kinetic parameters of hexose transport in hybrids between malignant ...

J. Cell Sci. 62, 49-80 (1983) Primed in Great Britain © The Company of Biologists Limited 1983

49

KINETIC PARAMETERS OF HEXOSE TRANSPORT IN HYBRIDS BETWEEN MALIGNANT AND NONMALIGNANT CELLS M. K. WHITE, M. E. BRAMWELL AND H. HARRIS Sir William Dunn School of Pathology, University of Oxford, Oxford 0X1 3RE, England

SUMMARY Matched pairs of isogeneic hybrid cells, in which one member of the pair was malignant and the other not, were used to examine the linkage between malignancy and functional alterations in hexose transport. The kinetic parameters of uptake of 2-deoxy-D-glucose were measured in a range of such hybrids, both human and murine. Some other malignant cell lines were also examined and were compared with non-tumorigenic derivatives of tumour cells selected by exposure to the lectin, wheat-germ agglutmin. In every case, malignancy, as denned by the ability of cells to grow progressively in vivo, was found to be linked to a decrease in the Michaelis constant of hexose uptake. Independent measurement of the transport and phosphorylation reactions involved in hexose uptake revealed that this decrease was determined by the membrane transport system. The difference in Michaelis constant between malignant and non-malignant cells was observed with 3-0-methylglucose, a hexose that is transported into the cell but not further metabolized. The activity of hexokinase in cell homogenates was higher than the level that would be required to cope with transport and showed no correlation with tumorigenicity. Measurement of the uptake of D-glucose itself, by a rapid nitration centrifugation method, gave results similar to those obtained with 2-deoxy-D-glucose. INTRODUCTION

Hybrids generated by the fusion of malignant with non-malignant cells provide a segregating genetic system in which malignancy, defined as the capacity of cells to produce progressive tumours in a suitable host, can be correlated with biochemical or cytological parameters measured in vitro (Harris, 1971; Straus, Jonasson & Harris, 1977; Watt, Harris, Weber & Osborn, 1978; Stanbridge& Wilkinson, 1978). Initially such hybrids are usually non-malignant, but on continued cultivation in vitro they may give rise to segregants in which malignancy has re-appeared (Klein, Bregula, Wiener & Harris, 1971; Wiener, Klein & Harris, 1971, 1974a; Stanbridge, 1976; Sager & Kovac, 1978). Hybrids in which malignancy is suppressed and malignant segregants derived from them thus constitute isogeneic matched pairs, which provide a screen to test whether or not a particular cellular property co-segregates with malignancy. In this way, Bramwell & Harris (1978a) showed that malignancy is systematically linked to a structural alteration in the carbohydrate moiety of a particular membrane glycoprotein. Further characterization of this glycoprotein provided circumstantial evidence that it had some role in glucose transport (Bramwell & Harris, 19786; Bramwell, 1980; Bramwell & Atkinson, 1982; Gingrich, Wouters, Bramwell & Harris, 1981o,6).

50

M. K. White, M. E. Bramwell and H. Harris

In the present work, isogeneic matched pairs of hybrids were examined to see whether any functional alteration in hexose transport was associated with malignancy. Non-malignant variants derived from tumour cells by selection for resistance to the lectin wheat-germ agglutinin (WGA) were also examined. There has been little previous work, and no formal kinetic study, on the relationship between glucose uptake and tumorigenicity. However, many investigations of hexose uptake have compared virus-transformed cells with untransformed controls. Hatanaka, Huebner & Gilden (1969) were the first to report that the uptake of glucose was enhanced on viral transformation. In this study, and several other early studies, transformation was associated with a large reduction in the apparent Km for glucose uptake (Hatanaka & Hanafusa, 1970; Hatanaka, Todaro & Gilden, 1970; Hatanaka, 1971; Hatanaka, Gilden & Kelloff, 1971). However, since glucose was the hexose used in these studies, uptake rates reflect both transport and subsequent metabolism, so that it is difficult to interpret the apparent Km values (for discussion see Plagemann & Richey, 1974). Many other studies have used 2-deoxy-D-glucose, an analogue of glucose that is transported into cells, phosphorylated by hexokinase, but not further metabolized (Renner, Plagemann & Bernlohr, 1972; Weber, 1973). Kinetic analysis of the enhancement of uptake of this sugar on viral transformation has given conflicting results. Some investigators have reported a reduction in the Km of uptake (Hatanaka, Augl & Gilden, 1970; Hatanaka & Hanafusa, 1970; Hatanaka, 1971; Hino & Yamomoto, 1971;Bader, 1972; Gazdaretal. 1972; Bradley & Culp, 1974; May, Somers & Kit, 1974). However, it has been suggested that this fall in Km is due to a failure to ascertain properly the initial rates of uptake (Plagemann, 1973). More careful analysis of deoxyglucose uptake kinetics has shown that transformation results in an increase in the VmiX for hexose uptake, the Km remaining constant (Isselbacher, 1972; Plagemann, 1973; M. J. Weber, 1973; Kletzien&Perdue, 19746, 1975; Royer-Pokoraef a/. 1978). This conclusion is supported by studies of the uptake of 3-O-methyl-D-glucose, which is transported into cells but not further metabolized (Renner et al. 1972; Weber, 1973). Again, a kinetic analysis indicated that the Vm^x was increased on viral transformation, but the/Cm was unaffected (Venuta & Rubin, 1973; Weber, 1973; Kletzien & Perdue, 19746). Viral transformation in vitro thus appears to result in an increased Vmex of the membrane hexose transport system and is apparently due to an increase in the number of carrier molecules in the cell membrane (Salter & Weber, 1979; Salter, Baldwin, Lienhard & Weber, 1982). In the present study the relationship between tumorigenicity and hexose transport was explored, initially by measuring the kinetic parameters of uptake of 2-deoxy-Dglucose. Significantly lower Km values were found for all tumorigenic cells, whether hybrid or not, compared to their non-tumorigenic homologues. Further studies of this phenomenon with 3-O-methyl-D-glucose confirmed that the reduction in Km was due to a change in the transport of hexose across the cell membrane. Finally, similar Km values were obtained with D-glucose, where the uptake was measured by a rapid assay method in which subsequent metabolism did not significantly complicate the rate measurements.

