Inhibition of Anaerobic Glycolysis in Ehrlich Ascites Tumor Cells by 2-Deoxy-ducose* = MARSHALL W&RENBERG~ AND JAMES F~OGG (Department of Btblqicd Chntktrg, Medical School, Uniaersily of Michigan, Ann Arbor, Mich.) Little is known of the mechanisms by which %deoxy-n-glucose (a-DG) inhibits anaerobic gly- colysis in yeast cells (1, 9, 18, 14) and in brain, diaphragm, and liver slices (15). This compound also inhibits anaerobic glycolysis in various tumors (15), prolongs the survival time of mice bearing the Krebs ascites carcinoma (a), and causes a decreased rate of growth of several tumors in rats or mice (2, 8). The inhibition with yeast (1) is competitive if either glucose or fructose serves as the substrate. 2-DG is phosphorylated by the hexokinase of yeast (l), brain (lo), and tumor (15), but it is not catabolized further by these tissues. 2-DG has been claimed to inhibit the transport of monosaccharides through cell membranes (1,9). During the course of this inves- tigation Wick et al. (12) reported that 2-deoxy-n- glucose-6-phosphate (2-DGP) competitively inhib- its phosphohexoisomerase in mammalian tissues. The observation of a step preceding glycolysis and possibly involved in hexose transfer into as- cites tumor cells (5,6) prompted the study of 2-DG to characterize further the transport step in hexose utilization. The &dings of Wick et al. (12) were confirmed for the ascites tumor, and, in addition, 2-DGP was found to inhibit glycolysis at a point subsequent to the formation of hexose diphos- phate. The possibility remains, however, that 2- DG inhibits hexose transport, since the enzyme inhibitions do not fully explain the inhibitory effects on glycolysis in cells. MATERIALS AND METHODS The Ehrlich ascites tumor (Hauschka, clone 9 [S]) was grown in white Swiss mice. At the 8th day after inoculation with 0.2 ml. of ascites tumor, the tumor cells were collected *This investigation was supported in part by American Cancer Society Institutional Grants Nos. 2%C and 49-D. The material was taken in part from a dissertation submitted to the Horace H. Backham School of Graduate Studies by Marshall Nirenberg in partial fultilhnent of the requirements for the degree of Doctor of Philosophy in the University of Michigan, 1957. t Present address, National Institute of Arthritis and Met- abolic Diseases, National Institutes of Health, Bethesda, Md. Beceived for publication October 29, 1967. in a heparinixed syringe, centrifuged at 400 X g for 5 minutes, and then washed twice with Krebs-Binger bicarbonate b&r at pH 7.4. Small amounts of contaminating erythrocytes which collected at the bottom of the centrifuge tube were removed each time with a capillary pipette. After suspension to about 93 per cent (v/v) concentration in alkaline, isotonic KC1 (4). tbe cells were homogenized either by application and rapid release of a pressure of 1600 lb/sq in of Na in a small stainless steel cylinder at room temperature or by 100 passes in a Brendler homogenizer at 6' C. Anaerobic glycolysis was measured in a Warburg apparatus at 37' C. with an atmosphere of 95 per cent Nx - 5 per cent Cot (11). Experiments with intact cells were performed in Krebs-Ringer bicarbonate buffer; those with homogenates in a medium mod&d from that of LePage (4) (Table 9). The manometric measurements were confirmed by simul- taneously determining the disappearance of hexose from the medium (5, S).l Tumor extracts containing both hexokinase and phosphohexokomerase were obtained by centrifuging a homogenate at 90,900 X g for SO minutes at 5" C. A lipide layer rising to the top was discarded. The active enzymes were contained in the aqueous supemate. The activities of hexokinase and phosphohexoisomersse were measured by di- rect spectrophotometric assay (7) (Chart 1). The formation of reduced triphosphopyridme nucleotide (TPNH) was fel- lowed by measuring the absorption at 840 rnp in a Model DU Beckman spectrophotometer with a photomultiplier at- tachment. The sample of glucose-&phosphate dehydrogensse (Sigma) used in the assay was free of hexohinase and pbos- phohexoisomerase activity. Doubling the amount of all rest- tants except the tumor enzyme did not increase the rate of the reaction, whereas doubling the amount of the tumor enzyme increased the reaction rate. The tumor extracts aP- parently contained phosphogluconate debydrogenase, since the ratio, TPNH productiou/hexose consumption, was sp proximately 2. Glucose and fructose were Pfanstiehl reagents. a-DG wss kindly supplied by Dr. Harold Blumenthal. LDGP was Pre- pared by incubating 111 pmoles %-DG with 9 mg. *Yeast hexokinase (Sigma), 150 moles diiodium adenosine trIPhose phate (ATP) (Sehwarz), and 309 pmoles MgCb in 9.9 ml. of a 0.9 Y phosphate buffer at pH 7.4. After SO minutes of incubation at SO" C., the solution was placed in a boiling water bath for 9 minutes, and the precipitated protein n-as then removed by centrifugation. The solution was stored at - lb" C. FRlctose-&phosphate (F-6-P) (Nutritional Rio- chemical Corporation) was neutralized to pH 7.0 with &C's before use. RESULTS When equimolar concentrations of %-DG and glucose were supplied to intact Ehrlich cells, s 1 M. W. Nirenberg and J. F. Hogg, unpublished results' 518 NIRENBERG AND HOQG-Tumor Inhibition by d-Deoxg-r>-Glucose 519 reduced rate of glucolysis was observed. Under &ilar conditions, %DG completely inhibited fruc- tolysis (Table 1). In homogenates (Table 8), when a-DG alone was used as a substrate, CO2 was displaced from the bicarbonate buffer, presumably ss a result of the conversion of %DG to %DGP. The measurement of acid production in homog- enates, then, is not a true measure of inhibition of the entire pathway of glycolysis by this com- pound. When the amount of CO2 produced by the conversion of %DG to R-DGP was subtracted from the CO% obtained with substrates plus %DG, 0.8 ,- 0.7 - 20.6 - : m ;0.5- c z g 0.4 - 1 ;0.3- 0" 40.2 - 0.1 - 0 W!NUTES CEURT I.-The effect of Pdeoxy-n-glucose-f&phosphate upon Ehriich ascites tumor phosphohexoisomerase. 0, a.7 ~olea F-6-P; 0, a.7 /m~oles F-6-P + 22.0 qoles a-DGP; 0, 24.0 eoles 9-DGP. The medium contained, in 3.0 ml.: appropriate substrates, 100 /.unoles tris(hydroxymethyl)- aminomethane bulfer (pH 8.0), 6 ~olea ATP, 10 pmole~ MgCh, 1.2 mnoles triphosphopyridme nucleotide (TPN) (Schwars), 6 holes KF, 0.2 K units of glucose-t?-phosphate dehydrogenase, and sufhcient tumor extract (0.05-0.10 ml.) to give the desired reaction rate. a large inhibition of glycolysis became apparent. The maximum rate of acid production, however, was still observed with %DG present. Therefore, since glucose might prevent phosphorylation of %DG, it was possible that no inhibition had occurred in the homogenates. To eliminate this qualification of conclusion, more specific measurements were made. First, the effect of %DG upon tumor hexokinase was tested spectrophotometricaJly. An equimolar con- centration of %DG had little effect upon glucose phosphorylation but had a large inhibitory effect with fructose, the change of optical density per minute being 0.11-0.13 for all combinations except fructose + %DG, when the rate was 0.074. These results could be explained by the possibility (a) that %DG competed more successfully with fruc- TABLE 1 THE EFFECT OF %DEOXY-D-GLUCOSE UPON THE RATE OF ANAEROBIC GLYCOLYSIS OF EERLICH AXITES TUMOB Cm Each vessel contained a.9 ml. of a S per cent cell suspension which was incubated for 90 minutes. No endogenous glycolysis was observed. COz/hr/mg dry wt Per cent Additicm per