HEXOSE TRANSPORT IN ASCITESTUMOR CELLS
Aug. 20, 1958 [COSTRIBUTION FROM
THE
DEPARTNEST OF
~ I O L O C I C A LCHEMISTRY,
MEDICAL S C H O O L , UNIVERSITY
4407 O F &lICHIGAN]
Hexose Transport in Ascites Tumor Cells’ BY MARSHALL W. NIRENBERGAND JAMES F. H O G G ~ RECEIVED APRIL3, 1958 The competitive inhibition of hexose utilization in ascites tumor cells by the non-utilizable hexose, galactose, has been ascribed to competition for a binding site which appears to be a functional part of the transport mechanism. Galactose inhibited neither hexose utilization by homogenates nor hexokinase action in extracts of the ascites tumors. In Gardner cells where hexose transport was shown to limit the rate of glycolysis, kinetic analysis served to identify the galactose-sensitive reaction with a step by which the sugar gains access t o the glycolytic enzymes. The relative affinities for the binding site are in the order: glucose > galactose = fructose. Several steroid compounds also were found to inhibit hexose utilization in a manner resembling in several respects the galactose inhibition.
Although chemically mediated transport of monosaccharides is shown more easily for complex systems such as the iritestinal and renal tubular walls, sugars apparently enter many cells by processes which are a t least in some respects similar. The studies of LeFevre, Wilbrandt and other^^-^ have shown that sugar entry into human erythrocytes apparently involves a reversible combination of the sugar with a membrane constituent. This conclusion has been extended to foetal erythrocytes of some other mammals* and to yeastg.10and bacteria.” Evidence for a comparable mode of hexose entry into Ehrlich ascites tumor cells was presented in a preliminary communication. More recently, Crane, et a1.,12 have verified and extended these results by means of a tracer technique. Although sugar transfer often has been supposed to depend upon phosphorylation, the insulinstimulated step in sugar uptake appears to precede such reaction.13-17 Other reasons for doubting that the transfer reaction is a phosphorylation are a poor correlation for the intestine between the transfer rates and phosphorylation rates of sugars18 and the concentrative transfer of 1-deoxy- and 6-deoxyD-glUCOSe from an intestinal pouch.1g (1) Portions of this work have been the subject of a preliminary cummunications and have been presented before the meeting of the American Society of Biological Chemists, Chicago, April, 1957.4 This material was taken in p a r t from a doctoral dissertation submitted t o t h e Horace H . Rackham School of Graduate Studies by Marshall W. Sirenberg in 1957. T h e investigation was supported in part by American Cancer Society Institutional Grants Nos. 22C and 22D and by the Michigan Memorial-Phoenix Proejct No. 45. ( 2 ) T o whom inquiries should be directed. ( 3 ) M. W.Nirenbergand J. F. Hogg, THISJOUXNAL, 78, 8210 (1958). (4) h.1. W. Nirenberg and J. F. Hogg, Federalion Proc., 16, 227 (1957). ( 5 ) P. G. LeFevre, Symp. Soc. Exjit. Bioi., 8 , 118 (1954). ( 6 ) W Wilbrandt. ibid., 136 (1964). (7) D. Reinwein, C. F. Kalman and C. R. Park, Fpdevatioii Proc., 16, 237 (19.57). ( 8 ) F. Bowyer and It’. F. Widdas, Disc. Foraday SOC., 21, 251 (195G). (9) A . Rothstein, S y m p . SOL.Erpt. Biol., 8 , 186 (1954). (10) E. R. Blakely and P. D. Boyer, Biochem. Biophys. Acta, 16, 576 (1955). (11) J. Monod, in “Enzymes: Units of Biological Structure and Function,” Academic Prcss, Inc., h’ew York, N. Y., 1956, p. 7. (12) R . K. Crane, R. A. Field and C. F. Cori, J. B i d . Chem., 224, 819 ( 1 9 5 7 ) . (1:O R. Levine .irid A r . S . Goldstrin, K r c . P r o p . Ilormoirp Rcs., 11, 3.13 (1955). (14) A. E. Kenold, A . B. Hastings, F. B. Nesbett and J. Ashmore, J. Biol. Chem., 213, 135 (1955). (1.5) C. R. Park, J . Clin. Invesf., 32, 593 (1953). (16) C. R . Park and L. 13. Johnson, A m . J . Physiology, 182, 17
(1955). (17) C. R. Park, J. Bornstein a n d R. L. Post, ibid., 182, 12 (1955). (18) A. Sols, REV.espaA. jisioi., 11, 277 (1955). (19) R. E;. Craue and S. M. Krane, Biochem. Bioplzys. A d a , 20, 568 (1966).