Hexose transport in hybrid cells

51

Preliminary reports of this work have been published (White, Bramwell & Harris, 1981, 1982).

MATERIALS

AND

METHODS

3

2-Deoxy-D-[l- H]glucose, 3-O-methyl-D-[l-3H]glucose, D-[U-l4C]glucose, L-[l-14C]glucose, inulin-[HC]carboxylic acid and 2-deoxy-D-[l-14C]glucose were obtained from Amersham International Ltd, Bucks, U.K. L-[l-3H]glucose was obtained from New England Nuclear, Boston, Massachusetts, U.S.A. Unlabelled 2-deoxy-D-glucose was from Calbiochem-Behring Corporation, La Jolla, California, U.S.A.; and L-glucose and 3-O-methyl-D-glucose were from Sigma Chemical Company Ltd, Poole, Dorset, U.K.

Cell lines Mouse PG19: hypoxanthine-guanine phosphoribosyl transferase-deficient (HGPRT") derivative of a spontaneous melanoma arising in a C57B1 mouse (Jonasson, Povey & Harris, 1977). SEWA: polyoma virus-induced osteosarcoma arising in an A/Sw mouse (Sjogren, Hellstrom & Klein, 1961). A9HT: malignant derivative of the A9 cell line selected by passage in vivo (Wiener, Klein & Harris, 1973). YACIR: HGPRT" derivative of a Moloney virus-induced lymphoma arising in an A/Sn mouse (Fenyo, Klein, Klein & Swiech, 1968). TA3 Hauschka: spontaneous mammary carcinoma arising in an A/Ha mouse (Hauschka, 1953). TA3HaB: HGPRT" derivative of TA3 Hauachka (Wiener, Klein & Harris, 19746). MSWBS: methylcholanthrene-induced sarcoma of the A/Sw mouse (Klein & Klein, 1958). P3/NS/l-Ag4: HGPRT" derivative of a myeloma arising in a BALB/c mouse (Kohler, Howe & Milstein, 1976). 3T3A31 clone 9: fibroblast-like embryonic culture from BALB/c mouse (Antoniades, Stathakos & Scher, 1975). PG19-G: derivative of PG19 selected for ability to grow in a low concentration of glucose (Bramwell, 1980). SV-3T3-B: SV40 virus-transformed derivative of the BALB/c 3T3 cell line (Todaro, Habel & Green, 1965). SV-3T3-BT: SV-3T3-B tumour explanted from a nude mouse. PG19-WGAR clone C2: derivative of PG19 resistant to 10/jg/ml WGA (Chan, 1980). A9HTWC: non-malignant WGA-resistant A9HT derivative (Chan, 1980). A9HTWD: malignant WGA-resistant A9HT derivative (Chan, 1980). Rat N R K T G 3 : HGPRT" rat kidney cell line (NRK) (Marshall, 1980). 3.B77.Sc4: NRK cell line transformed by Rous sarcoma virus strain B77 (Marshall, 1980).

Human MRC5: fibroblast strain from lung of male foetus of 4 months gestation (Jacobs, Jones & Bailie, 1970). S1814: fibroblast strain from male foetus (Klinger, 1980).

52

M. K. White, M. E. Bramwell and H. Harris DS1: fibroblast strain from spontaneous male abortus (Sir William Dunn School of Pathology, unpublished). GM1604: foetal fibroblast cell strain (Der & Stanbridge, 1981). WI38: fibroblast strain from lung of female foetus of 3 months gestation (Hayflick & Moorhead, 1961). RT112/84: cell line derived from carcinoma of bladder (Marshall, Franks & Carbonell, 1977). HeLa (Flow): cell line derived from carcinoma of cervix (Gey, Coffman & Kubicek, 1952) supplied by Flow Laboratories, Irvine, Scotland, U.K. HeLa spinner: derivative of HeLa selected for growth in spinner culture (Sir William Dunn School of Pathology, unpublished). HeLaD98F908A3: HGPRT" HeLa derivative (Klinger, 1980). HeLa D98AH2: HGPRT" HeLa derivative (Der & Stanbridge, 1981). H29/219: cell line derived from a carcinoma of the rectum (Marshall et al. 1977). H.Ep.2: cell line derived from a carcinoma of the larynx (Moore, Sabachewsky & Toolan, 1955). RPMI2650: cell line derived from a carcinoma of the nasal septum (Moore & Sandberg, 1964). HT55F: cell line derived from a carcinoma of the colon (donated by Professor J. F. Watkins, University of Wales, unpublished). H T U 5 : cell line derived from a carcinoma of the colon (donated by Professor J. F. Watkins, unpublished). Daudi: lymphoid cell line derived from a patient with Burkitt's lymphoma (Klein et al. 1967).

