MAGNETIC RESONANCE IN MEDICINE 23,356-366 ( 1992)
13C NMR
Studies of Glucose Metabolism in Human Leukemic CEM-C7 and CEM-C1 Cells
JANF. M. POST,* ERNABAUM,AND EDWARDL. EZELL Department of Human Biological Chemistry & Genetics, University of Texas Medical Branch, Galveston, Texas 77550 Received May 2, 1990; revised December 3, 1990 Glucose metabolism of human leukemic cell lines CEM-C7 and CEM-C 1 was investigated in vivo by 13CNMR using "C-labeled glucose. Exact knowledge of glucose concentration, cell count, and cell viability of the cell suspensions made it possible to analyze glucose metabolism in detail. In both cell lines aerobic glycolysis accounts for virtually all glucose consumption. The use of D['3Cz]glucoseprovided a simple method to measure the glucose flux through the pentose phosphate pathway as 9% (CEM-C1 ) and 1 1 % (CEM-C7) of glucose channeled into glycolysis. The dexamethasone-sensitive CEM-C7 cells consume glucose at a rate about 50% higher than the dexamethasone-resistant CEM-CI cells. It is shown that this higher consumption correlates with a larger size of the CEM-C7 cells. Therefore in CEM cells the development of drug resistance does not seem to involve related changes in cell energetics. 0 1992 Academic Press, Inc. INTRODUCTION
It has long been known that malignant tissues have a high rate of aerobic glycolysis as compared to normal tissue ( 1 ). At first it was thought that this was a characteristic property of tumor cells due to an impairment in respiration, but later it was found that also normal cells in tissue culture produce lactate readily in the presence of air. Bissell et al. found that in both normal and virus transformed chick cells the ratio of lactic acid produced to C 0 2 evolved increased with increasing cell density (2). Also the pentose phosphate pathway became more prominent as cell density increased. However, under comparable conditions transformed cells always had a higher rate of (lactic acid produced)/( C 0 2 evolved) than normal cells. These results suggest that both malignancy and cell density play a role in increased aerobic glycolysis of mammalian cells in culture. The use of I3C-labeledcompounds for NMR studies is becoming increasingly popular to study various aspects of carbohydrate metabolism ( 3 ) . The reasons for this are twofold: ( 1 ) cell metabolism can be studied in vivo; (2) concentrations of several metabolites can be followed simultaneously. Important insights into bacterial anaerobic glycolysis ( 4 ) ,gluconeogenesis in rat liver cells ( 5 ) , or drug resistance in tumor cells (6, 7) have resulted from in vivo 13CNMR studies. In this paper we present I3C NMR results obtained with two human leukemic cell lines CEM-C7 and CEM-C 1. CEM-C7 is a wild-type cell line which is sensitive to the * To whom correspondence should be addressed. 0740-3 194/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reseNed.