Our investigation3 on the mode of hexose uptake in the Ehrlich ascites tumor has been augmented and comparable results have been obtained with the Gardner tumor which, however, showed a specific dependence of glycolytic rate on hexose concentration for a wide range of values. Furthermore, certain steroids have been found to exert a selective inhibitory action upon fructolysis. Experimental Materials and Procedures.-Both the Ehrlich carcinoma and Gardner lymphosarcoma ascites tumorsM were obtained from Dr. T. Hauschka.21 The Ehrlich tumor (clone 2) was carried in unselected white Swiss mice; the Gardner (6C3IIED) in C3H mice obtained from the Jackson Memorial Laboratories, Bar Harbor, Maine. At 7 to 9 days after the intraperitoneal injection of 0.2 ml. of ascites tumor, the ascites tumors were aspirated into a heparinized syringe, centrifuged 5 min. a t 400 X g, the fluid decanted and the contaminating erythrocytes removed by aspiration with a capillary pipet. The cells w-ere then washed twice with Krebs-Ringer bicarbonate solution,2* erythrocytes being removed after each centrifugation. Quantitative homogenization was obtained by subjecting a 3370 suspension of cells in isotonic KClZ3to the application and,rapid ielease of a pressure of greater than 1500 1b.l sq. in. of XZ in a small stainless steel tank a t room temperature.24 Initially, for comparison, homogenates also were prepared in a Brendler homogenizers at 5 ” . Soluble tumor hexokinase preparations nere obtained by centrjfuging the homogenates a t 20,000 X g for 30 min. a t 5 . Hexokinase was found in the aqueous supernatant fluid; the precipitate was almost devoid of hexokinase activity. Anaerobic glycolysis was measured in a Warburg apparatus shaken a t 37” with an atmosphere of 95% h’?-5% CO?.22 Tumor cells were suspended in Krebs-Ringer bicarbonate buffer2$ a t pH 7.45; tumor homogenates in a medium, modified from that of L e P a g P (Table I ) , which held the pH a t 7.4 under an atmosphere of 95‘5; N2-5Y0 COZ. Intact cells maintained linear rates for a t least 90 min.; homogenates for 40 min. Glucose26 and fructose were Pfanstiehl reagents. Glucosefree galactose was kindly furnished by Dr. A . Campbell. The galactose had been freed of glucose by allowing two successive lots of washed cells of baker’s yeast to ferment any glucose present. Additional purification was accomplished by twice-repeated treatment of a 10% solution with 30 g./l. of Norite. Using the hexokinase assay described below with yeast hexokinase, the purified galactose contained less than 3 pg. of glucose in 80 mg. The steroids tested were commercial preparations (Ciba or Schering) kindly supplied ( 2 0 ) G. Kiein, Expl. Ccli Res., 2, 518 (1951). ( 2 1 ) T. S. Hauschka, Trans. N . Y . Acad. Sci., 16, 64 (1953) ( 2 2 ) W. W. Umbreit, R. H. Burris a n d J. F. Stauffer, “Manometric Techniques and Tissue h‘letabolism,” Burgess Publishing Co.,hlinneapolis, Minn., 1949. ( 2 3 ) G. A. LePage, J . Bid. Chem., 176, 1009 (1948). (24) C. S. French and H. Milner, in S. P. Colorrick and N. 0. Kaplan, “>lethods in Enzymology,” New York, N. Y.,1, 64 (1955). ( 2 5 ) H. Brendler, Science, 114, 6 1 (1951). (26) All carbohydrates and their derivatives had tile D-configuration unless otherwise stated.