Measurement of deoxyglucose uptake Cells growing as a monolayer. The cells were brought into suspension in Eagle's minimal essential medium supplemented with 10% foetal calf serum or 5 % foetal calf serum plus 5 % newborn calf serum (5-15 (XlO4) cells per ml) and evenly distributed into 48 16-mm diameter tissue-culture wells (Coster, Cambridge, Mass., U.S.A.). A sample (2 ml) of cell suspension was added to each well 24 h before assay. The adherent cell monolayers were washed twice with 500/il of phosphate-buffered saline (PBS), and 500/il of a solution of 2-deoxy-D-[l-3H]glucose (0-1-5 mM, l-210mCi/mmol) in PBS added at 20 C C. After 1-10min at 20°C, the radioactive medium was aspirated and the monolayer washed twice with 600/*1 of PBS. The cell monolayers were dissolved in 440^/1 0-4MNaOH, neutralized with l0f.ll glacial acetic acid and then mixed with lOmlUnisolve 1 (Koch-Light, Colnbrook, Berks, U.K.). The activity of tritium in these samples was determined with a Packard scintillation counter. In each experiment six wells were assayed for each concentration of 2-deoxyD-glucose. Uptake was usually measured at six different concentrations. Non-specific retention of radiolabel was measured with L-[l-3H]glucose under parallel conditions. Measurements were made on six wells and averaged. The number of cells per well was determined by detaching the cells with trypsin, transferring them to 10 ml of Isoton (Coulter Electronics, Dunstable, Beds, U.K.) and counting 500/il samples with a Coulter counter (Coulter Electronics). Measurements were made on six wells and averaged. The velocity of sugar uptake, V\, at each substrate concentration, S;, and the variance of each velocity estimate, VarV,, were calculated by linear regression through the origin of the plot of 2-deoxy-D-glucose uptake against time. The Km and Vmtx values and their standard errors were calculated by a weighted linear regression of S/V against 5, essentially as described by CornishBowden (1976). The final weight applied to each Sj/V, was [Ki2/VarVi] [ V 2 ™ / ^ + Si)2). The validity of this approach was confirmed by inspection of residuals in V (Cornish-Bowden, Porter & Trager, 1978) and by comparison of the results with those obtained by non-parametric analysis of the same data (Porter & Trager, 1977). Cells in suspension. This method was used with cells that grew in suspension culture or ascites


53

in vivo and also with lymphocytes. The cell number was determined in a haemocytometer and viability estimated by dye exclusion. The cells were then distributed into microfuge tubes (3-20 (X 105) cells per tube), pelleted by centrifugation for 4s in a Beckman Microfuge B and washed with PBS. The uptake measurements were started by resuspending the cells in 300/d of 2-deoxy-D[l- 3 H]glucose. The same concentrations and specific activities of radiolabel were used as for cells in monolayers and incubation times ranged from 1-5 min. Cells were.collected by centrifugation (4s) and rapidly washed in ice-cold PBS. Non-specific retention of label was again measured with L-[l-3H]glucose. The kinetic parameters were calculated as for cells in monolayers.

Measurement of 3-O-methyl-D-glucose uptake Uptake of 3-O-methyl-D-glucose into cells growing as monolayere was measured in the same way as the uptake of 2-deoxy-D-glucose, but with the following modifications: (a) 3-O-methyl-D[l- 3 H]glucose (0-1-5 mM, 10-500 mCi/mmol, 50/jCi/ml) was added to the cells in a volume of 250/il at 20 C C; (b) uptake was terminated by washing the cells rapidly twice with ice-cold PBS; (c) shorter incubation times were used (< 2 min).

Measurement of hexokinase activity The activity of hexokinase in homogenates of cultured cells was determined essentially as described by Kletzien & Perdue (1974a). Homogenates were prepared by ultrasonication of washed cell suspensions in lml of PBS (107— 10s cells). The cells were given four 5-s pulses with an MSE100 W ultrasonic disintegrator at 0°C. Each pulse was separated by a 25 s interval. The reaction mixture for determining hexokinase activity consisted of lOmM-HEPES buffer (pH7-4), 10mM-ATP, 10mM-MgCl2l 10mM-KCl, 3mM-NADP + , 12mM-glucose, 2 units of glucos*-6-phosphate dehydrogenase and 10— 50/d of cell homogenate in a total volume of 1-5 ml. The reaction mixture was incubated at 20 °C and the reduction of NADP + was measured by the increase in absorbance at 340 nm. Reaction mixture without homogenate was used as a blank. The protein content of each sample was measured by the method of Lowry, Rosebrough, Farr & Randall (1951), with bovine serum albumin as a standard.

Measurement of D-glucose uptake Uptake of D-glucose during very short incubations (5-30 s) was measured by the silicone oil filtration centrifugation technique described by Werdan et al. (1980). In order to measure the uptake of glucose by this method, it is necessary for the cells to be in suspension. Monolayers of cells were detached with PBS containing 0-2% (w/v) EDTA (disodium salt). After cell number and viability were determined, cells were resuspended in uptake buffer at a density of between 5 X 10s and 1 X 106 cells per ml. Uptake buffer consisted of 137 mM-NaCl, 5-4 mM-KCl, 4 8 mM-NaHCO 3 , OlmM-EDTA, 0 0 0 1 % (w/v) phenol red, 20mM-HEPES and 25/iCi/ml 3 H 2 O (Amersham International Ltd), pH7-4. The uptake incubation was done in 400/xl polyethylene tubes (Beckman). These tubes contained 20^1 of 1 M-HCIO4 with an overlay of 70^1 of silicone oil. The silicone oil was a mixture of AR20 and AR200 oils in a ratio of 1:1-5 (Wacker Chemie, Munich). A total of 250[A of cell suspension was layered onto the silicone oil in each tube and the uptake incubation was started by adding 10 fA of D-[U-MC]glucose (7-260 mCi/mmol) to give a final concentration of 0-1 mM-5mM. After 5-30 a at 20 °C, the incubation was ended by centrifuging the cells for 10 s through the layer of silicone oil into the perchloric acid. This was done in a Beckman microfuge B. The activity of 14C was used as a measure of the amount of D-glucose associated with the cells, and the activity of 3 H as a measure of the volume of material passing through the silicone oil. The activity of each isotope in both the supernatant layer and the perchloric acid was determined with a Packard scintillation counter preprogrammed to give 14C and 3 H disintegrations per minute for doubly labelled samples. A significant amount of extracellular material passed through the silicone oil with the cells. The contribution of this to the total uptake was determined in two ways: (1) tubes were assayed with either L-[l-MC]glucose (61-4mCi/mmol, final concentration 0-1 mM) or inulin-[14C]carboxylic acid (7-8 mCi/mmol, final concentration 1 £JM), both of these beine excluded from the intracellular space; (2) uptake of D-glucose was extrapolated to zero time. L-[1- C]gluco«e measurements were found to give the most accurate estimate of extracellular space. The amount of glucose taken up into