356
357
'3C NMR STUDIES OF LEUKEMIC CELLS
synthetic glucocorticoid dexamethasone, which induces programmed cell death (apoptosis) by binding to a specific glucocorticoid receptor in the cytoplasm ( 8 ) . When the receptors are continuously occupied by active glucocorticoids for at least 24 h, the CEM-C7 cells will be permanently arrested in the G 1 phase of the cell cycle, where eventually they die ( 9 ) . A dexamethasone-resistant clone CEM-C1 has been isolated, which is remarkable in that it contains almost the same number of fully functional receptors as the CEM-C7 clone does. This suggests that dexamethasone resistance of the CEM-C1 cells has developed at a postreceptor level (10). We used I3C NMR to investigate if dexamethasone resistance correlates with possible changes in glucose metabolism. For both cell lines glucose is almost completely converted into enabled us to show that the lactate. The use of D-[ '3Cl]glucose and D-[ 13Cz]gluc~~e CEM-C7 cells consume glucose at 50% higher rate than the CEM-C1 cells. This higher rate corresponds with a larger size of the CEM-C7 cells. Furthermore we could assess the contribution of oxidative phosphorylation and pentose phosphate pathway to total glucose metabolism in both cell lines. MATERIALS AND METHODS
Cell Cultures CEM-C1 and CEM-C7 cells were kept frozen in liquid nitrogen at 5 X lo6 - 1 X lo7 cells/ml in Iscove's medium 25% heat-inactivated fetal calf serum (FCS) 10%DMSO. Cells were defrosted at 37"C, washed 1 X in Iscove's medium 5% FCS, and transferred to growth medium (Iscove's 10% FCS). After 24 h incubation at 37°C and 5% COz the cells were diluted in growth medium to 5-6 X lo5cells/ml for another 24 h. The cells were then checked for viability and morphology with trypan blue and counted in a hemocytometer and model ZM Coulter counter. A cell size distribution curve was obtained with a Coulter 256 Channelyzer. Mean cell diameters were determined using Coulter software on a personal computer interfaced with the Channelyzer. For further propagation the cells were diluted to 2-3 X l o 5 cells/ml in 200 ml Iscove's medium 5% FCS, using T-175 flasks. At this point 100 u/ml penicillin-100 gg/ml streptomycin was added to the cultures (GIBCO, 10,000 units penicillin, 10,000 pg/ml streptomycin). Cells were grown up to midlog phase ( lo6 cells ml-') in 200 ml for each NMR experiment.
+
+
+
+
+
NMR Measurements
To follow glucose metabolism in vivo by NMR the contents of one flask (2-3 X 10' cells) were centrifuged at 200g for 10 min. The pellet was resuspended by gently pipetting back and forth in 3 ml of a cold 1:1 mixture of Iscove's medium and 300 m M Hepes buffer (pH 7.5), following the procedure of Smith et al. (11). When
medium only was used for resuspending, the pH was found to drop considerably during the NMR experiments due to lactate production. The 1: 1 Hepes/Iscove mixture provides sufficient buffering capacity and the pH drops only on the order of 0.3 pH units over the course of about 2 h. Of the concentrated suspension prepared as described above, cell count and size distribution were obtained. Particles with diameter
13C NMR
Studies of Glucose Metabolism in Human Leukemic CEM-C7 and CEM-C1 Cells
JANF. M. POST,* ERNABAUM,AND EDWARDL. EZELL Department of Human Biological Chemistry & Genetics, University of Texas Medical Branch, Galveston, Texas 77550 Received May 2, 1990; revised December 3, 1990 Glucose metabolism of human leukemic cell lines CEM-C7 and CEM-C 1 was investigated in vivo by 13CNMR using "C-labeled glucose. Exact knowledge of glucose concentration, cell count, and cell viability of the cell suspensions made it possible to analyze glucose metabolism in detail. In both cell lines aerobic glycolysis accounts for virtually all glucose consumption. The use of D['3Cz]glucoseprovided a simple method to measure the glucose flux through the pentose phosphate pathway as 9% (CEM-C1 ) and 1 1 % (CEM-C7) of glucose channeled into glycolysis. The dexamethasone-sensitive CEM-C7 cells consume glucose at a rate about 50% higher than the dexamethasone-resistant CEM-CI cells. It is shown that this higher consumption correlates with a larger size of the CEM-C7 cells. Therefore in CEM cells the development of drug resistance does not seem to involve related changes in cell energetics. 0 1992 Academic Press, Inc. INTRODUCTION
It has long been known that malignant tissues have a high rate of aerobic glycolysis as compared to normal tissue ( 1 ). At first it was thought that this was a characteristic property of tumor cells due to an impairment in respiration, but later it was found that also normal cells in tissue culture produce lactate readily in the presence of air. Bissell et al. found that in both normal and virus transformed chick cells the ratio of lactic acid produced to C 0 2 evolved increased with increasing cell density (2). Also the pentose phosphate pathway became more prominent as cell density increased. However, under comparable conditions transformed cells always had a higher rate of (lactic acid produced)/( C 0 2 evolved) than normal cells. These results suggest that both malignancy and cell density play a role in increased aerobic glycolysis of mammalian cells in culture. The use of I3C-labeledcompounds for NMR studies is becoming increasingly popular to study various aspects of carbohydrate metabolism ( 3 ) . The reasons for this are twofold: ( 1 ) cell metabolism can be studied in vivo; (2) concentrations of several metabolites can be followed simultaneously. Important insights into bacterial anaerobic glycolysis ( 4 ) ,gluconeogenesis in rat liver cells ( 5 ) , or drug resistance in tumor cells (6, 7) have resulted from in vivo 13CNMR studies. In this paper we present I3C NMR results obtained with two human leukemic cell lines CEM-C7 and CEM-C 1. CEM-C7 is a wild-type cell line which is sensitive to the * To whom correspondence should be addressed. 0740-3 194/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reseNed.