440R
Vol. 80
hl. W. NIRENBERG AND J. F. HOGG
liy Dr. Burton L. Baker; nystatin (Squibb M y c o s t a h )
Iiy Dr. J . 0. Lampen. Analytical Methods,-sugars xvere rrleasured in samples deproteinized with Ba(OH)? and ZnS01.Z7 Reducing sugar was determined by the colorimetric method of Somogyi2e : i t i d fructose by the method of X correction proportional t o the galactose level was applied for the latter method. IIexokinase Lictivitv was measured h v :I direct snectrophotometric ass:iydu,31 x i t h a reaction mixture containing in a total volume of 3.0 ml. the following: hexose substrate, 100 pmoles of 1ris-(hydroxymetli)-l)-aniiiiomethane bufier (pH 8.01, 5 prnolcs of disodium adenosine triphosphate !.lTP) (Schwarz), 10 pmoles of hlgCl,, 1.2 pmoles of triphusphopyridine nucleotide (TPS) iSigrna'i, 3 pmoles of KF, 0.2 k' units of glucose 6-pliospliate dehydrogenase iSigiii:~),2.3 ing. of crude plicisp}io!iesoisonnerase (Sigma, r~iiiittctl f u r tumor extracts), and suflicient tumor hemprepiratiun \0.05-0.10 nil.) to give the required 11 rate. The formation o f reduced T P N ( T P S H ) was measured by the absorption at 340 mp iu a model DU Beckinan spectrophotometer using a photomultiplier attacliment. Both the glucose 6-phosphate dehydrogenase :tiid the phosphohexoisomerase \yere found t o be free of Iiesokinase. Of the reactants, only hexokinase was at a rate-limiting level. n'hen the concentration of fructose \vas made limiting (0.1-1.0 Krnoles per 3 ml.), the rate of 'TPNII formation was proportional to the concentration of fructose. The tumor extracts apparently contained phos~~liogluconate tlcliydrogenase sirice the ratio, TPSH producri~iii/fructoie consumption, was approximately 2.
CPOS
27) hl. Sumogyi, J . Bmi. Ci 128) 31. Somogyi, i b i d . , 1 9 5 , 19 (1952) ( 2 9 ) T. €1. Roe, i b i d . , 107, 15 (1934). l R O i 11. \V. Slein, G. T. Cori and C. 1'. C o n , zbad., 186, 703 (1950). (31) R . E;. Crane and A. Sols, in S. P. Colowick and N. 0. Kaplan, "hlethods , t i Enzymology." S e w York. S . Y., 1 , 277 (19.55). ( 3 2 ) In two experiments, the rate of fructolysis increased approximately 2 5 % between 10 and 40 pmoles Apparently a concentration of 10 pmoles per vessel just saturates the Ehrlich cells and a small alteration of cellrilar condition may re4ult i n r? very limited concmtratioii dependency. t.33) H I.ineae+\er and D. Burh, T H I S J U L R N A I . , 86, 688 (1!4.%41.
dry C Owt./hr. dmg.
Additions per vessel. pmoles
.~
Results and Conclusions Ehrlich Carcinoma Ascites Tumor.----In intact cells glucolysis and fructolysis proceeded at approximately equal rates (Table 1). Raising the amount of glucose or iructose i n the Warburg vessel from 11 to 2 i S pmoles did not increase significantly the rate of glycolysis. Lesser amounts precluded accurate manometric measurements of rate. Galactose was not metabolized, since i t did not serve as a glycolytic substrate and, when incubated with the tumor cells, tiid not disappear progressively from the n i e ~ l i u m . ~ -11though galactose had no effect upon glucolysis (Table I), it markedly inhibited fructolysis in intact cclls. The inhibition of fructolysis mas completely reversed by additional fructose. Plotting results for fructolysis according to Linemeaver and R ~ r (Fig. k ~ 1) ~ showed that the galactose inhibition was competitive, thus indicating a common binding site for galactose arid fructose. Such a common binding site was indetectible in homogenates; here galactose failed to inhibit fructolysis. Glucolysis and fructolysis proceeded ;it about equal rates, these being equally high in honiogeiiates and intact cells (Table I). The manometric results mere confirmed by simultaneously determining the disappearance of fructose from the medium (Table 11). Both methods of analysis demonstrated clearly that galactose inhibited fructolysis in intact cells but not in homogenates. The hexokinase step was examined more critically by another technique, namely, a spectro -
TABLE I SUBSTRATE, COSCEXTRATIOX A N D GA~ACTO~E . ~ X A E R O B I C GLYCOLYSIS I N EHRLICIr ASCITESTUMOR
EPFEcr OF
MI.