54


the intracellular space was calculated by subtracting the extracellular glucose from the total amount of glucose in the perchloric acid fraction. This was then divided by the volume of the intracellular space to give the amount of glucose taken up per unit volume of intracellular space. This was expressed in picomoles per nanolitre. A more detailed account of these calculations is given by Werdanefa/. (1980). Six nine-point time courses of D-glucose uptake against time at different D-glucose concentrations were measured. Rates of uptake and their variances were again calculated by linear regression through the origin, and the weighted linear regression method was applied to these data to determine Km and Vmu , as described for 2-deoxy-D-glucose.

Effect offluoride on glucose uptake Werdan et al. (1980) demonstrated that D-glucose uptake measurements made by this method may not be affected by loss of intracellular isotope due to glucose metabolism. To test this, YACIR cells were preincubated with 1 mM-sodium fluoride for 20min at 20 °C. Under these conditions fluoride gave a 90 % inhibition of glycolysis as determined by lactate production (measured by the method of Everse, 1975), with no significant loss of viability. Uptake of D-glucose was measured with D-[U- H C]glucose (0 - l ITIM, 300mCi/mmol) by the method described and compared to that in untreated YACIR cells.

Measurement of 2-deoxy-D-glucose uptake by silicone oil filtration centrifugation The uptake of 2-deoxy-D-glucose into YACIR cells was measured with 2-deoxy-D-[l-14C]glucose (O'lmM, 58-SmCi/mmol) and compared to D-[U-14C]glucose (0-1 min, 300mCi/mmol) uptake into the same batch of cells.

Measurement of L-glucose uptake by silicone oil filtration centrifugation The silicone oil filtration centrifugation method was modified in the following ways to measure L-glucose uptake: (a) L-[l- H C]glucose (61-4mCi/mmol) was added to give a final concentration of 0-1 mM; (b) the length of incubation was increased to between 60 and 180 min, as the uptake is very slow; (c) between three and six tubes were assayed in each experiment, and the rate of uptake calculated by dividing the mean amount of L-glucose taken up by the time of incubation; the uptake of D-glucose (O'l mM, 300 mCi/mmol) was measured in the same batch of cells for comparison. RESULTS

Uptake of 2-deoxy-D-glucose The time courses of deoxyglucose uptake into cells were usually found to be linear for 5—10 min, thus allowing the rate of uptake to be calculated by linear regression. The variation of these rates with deoxyglucose concentration obeyed MichaelisMenten kinetics for most cell lines. This allowed the determination of the kinetic parameters of deoxyglucose uptake in a large range of cell lines. Tables 1 and 2 give the Km and Vmix values for the non-human tumour cells and the non-malignant cells tested. Included in these tables are the parental tumour cells and the diploid cells that were fused to give the hybrids investigated. The Vmax values are highly variable ranging from 4-5 to 495 nmoles/106 cells per h. Comparison of Tables 1 and 2 shows that there is no correlation between tumorigenicity and a high Vmmx in vitro. Where replicate determinations were made, there was variability in the replicate V™,, values, but less than that between different cell types. Variation in Vm^t is presumably due to factors such as cell density and other conditions of culture. Some cell types (e.g. YACIR, BALB/c fibroblasts and A/Sn fibroblasts) give a reproducible Vmax value.


55

Table 1. Kinetic constants of hexose uptake in malignant mouse cells V v

No. of concentrations assayed

Km (miu)

(nmol/10 6 cells perh)

8-8 X10 2-5 x 10s 1-2X10 5 1-5 x 10s 1 0 x 10s 7-8 x 104 9-7 x 104

6 4 6 6 6 4 6

1-104 ±0-066 0-608 ±0-114 0-992±0-119 1-364 ±0-074 0-779 ± 0-084 0-834 ±0-162 1-126 ±0-079