vetmel cells inbibi- (rmoles) (PI.) tioo 11 Glucose 26.8 11 Fructose 21.1 11 Glucose+11 %DG 19.4 28 11 Fructme+ll %DG 0 100 TABLE 2 THE EFFECT OF P-DEOXY-D-GLUCOSE UPON THE RATE OF ANAEROBIC GLYCOLYSIS OF AN EFLRLICH AWITIB TUMOR HOMOGENATE Each vessel contained S.S ml. of a 3 per cent tumor homo e- nate. CO, production was measured at 20-minute intervals f or 80 minutes. The medium au&ted ofz 0.00%4 M K%HPCh, O&&5 M KHCOa, 0.04 M nicotinamide (Merck), 0.009 M diso- dium adenosine triphosphate (ATP), 0.0004 M diphosphopyri- dine nucleotide (DPN) (Pabst), 0.0095 M potassium fruc- tose-l,&diphosphate (HDP) (Schwarx), (Schwarz). and 0.01 M MgCla. 0.0005 M pyruvate COJhr/mg Apparent CwwwT dwwt inhibi- Additiona per vessel drJld corrected tion (fimoles) (rl.) for !&DG (per cent) None 8.4 16.7 a-DG 2S.8 16.7 Glucose 3T2.a 16.7 Fructose 31.7 16e7DGu0Jse+16.7 34.a 10.4 68 16i~D~tme+16.7 93.6 5.7 82 tose for hexokinase or (b) that %DGP inhibited the phosphohexoisomerase reaction required in the assay with fructose. The latter possibility was checked by using F-6-P as substrate in the same system, thereby permitting an assay for phosphohexoisomerase. A high level of %DGP completely inhibited the tumor phosphohexoisom- erase (Chart l), the inhibition being reversed by higher levels of F-6-P (Chart a). Thus the inhibition of homogenate glycolysis by %DG (Ta- ble 2) was veraed. The inhibition of tumor phosphohexoisomerase 220 Cancer Research Vol. 18, June, 1958 by 2-DGP could explain the inhibitory action of 2-DG upon glucose utilization by the cells, but it could not explain the even greater inhibition of fructose utilization. Fructolysis via the Emb- den-Meyerhof pathway, in contrast to the coupled system for the spectrophotometric assay, bypasses the phosphohexoisomerase step. This suggested that 2-DGP was also inhibiting one or more re- 0.7 - =t E0.6 - 0 1 2 W&ES 4 5 6 CHART P.-Reversal of adeoxy-o-glucose&-phosphate in- hibition of Ehrlich aacites tumor phosphohexoisomerase by fructose-&phosphate. 0, 2.5 holes F-6-P; A, 2.5 pmoles F-6-P + 13.9 rmoles 5~DGP; IJ, 20 junoles F-6-P + 13.9 pmoles %DGP. See Chart 1 for medium. actions of the Embden-Meyerhof pathway follow- ing the formation of F-6-P. This possibility was tested by adding various combinations of 2-DGP, F-6-P, and fructose-1,6-diphosphate (HDP) to tumor homogenates and determinii the rate of anaerobic glycolysis manometrically (Table 3). Since the phosphorylated inhibitor was used, the previous problem of acid production from 2-DG was eliminated here. A small amount of HDP was required in the medium routinely used for homogenate experiments in order to obtain an ade- quate rate of glycolysis (4); therefore a low rate of glycolysis occurred in the vessels containing no added substrate. The 2-DGP inhibited the gly- colysis of F-6-P, thus demonstrating inhibition of a step in the glycolytic pathway after the formation of F-6-P. Although one might expect 2-DGP to inhibit phosphofructokinase, 2-DGP again inhibited glycolysis greatly when additional HDP `was supplied as substrate. Since equal in- hibitions by 2-DGP were obtained with F-6-P or HDP, clearly 2-DGP can inhibit the Embden- Meyerhof pathway at some point after phospho- fructokinase action. DISCUSSION 2-DGP inhibition of HDP utilization and of phosphohexoisomerase clearly is a mode of in- hibition of cellular glycolysis by 2-DG. With in- tact tumor cells, however, fructolysis was much more sensitive to 2-DG inhibition than glucolysis. Furthermore, fructolysis does not pass through TABLE 3 TEE EFFECT OF %DEOXY-D-GLUCOsE-6PHOSPHATE UPON THE RATEOFANAEROBIC GLYCOLYSISOFAN EARLICHAXITESTUMORHOM~GENATE See Table !Z for conditions. Additiona per vessel hd~d COt/hr/mg dry wt ("1.) Per cent inhibition None (1.5`HDP ~II medium) l-l.4 2%. 