Intact cell-. 13.9 glucose 13.9 fructose 13.9 galactose 414 0 glucose 8 . 7 glucose 454,O galactose 16.7 fructose 1 6 , 7 fructose 222.0 galactose 1 6 . 7 fructose 444 0 galactose 2 7 , 8 fructose -I- 220.0 galactose 5 5 . 6 fructose 220.0 galactose
+ + + +
24.5
'76.7 0 29.0 29.1 25.3 8 7 4.S 15.3 26.7
Homogenates" 16.7 glucose 24.7 1G,7 fructose 2 4 . ti 1 6 , 7 fructose 443 galactose 25.o 144.0galactose 1.5 The medium consisted of: 0.0024 -11 I;2HP04, 0.025 KHCOB,0.05 J I nicotinamide (Merck), 0.002 JI disodium adenosine triphosphate (Schwarz) , 0.004 II1 diphosphopyripotassium fructose 1,6dine nucleotide (Pabst), 0.0005 diphosphate (Schwarz), 0.0005 Msodiurn pyruvate (Schwarz) and 0.01 31 hfgCll. Rates of glycolysis in the homogenates were corrected for a blank glycolysis of approximately 10 pl. C02/mg. dry wt. homogenate/hr. due to the fructose 1,B-diphosphate in the medium. Therefore the figures given represent the complete Embden-Meyerhof pathway of glycolysis starting from glucose or fructose.
+
(i
photometric assay of hexokinase activity. Both glucose and fructose, a t the same concentrations used for measurements of glycolysis were found to be phosphorylated at equal rates and levels of galactose high enough to inhibit fructolysis by 807, or more in a cellular system did not affect the hexokinase action even when the fructose concentration was rate-limiting (Fig. 2). Therefore galactose did not inhibit fructolysis by competing for hexokinase. TABLEI1 COrVlPARISOS BET\T'EES h I A K O M E T K I C DETERMIXATIONS O F ASAERODICFRUCTOLYSI? AND MEASL,REMENTS OF FRUCTOSE
nISAPPEARASCE .%SCIlE$
Substrate, pmoles
Intact cells 10.7 fructose 1 6 . 7 fructose galactos? Homogenates 1 1 . 1 fructose I I 1 fructose galactose
FKOM
1'HE
>,IEDIU>f
IS
EHRLICA
TUMOR CELLS A N I ) HOBIOGEKATES
+ 1%)
+ 3%i1
Theoretica I fructose
I:ructos? dis:ippearancP. Umole:.
Extra CU: produced, ul
appearance." pmoles
7.2
29ti
ti t i
#i ,A , I
135
::
.j.t i
31)-i
1 Ii
5 6
190
4 .8
diS-
li
The theoretical fructose disappearance was calculated from the manometric data by assuming that one pmole of fructose yields 2 pmoles of lactic acid and therefore should produce 44.8 pl. of C 0 2 . a
-In alternative to concluding that galactose and fructose compete for a common site is the possibility that irituc-1 ~ I ~ H L Ocells Y but not howloge7latc.~
Aug. 20, 1958 4
HEXOSE TRANSPORT IN ASCITESTUMOR CELLS
4409
0.9
GALACTOSE, p M/WL. T
0.0 0.1
i a ! m
0.6
c 4
* 0.5 I-
v)
z
0.4 -1 4
V
0.3 0
0
0.2
Fig. 1 --il Lineweaver-Burk plot of galactose inhibition of fructolysis in Ehrlich ascites tumor cells. The substrate concentration is expressed as @molesper ml. and the velocity is expressed as pmoles COn produced per mg. dry weight of tumor cells per hr.
I
0
I
I
I
I
I
I
I
I
1 2 3 4 5 6 7 8 9 W I NUTES.
Fig. 2.-The effect of galactose upon phosphorylation of fructose by Ehrlich ascites tumor hexokinase: 0 , no substrate; 0 , 355 @moles galactose; V, 1 @molefructose; X, 1 pmole fructose 125 @molesgalactose; A, 11 pmoles 355 pmoles fructose or glucose; 0,11 pmoles fructose galactose.