28-6 ± 0-9 961 ± 7-8 108-2 ± 7-8 83-2± 3-0 81-7± 4-6 132-5 ±10-8 60-8 ± 2-6

YAC1R

Suspension culture

5 6

1-154 ±0-244 1-101 ±0-033

46-9 ± 6-3 44-7 ± 0-8

A9HT

2-1 X 10s

5

1-170 ±0-079

131-4 ±

SEWA

Suspension culture

6 4

0-520 ±0-073 0-683 ±0-008

200-0 ±20-3 494-7 ± 2-6

Cell type PG19

SEWA TA3HaB TA3

Cell density (cells/cm2) 4

majc

6-3

Ascites

5

0-369 ±0-100

299-2 ±18-4

2 0 X 105

5

0-762 ± 0-078

291 -2 ±16-7

Suspension culture

6

0-377 ±0-042

122-8 ± 5-3

Ascites

5

0-870 ±0-024

134-4 ± 2-1

P3/NS/l-Ag4

Suspension culture

6

1-495 ±0-229

47-7 ± 3-2

MSWBS

Suspension culture

4

1-378 ±0-268

161-4 ±18-5

6

0-846 ±0-049

TA3

PG19-G

5

1-5 X 10

50-7 ±

1-8

There is a clear difference in/Cm between the tumour cells and the normal cells. The Km values of the tumour cells range from 0-37 to 1-49 HIM while those of the nonmalignant cells range from 1-61 to 3-69mM. There is no unique Km value for either malignant or non-malignant cell types. For example, there are differences in Km between fibroblasts derived from different inbred strains of mouse, which is probably due to genotypic variation since replicate determinations were made with cells from different animals. There is reasonable agreement between replicate determinations on the same cell type indicating the reproducibility of the/C value. The small variation between replicates can probably be accounted for by experimental variation. The question of whether the reduced Km value observed for the tumour cells is linked to malignancy can now be explored by comparing matched pairs of hybrids as described above. Deoxyglucose uptake by lymphoma X fibroblast hybrids The hybrids described in this section were derived from clone 1G. This was produced by the fusion of YACIR lymphoma cells with CBAT6T6 fibroblasts and has been described previously (Jonaason et al. 1977). The clone originally was nontumorigenic but, after several weeks cultivation in vitro, generated malignant