356
357
'3C NMR STUDIES OF LEUKEMIC CELLS
synthetic glucocorticoid dexamethasone, which induces programmed cell death (apoptosis) by binding to a specific glucocorticoid receptor in the cytoplasm ( 8 ) . When the receptors are continuously occupied by active glucocorticoids for at least 24 h, the CEM-C7 cells will be permanently arrested in the G 1 phase of the cell cycle, where eventually they die ( 9 ) . A dexamethasone-resistant clone CEM-C1 has been isolated, which is remarkable in that it contains almost the same number of fully functional receptors as the CEM-C7 clone does. This suggests that dexamethasone resistance of the CEM-C1 cells has developed at a postreceptor level (10). We used I3C NMR to investigate if dexamethasone resistance correlates with possible changes in glucose metabolism. For both cell lines glucose is almost completely converted into enabled us to show that the lactate. The use of D-[ '3Cl]glucose and D-[ 13Cz]gluc~~e CEM-C7 cells consume glucose at 50% higher rate than the CEM-C1 cells. This higher rate corresponds with a larger size of the CEM-C7 cells. Furthermore we could assess the contribution of oxidative phosphorylation and pentose phosphate pathway to total glucose metabolism in both cell lines. MATERIALS AND METHODS
Cell Cultures CEM-C1 and CEM-C7 cells were kept frozen in liquid nitrogen at 5 X lo6 - 1 X lo7 cells/ml in Iscove's medium 25% heat-inactivated fetal calf serum (FCS) 10%DMSO. Cells were defrosted at 37"C, washed 1 X in Iscove's medium 5% FCS, and transferred to growth medium (Iscove's 10% FCS). After 24 h incubation at 37°C and 5% COz the cells were diluted in growth medium to 5-6 X lo5cells/ml for another 24 h. The cells were then checked for viability and morphology with trypan blue and counted in a hemocytometer and model ZM Coulter counter. A cell size distribution curve was obtained with a Coulter 256 Channelyzer. Mean cell diameters were determined using Coulter software on a personal computer interfaced with the Channelyzer. For further propagation the cells were diluted to 2-3 X l o 5 cells/ml in 200 ml Iscove's medium 5% FCS, using T-175 flasks. At this point 100 u/ml penicillin-100 gg/ml streptomycin was added to the cultures (GIBCO, 10,000 units penicillin, 10,000 pg/ml streptomycin). Cells were grown up to midlog phase ( lo6 cells ml-') in 200 ml for each NMR experiment.
+
+
+
+
+
NMR Measurements
To follow glucose metabolism in vivo by NMR the contents of one flask (2-3 X 10' cells) were centrifuged at 200g for 10 min. The pellet was resuspended by gently pipetting back and forth in 3 ml of a cold 1:1 mixture of Iscove's medium and 300 m M Hepes buffer (pH 7.5), following the procedure of Smith et al. (11). When
medium only was used for resuspending, the pH was found to drop considerably during the NMR experiments due to lactate production. The 1: 1 Hepes/Iscove mixture provides sufficient buffering capacity and the pH drops only on the order of 0.3 pH units over the course of about 2 h. Of the concentrated suspension prepared as described above, cell count and size distribution were obtained. Particles with diameter