produce a glycolytic inhibitor in the presence of + galactose. This was tested by incubating cells + for 45 min. with a mixture of galactose and fructose (ratio of 300/1). The cells should then have contained the supposed inhibitor. After centrifugation the cells were homogenized and lyo- cently by means of a different technique.13 \Vith philized. X suspension of tumor cells incubated Gardner cell homogenates, the results agreed in all without galactose was treated in the same way. respects with the results for Ehrlich cell homogThe rate of fructolysis by a fresh homogenate was enates, including Fig. 2 . Therefore the Gardner then determined after adding either the galactose- cells seem also to contain the reaction preceding treated or the control homogenate. Since the ex- glycolysis. periments yielded 31 and 30 PI. COn/mg./30 min., GALICTOSE.pW/ML. respectively, the galactose-treated cells did not contain a glycolytic inhibitor. Therefore we must return to the concept of a common binding site for / 20 the two sugars. Xlthough one would anticipate this site to be on an enzyme, galactose inhibited 16 ’. i l 17 neither fructolysis in homogenates nor phosphorylation in extracts of the cells. These results then r ‘-12. suggest that in cell suspensions an unknown initial reaction must occur before the sugars can enter glycolysis. Plausibly this reaction may participate 19 0 in the transport of fructose into the cell. Gardner Lymphosarcoma Ascites Tumor.--Here again galactose was not glycolyzed whereas both fructose and glucose served as glycolytic substrates 0 .6 1.2 l.E 2.4 3.0 3.6 4.2 4.8 5.4 1/s x 10 2 (Table 111). I n these cells, as in the Ehrlich cells, galactose inhibited fructolysis and again galactose Fig. 3.- Lineweaver-Burk plot of galactose inhibition of was inhibiting competitively (Fig. 3). *4t very fructolysis in Gardner lymphosarcoma ascites tumor cells. high galactose/glucose ratios (200/1), however, an The substrate concentration is expressed as pmoles per ml. inhibition of glucolysis also was observed with and the velocity as pmoles CO? produced per mg rir\ these cells, the inhibition being fully reversed by weight of tumor cells per hr. additional glucose. Although glucolysis inhibition was not obtained in Ehrlich cells a t the highest A distinct and significant feature in the Gardner ratios experimentally feasible, competition between cell, however, is the dependence of the rate of COSglucose and galactose for entry has been shown re- production upon the level of fructose or glucose
241 ~~
4410
hf. W. NIRENBERG AND J. F. HOGG
taken, although glycolysis was always more rapid with the latter sugar (Table 111). This concentration dependency, as well as the difference in rate with the two sugars, disappeared upon homogenization, both rates being substantially increased (Table 111). The lack of concentration dependency in tumor homogenates has been observed previously by LePage. 3 4 ,Ilthough the accelerntion of glucolysis upon homogenization possibly could be explained by an increased availability of coenzymes or phosphate, d5 this can hardly be true for the much greater acceleration (30-fold) of fructolysis. This acceleration upon homogenization means that a barrier has been removed. The barrier, however, is the necessity that the sugar pass through a reaction, one in which galactose competes and for which the sugars have affinities in the order: glucose > galactose = fructose. The disappearance of the substrate specificity and concentration dependency upon homogenization means that this otherwise necessary initial reaction is associated with cellular integrity and is rate-limiting to sugar utilization by the intact cell. These features of the initial reaction appear to describe a transport reaction. The same conclusion may be drawn independently from the results of kinetic studies with the intact cells alone (Figs. 3 and 4’1. The line de-
VOl. 80 TABLE
111
EFFECT OF SUBSTRATE, CONCENTRATION A N D OF GALACTOSE UPON AXAEROBICGLPCOLYSIS IN THE GARDXERASCITES
TUMOR Additions per vessel, &moles
Intact cells 16.7 glucose 16.7 fructose 1 6 , 7galactose 5 . 5 glucose 6 5 . 5 glucose 278 glucose 5 . 5 glucose 1100 galactose 5 5 . 5 glucose 1100 galactose 5 5 . 5 fructose 167 friictose 2 i 8 fructose \ - a l of a barrier t o entry. ( 3 6 ) J . J. lllutll a n d L). J. Jenderi, i l v ~ h b’ioclwn. . U i o p h j s . , 6 6 , 316 (1957).
dry C Owt./hr. dmg. All.