M. K White, M. E. Bramwell and H. Ham's

0 0 0

~~~~

0 0 0 0 0

0 0

+I ti +I

+I ti +I ti +I

ti ti +I

N-b

~ m w - c o cob W N m-.

v,

2

N

0

0 0 0

3 ~ 2

W

W v l N

*-?

N ~ U ~

b

N

0 0

0

0

8; $ $ 8

+I +I ti ti +I

0

+I ti +I

-s 3 2

0

0 0

ti +I

+I

ti +I

0

0

0 G ? "Z


57

segregants, which overgrew the cultures. Early passages of clone 1G were recloned and the secondary clones tested for tumorigenicity. Two of these secondary clones, 1G1 and 1G8, were selected for further study. Clone 1G8 produced no tumours with inocula of 5 X 105 cells per mouse, while clone 1G1 produced tumours in about 90% of the animals with this inoculum. Three such tumours derived from clone 1G1 were explanted and grown in vitro (clones 1G1T1, T2 and T3). The kinetic constants of deoxyglucose uptake for clone 1G8 and these three tumours are given in Table 3. Hanes (1932) plots of the uptake kinetics of clones 1G8 and 1G1T2 are shown in Figs 1 and 2. The Km for deoxyglucose uptake in the three tumours is approximately one third of that of the non-tumorigenic hybrid 1G8. When the deoxyglucose uptake kinetics of clone 1G1 were measured, it was found that the reciprocal Hanes plots were markedly non-linear (Fig. 3) although the tumours derived from 1G1 gave linear plots (e.g. see Fig. 2). This curvature was reproducible and there are several possible explanations for it: (a) in addition to facilitated diffusion, deoxyglucose might enter clone 1G1 cells at a significant rate by non-carrier-mediated diffusion; (b) clone 1G1 cells might possess two uptake systems with different Michaelis constants; (c) clone 1G1 might be a mixed population of cells with differing Michaelis constants for deoxyglucose uptake. Table 3. Kinetic constants of hexose uptake in malignant and non-malignant lymphoma X fibroblast hybrids

Hybrid

Tumorigenicity


Clone 1G8 Clone 1G8 Clone 1G8

4 4 5

Clone 1G1

6

* IIIU

Cell density (cells/cm2) 4

5-7 X 10 6-4 X 104 4-6 X 104

Km (mM)

(nmol/10 6 cells perh)

3-214 ±0-255 3-500 ±1-020 3-202 ±0-182

140-7 ± 8-9 294-5 ± 57-2 203-3+ 8-8

Non-linear reciprocal plot 4

Clone Clone Clone Clone

1G1T1 1G1T2 1G1T2 1G1T3

6 5 6 6

4-7 5-5 1-1 8-9

Clone Clone Clone Clone Clone Clone Clone Clone Clone

1G1A 1G1E 1G1F 1G1G 1G1G 1G1H 1G1H 1G1J 1G1J

6 6 5 4 4 4 6 5 6

1-6 X 8-9 X 2-1 X 8-9 X 8-8 X 2-1 X 9-3 X 1-9 X

Non-linear reciprocal plot 105 1-461 ±0-208 37-7 ± 631 ± 1-198 ± 0 1 6 5 104 13-5 ± 1-087 ±0-170 105 48-0 ± l-312±0-155 104 41-6± 104 0-666 ±0-107 22-9 ± 1051 ±0-097 105 104 57-6 ± 0-938 ±0-198 37-1 ± 105 1-353 ±0-203

Clone lG8a Clone lG8b Clone lG8c

5 5 5

7-5 1-5 6-0

X X X

104 105 104

0-851 ±0-068 1-413 ±0-263 0-835 ±0-047

74-8 ± 2-9 62-8 ± 8-5 1461 ± 4-2

Clone lG8bTl Clone lG8bT2

6 6

5-2 6-2

X X

104 104

1-281 ±0-168 1-349 ± 0 1 1 0

135-9 ± 6-8 179-5 ± 8-8

X X X X

10 104 105 104

0-725 ±0-098 1-088 ±0-336 0-888±0115 1-063 ±0-157

106-1 ± 7-6 86-9+13-9 156-2 ± 8-9 242-6 ±21-9 3-2 4-8 1-4 4-6 3-6 0-9 6-5 3-4

CEI.62

58

M. K. White, M. E. Bramwell and H. Harris S V (mM (/imol/10 cells per h)~ 60-

50-

40-

30-

20-

10

-5

-4

-3

-2

-1

0 S(rnM)

Fig. 1. Reciprocal plot of 2-deoxy-D-glucose uptake in the non-malignant hybrid YACIR X CBAT6T6 fibroblast clone 1G8. The x intercept is -Km. Bars equal two standard errors. Standard error of 5 / V i s (S./V?) X VarV,.

s V (fimol 106 cells per h)~')

60-

50-

40-

30-

20-

10-

-

5

-

4

-

3

-

2

-

1

0 S(mM)

1

2

3

4

5

Fig. 2. Reciprocal plot of 2-deoxy-D-glucose uptake in the malignant hybrid YACIR X CBAT6T6 fibroblast clone 1G1T2.


59

S V (^mol/10° cells per h)"1) 60-

50-

40-

30-

20-

10-

-5

-4

-3

-2

-1

Fig. 3. Reciprocal plot of 2-deoxy-D-glucose uptake in the malignant hybrid YACIR X CBAT6T6 fibroblast clone 1G1.

The first possibility was tested by measuring the amount of L-glucose (at a concentration of 0-1 mM) that was taken up by 1G1 cells in 10 min. It was found that the L-glucose uptake was less than 2 % of that of deoxyglucose at the same concentration. This non-specific uptake cannot therefore account for the degree of curvature obtained. In order to test whether clone 1G1 was a mixed population of cells, this clone was recloned to give tertiary clones 1G1A,E,F,G,H and J. When these clones were assayed, they each gave linear reciprocal plots with the exception of clone 1G1A, which still gave a curvilinear plot. The kinetic constants of the tertiary clones are given in Table 3. These results support the view that clone 1G1 is a mixed population containing non-malignant cells and malignant segregants. Thus, on recloning, or on selection of the malignant subpopulation by growth in vivo (clones 1G1T1, T2 and T3), linear reciprocal plots with low Km values are obtained. As no tumours were produced by direct injection of clone 1G8, it was of interest to attempt to select malignant segregants from this clone in order to compare their uptake kinetics with those of 1G8. To this end, subclones were selected for their ability to grow in agarose, in the hope that such subclones might be enriched for cells capable of progressive growth in vivo. From three dishes, each seeded with 105 cells in agarose as described by Steinberg & Pollack (1979), five primary colonies were obtained, of which three survived isolation and further subculture (clones lG8a, b and c). These gave 80-100 % take incidences at 5 X 105 cells per mouse. Hence, three tumorigenic derivatives of 1G8 were prepared, each arising by a separate segregation

60