0 %DGP 5.8 7.5 F-6-P 92.6 7.5 F-B-P+%.O %DGP 7.5 7.5 HDP 17.0 7.5 HDP+ea.O %DGP 5.e 49 66 69 the phosphohexoisomerase step. Therefore, the action of 2-DGP cannot fully explain 2-DG in- hibition of cellular glucolysis. Possibly 2-DG in- hibits fructolysis by serving as a competitive sub- strate for hexokinase. The Michaelis-Menten con- stants of brain hexokinase for D-ghlCOSe, 2-DG, and D-flUCtOSe are 8.0 X 10" M, 2.7 X 10s5 hl, and 1.6 X lo+' BA, respectively (10). Therefore, at an equimolar concentration, 2-DG should com- pete effectively for hexokinase with fructose, but not with glucose. Alternatively, 2-DG may com- pete with glucose and fructose for the hexose transfer mechanism, fructolysis being the more inhibited because fNCtOSe has the lesser affinit? for the binding site (5, 6). Regardless of what may be the additional mode of inhibition b>* 2-DG, its use as a tool for the study of hexose transport is seriously q.ualified by its conversioll to an effective inhibitor of glycolytic enzymes. SUMMARY EquimoIar.concentrations of 2-deoxy-n-glu~ose inhibited anaerobic glycolysis of Ehrlich ascItes NIRENBERG AND Hoac-Tumor Inhibition bg d-Deoxy-D-&.xose tumor cells when either n-glucose or n-fructose 6. -. Hexose Uptake in Ascites Tumor Cells. Fed. was used as the substrate, glycolysis with the Pm., 16: 2277.1057. latter being more strongly inhibited. %Deoxy- 7. SLEIN, M. W.; Corn, G. T.; and Corn, C. F. A Compsra- s-glucose did not inhibit anaerobic glycolysis pri- tive Study of Hexokinase from Yeast and Animal Tissues. J. Biol. Chem.. 166:783-80. 1050. iarily at the monosaccharide tram&&t level-In- 8. stead it was converted to Sdeoxy-n-glucose-6- phosphate by tumor hexokinase and the latter compound inhibited both phosphohexoisomerase 9. and a further step in the Embden-Meyerhof path- way subsequent to phosphofructokinase action. The Michaelis-Menten constants of hexokinase lo* suggest that !%deoxy-n-glucose may also inhibit 11 hexokinase action on n-fructose. REFERENCES 14. 1. CRAMER, F. B.. and WOODWARD, G. E. 3-Desoxy-n-glucose as an Antagonist of Glucose in Yeast Fermentation. J. Franklin Inst., 263:954-60, 1959. 13. 9. ELY, J. 0. %Deoxy-n-glucose as an Inhibitor of Cancerous Growth in Animals. J. Franklin Inst., 268:X7-60, 1954. 3. HAUSCHKA, T. S. Cell Population Studies on Mouse As- l4 * cites Tumors. Trans. N. Y. Acad. SC.. 16:64-75,195S. 4. LEPAGE, G. A. Glycolysis in Tumor Homogenates. J. Biol. Chem., 176:1009+?0, 1948. 16. 5. NIREXBERG, M. W.. and HOGG, J. F. On the Mode of Hexose Uptake by Ascites Tumor Cells. J. Am. Chem. SOL. 78:6210, 1956. ~OKOLOFF, B.;`Cw, @UELROR, C.; and BZACH, J. The Effect of a Glucose Analog on the DNA Content in Tumor Tissue. Proc. Am. Assoc. Cancer Research, 2: 148.1956. Sona, A. Selective Fermentation and Phosphorylation of Sugars by Sauternes Yeast. B&him. et Biophys. Acta, 20: 6W38.1956. Sona, A., and Cq R. K. Substrate SpecifZty of Brain Hexokinase. J. Biol. Chem., 210:581-95,1954. Uaa~msrr, W. W.; Bunma, R. H.; and STAUFFER. J. F. Manometric Techniques and Tissue Metabolism. Min- neawlis. Minn.: Burcess Publishinx Co.. 1949. WICK. A. N.; DRUISY: D. R.; Na&a, H. I.: and WOLFE, J. R. Localization of the Primary Metabolic Block produced by %Deoxyglucose. J. Biol. Chem., 224:963-60, 1957. WOODWAED, G. E. D-Desoxy-n-ghmose as an Inhibitor in the Aerobic GIucose Metabolism of Yeast. J. Franklin Inst., 264:653-65.1052. WOODWAXD, G. E.; Cn, F. B.; and HUDSON. M. T. Carbohydrate Analogs as Antagonists of Glucose in Car- bohydrate Metabolism of Yeast. J. Franklin Inst., 266: 677-87.1953. WOODWARD, G. E., and HUDSON, M. T. The Effect of 2-Desoxy-n-glucose on Glycolysis and Respiration of Tu- mor and Normal Tissues. Cancer Research, 14:599-605, 1954.