14.0 0.75 0 9.33 15.2 21.3 5.28 14.6 3.30 8.20 10.5 1.60
21.6 22.8 0 20.9 17.7 25.5 24.1
2T.1 25.8
where S is the concentration of substrate; Ire, the experimental velocity a t S ; V,, theFaxima1 experimental velocity for the system; K , the reciprocal Michaelis-Menten constant for the one intracellular enzyme which is responsible for the observed rate (presumably hexokinase in this case) ; T O , the cell radius and I1 is the permeability constant of the cell. I n this form of the equation the diffusion constant within the cells is assumed to be comparable to that for the substrate in aqueous solution. Plotting S / V , us. l / ( V m - ITe), the intercept on the S / V e axis (Fig. 4) where 1/( V m Ve) equals 0 gives the value of ro/311. Therefore the increasing value of this intercept with rising galactose level supports the conclusion that galactose reduced the permeability of the cell to fructose (see Addendum). Other Inhibitors of Tumor Glycolysis.-The i l l hibitory action of 2-deoxy-~-glucoseis considered e l s e ~ h e r e . ~ 7In contrast to yeast,3* the tumor glycolysis was unchanged in the presence of the Q antibiotic, nystatin. As with e r y t h r ~ c y t e s , ~insulin had no measurable effect upon hexose uptake in the Gardner cell or, in agreement with Crane, et a/.,’? in the Ehrlich cell. Certain steroid compounds, however, were effective inhibitors of fructolysis, but not of glucolysis, in the ascites tumor cells (Table IV) . 4 0 -1lthough the substrate speci(37) R1. Nirenberg and J. P. Hogg, Cancer Res,, in press. (38) J . 0. I.ampen, E. R . Morgan and P. C . Slocum, FedernLiuiL P v o r . , 16, h o . 1, 96’3 (1933).
(39) W Wilbrandt, E. Guensberg and H. T,auener, Ifel;,. Physiol. el pharmacol. Acta, 6, C20 (1947). (40) T h e selective inhibition of fructolysis by steroids has been \er!f i e d fur t h e melanoma S-91 (solid) and Krebs-2 (ascites) tumor b y Wood and Rurk (“Pigment Cell Biology,” Academic Press, Inc., New York, S. Y . , 1958. (Report of t h e Pigment Cell Conference,
Aug. 20, 195s
HEXOSETRANSPORT IN ASCITES TUMOR CELLS
ficity, the reversal of inhibition by increased fructose concentration and the loss of inhibition on homogenization resemble closely the galactose inhibition, the equal effectiveness of testosterone (and deoxycorticosterone) with either tumor is a marked contrast to the results for galactose. This difference means either that steroids and galactose act a t different sites on the transport mechanism or, less probably, that steroids do not affect the cellular permeability. Steroids of quite diverse hormonal activities were found to inhibit tumor fructolysis. This finding, however, does not rule out the possibility that the testosterone inhibition of fructolysis is the mode of control of fructose secretion in the male accessory glands by this hormone.41
4411
Discussion
Three assumptions necessary to our conclusions are that anaerobic CO2-production is a specific index of hexose uptake, that no important galactosesensitive unknown routes of glucose aiid fructose catabolism occur in these cells, and that the substrate-binding characteristics of hexokinase do nu 1 change upon homogenization. That the first assumption is substantially correct is demonstrated by the parallel analytical results of Table 11. Furthermore hexoses cannot be converted to glycogen or arise from endogenous glycogen in these tum o r ~ ,possibly ~? because phosphorylase is lacking.4 3 Although the hexose monophosphate shunt does occur a t low level in these tumors,44calculations based upon the distribution of carbon-14 from 1-, TABLE I\' 2-, or 6-labeled glucose into CO2 and lactic acid in Ehrlich ~ells,~ or5 into ribose phosphate in Gardner EFRECT OF STEROIDS AND OTHER COMPOUNDS ON AVAEROBIC cells,46 show that 90y0 or more of the glucose cataGLYCOLYSIS IN ASCITES TUMORS" bolism follows the Embden-Meyerhof pathway. cod mg I n homogenates T P N addition is necessary to elicit dry U't / catabolism via the shunt.47 pg./ SubPZ,/ hr. The second assumption is more difficult to supml. Test compound 6 strate rl . port directly because the process of homogenizaGardner cells tion conceivably could destroy an unknown cataNone Glucose 4.4 25 bolic pathway of hexoses. Glycolytic rates in 15 Glucose 4.4 22 Testosterone tumor homogenates, however, have always equaled Fructose 44.0 7 . 