event. Table 3 shows the Km and Vmax values for these clones and for tumours derived from them (clones lG8bTl and T2). Both the clones themselves and the tumours produced by them have Km values two to three times lower than that of the nontumorigenic clone 1G8 from which they were derived. In this cross, the low Km for deoxyglucose uptake characteristic of the malignant parent is associated with tumorigenicity in the hybrids. Tumorigenic derivatives obtained directly by the inoculation of hybrid cells into the animal or indirectly by selection in semi-solid medium systematically show a reduction in Km compared to the non-tumorigenic hybrids from which they were derived. Deoxyglucose uptake by melanoma X fibroblast hybrids Clone 7 and clone 8 are non-tumorigenic hybrids produced by fusion of the PG19 melanoma derivative and diploid fibroblasts homozygous for the T13H translocation (Jonasson et al. 1977). The kinetic constants of these clones and a tumour produced from clone 8 are given in Table 4. The hybrids in which tumorigenicity is suppressed show/Cn values that are much higher than the parental PG19 tumour cells; the tumour derived from clone 8 shows a marked reduction in Km compared with that of clone 8 itself. Deoxyglucose uptake by a melanoma derivative selected for resistance to wheat-germ agglutinin Several derivatives of PG19 cells have been produced by selection for their ability to grow in the presence of the cytotoxic lectin WGA (Bramwell & Harris, 1978a). One such line (C) produced no tumours in 17 animals with inocula of 5 X 104 cells per animal. This line was recloned and one of the secondary clones, PG19WGAR clone C2, was assayed. This non-tumorigenic clone had a Km approximately 2-5-fold higher than the PG19 cells from which it had been derived (Table 4). The production of non-tumorigenic clones by selection with WGA provides an additional test for the Table 4. Kinetic constants of hexose uptake in melanoma X fibroblast hybrids and in a lectin-resistant melanoma derivative

Cell type

No. of Tumori- concentrations Cell density genicity assayed (cells/cm2)

Km (mM)

K™, (nmol/10 6 cells per h)

PG19XT13HT13H Clone 7

-

6

2-lxlO5

2-348 ±0-339

1950 ±16-4

PG19XT13HT13H Clone 8

-

4 6

5-3 x 104 7-9 xlO 4

3-587 ±0-572 2-533 ±0-466

160-5 ± 18-4 132-5 ± 16-2

PG19XT13HT13H Clone 8T1

+

6

2 1 x 10s

1-480 ±0-051

204-3 ± 3-8

PG19WGAR Clone C2

-

5

2 0 x 105

2-400 ±0-230

173-5 ±11-5


61

association of markers with malignancy (Bramwell & Harris, 1978a). In this case, loss of tumorigenicity of the PG19 melanoma is associated with loss of the low Km characteristic of tumour cells. Deoxyglucose uptake by fibrosarcoma X lymphocyte hybrids The tumorigenicity of the malignant L-cell derivative A9HT can be suppressed by fusion with mouse diploid lymphocytes (Wiener et al. 1974a). Assay of two such suppressed clones (A9HT X C57B1 lymphocyte clones 3 and 4) showed their Km values to be of the order of 2 HIM (Table 5), while the parental tumour cell had a Km of the order of 1 mM (Table 1). A tumour derived from another such clone (A9HT X C57B1 lymphocyte clone 2T1) resembled the parent tumour cell A9HT in its Km value (Table 5). Deoxyglucose uptake by fibrosarcoma derivatives selected for resistance to wheatgerm agglutinin Selection of A9HT cells for resistance to WGA produced some derivatives that had lost tumorigenicity and others that had retained it (Chan, 1980). Clones A9HTWC and A9HTWD were both resistant to WGA at a concentration of 25^ig/ml, but whereas clone WC produced no tumours with inocula up to 6-6 X 106 cells per animal, clone WD remained as tumorigenic as the original A9HT cell line. Table 5 shows that the tumorigenic clone WD had a low .Km (0-731 ± 0-087 mM), whereas that of the nontumorigenic clone WC was more than twice as high (1-685 ± 0-151). Deoxyglucose uptake by polyoma virus-induced osteosarcoma (SEWA) X fibroblast hybrids As previously described (Wiener et al. 1971; Jonasson et al. 1977), hybrids between SEWA and diploid cells are highly unstable and generate malignant segregants at a high frequency, so that the tumorigenicity of the hybrid clones is variable. Table 5. Kinetic constants of hexose uptake infibrosarcomaX lymphocyte hybrids and in lectin-resistant fibrosarcoma derivatives

Cell type A9HT X C57B1 lymphocyte clone 3 A9HTXC57B1 lymphocyte clone 4 A9HT X CS7B1 lymphocyte clone 2T1 A9HT WC A9HT WD

No. of Tumori- concentrations Cell density genicity assayed (cells/cm2)

Km (miu)

(nmol/106 cells per h)

6

2-6 x 10s

2-240 ±0-160 247-6 ±13-2

6

2-5 xlO 5

2-080 ±0-120 451-3 ±20-8

5

2-OxlO5

l-123±0-046

6 6

1-8 xlO 5 1-8 xlO 5

1-685 ±0-151 133-9 ± 7-6 0-731 ±0087 120-4 ± 7-9

119-4 ± 2-8


62

Table 6. Kinetic constants ofhexose uptake in osteosarcoma X fibroblast hybrids Cell density (cells/cm2)

Hybrid clone


•3x 10= •2X10= •6X105 1-7 xlO 5 '•0x10= 1-6x10= 1-4 x 10= 2-5 x 105 1-3x10= 2-0x10=

Bl B Dl

D J2 C1T2 C1T1 FT3

J3 D3

5

6 5 6 6

6 6 6 6 6

Km (mM)

(nmol/106 cells perh)

•058 ± 0-086 167-6± 11-3 109-1 ± 2-6 ()-877 ± 0-035 1-390 ±0-040 115-2± 2-0 243-3 ±12-6 1-909 ±0127 276-9 ± 4-2 1-548 ±0-040 ;'•023 ±0-258 295-9 ±23-5 1-493 ±0-072 377-1 ±12-3 i-746±0-310 249-1 ±29-9 Non-linear reciprocal plots Non-linear reciprocal plots

Table 7. Effect of continuous culture in vitro on the kinetic parameters ofhexose uptake in osteosarcoma X fibroblast hybrids Hybrid clone Cl

F

No. of No. of days in continuous concentrations assayed culture 3 7 9 16 22 6 13 40

6 6 6 4 5 6 6 5

Cell density (cells/cm2)

Km (mM)


2-1 X 105 1-4 x 10s 16X10= 2-0x10= 2-9x10= 1-8x10= 2-4 x 10= 6-3 xlO 4

2-986 ±0-140 2-499 ± 0-267 1-896 ±0130 l-156±0-108 1-420 ±0-376 3-218 ±0-529 2-220 ±0-052 1-363 ±0-092

445-3 ±16-3 243-6 ± 19-1 354-6 ±17-2 129-5 ± 9-4 462-6 ±39-0 372-1 ±37-6 723-0± 11-0 240-6 ± 7-4

