated in two cultured subpopulations of either undifferentiated (Glc+ cells) or differentiated (Glc- cells) HT29 colo...
916KB Sizes 0 Downloads 0 Views