7 None or exceeded the rates in the cells (Tables I, 111); 15 Fructose 44.0 4 . 3 Testosterone therefore no loss of catabolic activity upon ho1 . 5 Fructose 44.0 6 . 7 Testosterone mogenization is apparent. Furthermore galactose 15 Fructose 44.0 4 . 5 Deoxycorticosterone has been shown to compete with ribose for uptake into Ehrlich ascites tumor ~ e l 1 s . I ~Since neither Ehrlich cells ribose48nor galactose is metabolized by the Ehrlich None Glucose 6 . 7 33 cell, they could scarcely compete in a catabolic reTestosterone' Glucose 6 . 7 32 20 action. Fructose 6 . 7 33 None The third assumption is necessary because a 20 Fructose Testosteroned 6.7 9.5 hexokinase with lower and differential hexoseTestosterone Fructose 67.0 32 20 affinities in the intact cell, which are increased 5 Fructose Testosterone 6 . 7 25 upon homogenization, could represent the obligaA1.'-Androstene-3,17tory transport reaction.49 Two factors make hexodione Fructose 6 . 7 22 20 kinase unsuitable for this role: sugars which are not A4-Androstene-3,17phosphorylated by hexokinase pass through the 20 Fructose 6 . 7 16.0 dione transport-reaction,12and erythrocytes transport su25 Diethylstilbestrol Fructose 6 . 7 12.0 gars well despite very slow phosphorylation.6J 25 Fructose 6 . 7 22 Progesterone Also the failure of cell fructolysis to equal homogDeoxycorticosterone Fructose 6.7 8.2 20 enate fructolysis even at 0.1 Jf fructose argues 25 Podophyllin Fructose 6 . 7 20 against this possibility.4g Gardner cell homogenate' The structural specificity both of passage and of Xone Fructose 3 . 7 27 inhibition of passage eliminates diffusion as the 20 Fructose 3 . 7 28 Testosterone rate-limiting step in hexose transport, as does also Deoxycorticosterone Fructose 3 . 7 27 20 the non-linear relationship between hexose concentration and utilization in the Gardner cell. Other In comparable experiments, estradiol, cortisone, cortisol, insulin, nystatin and epinephrine were without effect r e s ~ l t also s ~ have ~ ~ ~excluded ~ this possibility in ason either tumor. * Alcoholic solutions of the steroids were cites tumors. diluted 20-fold in Krebs-Ringer bicarbonate buffer imme-
diately before use. Each Warburg vessel received 0.5 ml. of steroid dispersion or of the alcoholic buffer. In comparable experiments, diethylstilbestrol, deoxycorticosterone, progesterone and podophyllin also had no significant effect on glucolysis. Three different preparations (Ciba or Schering) gave the same result; one of these contained an extraneous IO-fold quantity of cholesterol as diluent. All compounds listed in this table were found to be inactive also in Ehrlich cell homogenates a t the concentrations shown. Houston, Teras, November 14, 1957)). In addition, inhibition of glucolysis a t higher levels of steroids was found. (41) T. M a n n , "The Biochemistry of Semen," Methuen, London, 1954.
(42) hZ. W. Nirenherg and J. F. Hogg, unpublished. (43) L. Slechta, A. Jakuhovic and F. Sorm, Coll. Czechoslovak Chem. Comtn., 21, 24 (1956). ( 4 4 ) E. Racker, Ann. N. Y.Acnd. Sci., 63, 1017 (1956). (45) C. E. Wenner and S. Weinhouse, J . Biol. Chem., 222, 399 (1956). (46) C. E. Wenner, J. Hackney and J. Herbert, Proc. Amer. Assoc. Cnncer Res., 259 (1957). (47) E. Kravitz and A. J. Guarino, unpublished. (48) E. Kun, P. Talalay and H . G. Williams-Ashman, Cancer Res, 11, 855 (1951). (49) J. G. Kaplan, Ezpt. Cell Res., 8 , 305 (1955). (50) W. D. Yushok and W. G. Batt, Abstracts, 130th Meeting, American Chemical Society, p. 22-D,Atlantic City, September 1 6 21, 1956.
-1 selective inhibition of fructose permeability by a steroid, cortisone, was reported for the foetal erythrocytes of sheep.bl Bacila and GuzmanB a r r ~ npreviously ~~ found no effect of cortical steroids on Ehrlich tumor glucolysis, but glucolysis in the mouse diaphragm was inhibited a t low substrate concentration, the inhibitory action being attributed to a reversible ketosteroid reaction with sulfhydryl groups. Since various sulfhydryl reagents inhibit sugar transport in erythrocyte^,^ a possible mode of steroid inhibition is suggested. In agreement with this suggestion, the ketoses react much more poorly with cysteine than do the al~Ioses.5~ Addendum By J . J. B L U M ~ ~ The purpose of this addendum is to discuss briefly the ambiguities inherent in the kinetic analysis of constrained enzyme systems when inhibitors are present. Consider an enzyme, E, uniformly distributed in a system bounded by a membrane of permeability IT and with an internal diffusion constant D for the substrate whose (maintained) external concentration is S. The enzyme is supposed to catalyze the following reaction in the presence of an inhibitor I. E
k1 kz + S I_ J3.S + E + P
+I E.1
k.
,
1,
=
k&
for the rate oi fructolysis (cf. Fig. 3). Alternatively galactose may act to decrease the permeability, as is indicated by the lack of inhibition and loss oi concentration dependency in homogenates. If one first assumes that the permeability term is of negligible importance ( H 4 03 ) , then the above equation is a rearranged form of the MichaelisMenten equation with competitive inhibition, and the data of Fig. 4 give a value of K = 4.0 liters/rnole (computed from the slope of the line with 0 galactose). The values of K , computed from the slopes of the lines for 19, 77 and 115 pmoles ml. galactose are, respectively, (negative), (negative) and 20 liters/mole. Thus this assumption does not yield consistent values for K l , even though consistent values are obtainable from the Lineweaiw-Burk plot (Fig. 3 ) . The Lineweaver -Burk plot used OII whole cells, however, simply indicates a coinpetitive action of galactose, and offers no possibility of deciding whether the inhibitor affects permeability or an intracellular eniyme, or both If one assumes that the galactoqe affects only the , k , -- K,’ =. O ) , the slopes and the extrapolated intercepts of the lines in Fig. 4 yield the following values of the parameters, taking . ~ the ~ the radius of the cells as 5.2 X 10 - 4 c ~ nand cellular volume a s 4 5 x 10- cc. per mg. dry weight K, I . /mole
H (cm.1sec.j X I O + :
--
1 0 4 6
3 9
19 I 1
( 3 0-6 7 )
113
1 2
0 45 i Segative)
GaLictose, fimoles/ml.
+1
0
E.S.1
In the previously published method36 for obtaining R, Vm, D and H from experimental data, the presence of inhibitors was not considered. It can be shown that in the presence of an inhibitor the appropriate equation is
+
.I_~-_-
YO2
+
70
( p a)(p 4)D + G 7 3 i R where p = - 1, 0, 1 for a tissue slice, cylindrical cell and spherical cell, respectively. For the Gardner tumor cell fi = 1, and i t may be assumed that the term containing D is negligible compared to the permeability term. There are two opposite ways in which the inhibitor (galactose) may act. Possibly galactose acts as a competitive inhibitor (Ki’= 0) of the intracellular enzyme responsible
+
(51) A . Plrtscher, P. \’an Planta and U’. A . 1-Iunzing.r. If,!:, f i h r s i o l . e l pkaumacol. Acta, 13, 18 (1955). (52) M. Racila a n d E. S. Guzman-Barron, E i r i i o r r i n o l o g y , 64, .501
(ins$). (.3) PiI. 1’. Schubert, J . Bid. C ~ P I I L 130, . , 801 (1939). (.W Dept. of Riolorical Chemistry and l l e n t a l Health Institute.
1 4
Except for the highest value of galactose, where the data do not fit either assumption-presumably because some other phenomenon is occurring--the values of K are reasonably constant and the permeability decreases with increasing galactose concentration. IT-e have previously” emphasized the need for a large number of accurate determinations of the steady-state rate if one is to get reliable estimates o f the parameters. Furthermore the equation required to get K , l T mand €3 from kinetic data C)II whole cells is necessarily more sensitive to experimental variation than the Lineweaver-Burk equation. The presence of an inhibitor in the systeiii further increases the requirement for many accurate measurements if ambiguity is to be avoided Nevertheless the present relatively limited datx tend to support the conclusion that galactose decreases the permeability of the Gardner ascites tumor cell t o fructose and provides an order of magnitude estirncite of the permeability. \ v \ .IRBDR, M I C I I I\ ,\ ~