Hybrids that initially fail to produce tumours become tumorigenic again on cultivation in vitro. Fusion of SEWA with Rb7BnR/Rb7BnR fibroblasts gave nine hybrid clones (clones B, Bl, Cl, D, Dl, D3, F, J2and J3). These hybrids were assayed and gave Km values ranging from 0-8 to 2-9 mM (Tables 6 and 7); in some cases the reciprocal plots were curvilinear, as described for YACIR X CBAT6T6 clone 1G1. These clones were obviously heterogeneous, presumably due to varying degrees of segregation. Two of the clones (Cl and F), which initially gave high Km values, were grown continuously in vitro and their Km values determined at intervals. The results (Table 7) show that, on continued cultivation, the Km values fell progressively. Clone F, tested in vivo, initially produced no tumours, but later became tumorigenic. In addition, the morphology and cell volume of both clones were altered during this period of cultivation. Thus, these clones appear to be rapidly segregating populations and this segregation is associated with a fall in Km . Three tumours produced by inoculation of early passages of these clones were also assayed (C1T1, C1T2 and FT3). These tumours gave lower Km values than the cells inoculated, indicating a selection for lower Km in vivo (Table 6).


63

Table 8. Kinetic constants of hexose uptake in human fibroblasts

Cell type MRC5 S1814 DS1 GM1604 WI38

Passage no. P35 P6 P4 P27

Cell density (cells/cm2) 1-7X 4-2 x 4-7 x 5-0 x 5-3 x

4

10 104 104 104 104

No. of concentrations assayed 6

5 6 6 5

Km (mm)


3-210±0-182 2-769 ±0-169 3-886 ±0-146 2-577 ±0-237 2-358 ±0-205

552 •o± 17-9 334 •8± 14-9 283 •4± 8-1 169 •4± 14-5 159 •8± 120

Table 9. Kinetic constants of hexose uptake in human cancer cell lines

Cell type

No. Of Tumori- concentrations genicity assayed


Km (miu) 1-630 ±0-123 1-271 ±0-100 0-984 ±0-219

Vn»* (nmol/106 cell per h)

RT112/84 HeLa flow

+

6

+

5 6

1-6 xlO 5 1-1 x 1054 8-6 x 10

HeLa spinner

+

4

1-3 x 105

1-517 ±0-190

44-1 ± 2-0 105-1 ± 4-9 133-5 ± 13-5 83-2 ± 7-9

+

5

7-8 x 104

1-260 ±0-046

105-3 ± 2-4

+

6

7-4 x 104

1-285 ±0-033

+

6

4

+

6 6

7-6 xlO 1-1 XlO5 1-1 xlO 5

1-161 ±0-120 1-278 ±0-051 1-598 ±0-097

7-1 xlO 4 3-4 xlO 4

1-303 ±0-151 2-382 ±0-174

95-8 ± 1-8 332-2 ±20-0 114-7 ± 2-1 253-0± 8-1 37-9 ± 2-9

HeLa D98 F908A3 HeLa D98 AH2 H29/219 H.Ep.2 RPMI2650

ND*

6

HT55F

ND

5

HT115

ND

5

Daudi

-

5

5-8 xlO 4 0-600 ±0-140 1-4 xlO 5 1-447 ±0-162 Suspension culture 2-342 ±0-150

301-0± 15-4 16-2± 2-1 141 ± 11 43-3 ± 11

• ND, not determined.

Deoxyglucose uptake by malignant and non-malignant human cells Tables 8 and 9 give the kinetic constants of deoxyglucose uptake for five human fibroblast cell lines and ten human cell lines of malignant origin. Included in these tables are the parental cell lines for two series of hybrids that were examined. Figs 4 and 5 give examples of the Hanes plots of the human cell lines. As was the case with the mouse fibroblasts, there is significant variation in the kinetic constants for the different types of human fibroblast. The Km values obtained for the human fibroblasts range from 2-3 to 3-9 mM. These values are much higher than those obtained for the human tumorigenic cell lines, which range from 0-9 to 1-6 mM. The tumorigenicity


64

S

v

6

(/imol/10 cells per h)~ 24-

20-

16-

4-

-

5

-

4

-

3

-

2

-

1

0

1

2

3

4

5

S(ITIM)

Fig. 4. Reciprocal plot of 2-deoxy-D-glucose uptake in DSl human foetal fibroblasts. S V M (/.mol/106 cells per h)"1)

6050-

40-

30-

20-

10"

-

5

-

4

-

3

-

2

-

1

0 S(mM)

1

2

3

4

5

Fig. 5. Reciprocal plot of 2-deoxy-D-glucose uptake in the human carcinoma cell line HeLaD98AH2.


65

of the human cell lines was assessed by their ability to grow progressively in nude mice. Tumorigenicity in human cells is thus associated with a reduction in the Km for deoxyglucose uptake. Four different sublines of the cervical carcinoma cell line HeLa were assayed. These are similar both in Km and Vmax, indicating that the kinetic constants have remained largely unchanged since these sublines diverged from the original HeLa cell lines (Gey etal. 1952). Two of the cell lines in Table 9 have unexpectedly high Km values (Daudi and HT55F). Daudi has failed to give tumours when injected into nude mice, while HT55F has not yet been tested for growth in vivo. Deoxyglucose uptake by human carcinoma X human fibroblast hybrids Hybrids between 6-thioguanine-resistant human cervical carcinoma derivatives and human diploid fibroblasts have been obtained from two sources. Five hybrids between HeLa D98AH2 cells and GM1604 fibroblasts were obtained from E. J. Stanbridge (Der & Stanbridge, 1978, 1981) and four hybrids between HeLa D98F908A3 and S1814 fibroblasts were obtained from H. P. Klinger (Klinger, 1980). Both investigators have shown that in these crosses tumorigenicity is stably suppressed but that rare malignant segregants can be isolated. Tumorigenicity in these hybrids was tested by inoculation into nude mice (Stanbridge, 1976; Klinger, 1980). Table 10 gives the kinetic constants of deoxyglucose uptake for these hybrids. Comparison of the hybrids in which malignancy is suppressed with their tumorigenic derivatives shows that there is a reduction in Km associated with tumorigenicity. Thus, in these crosses, a low Km value is associated with tumorigenicity. Effect of viral transformation on deoxyglucose uptake Since many investigators have compared the hexose uptake of transformed cells in vitro with that of their non-transformed counterparts, it was of interest to compare such cells in the present assay. Two pairs of transformed and non-transformed cells were investigated: (a) the rat kidney cell line, NRKTG3, and its Rous sarcoma virus (RSV)-transformed counterpart, 3.B77.Sc4 (Marshall, 1980); (b) the mouse fibroblastic cell line, 3T3 A31 clone 9 and its simian virus 40 (SV40)-transformed derivative, SV-3T3-B. In addition, a cell line derived from a tumour produced in a nude mouse by SV-3T3-B cells was assayed (SV-3T3-BT). These results are presented in Table 11. The following conclusions can be drawn: (1) transformation of NRKTG3 by RSV results in a 2-7-fold increase in V ^ with no significant alteration in the/Cm; (2) transformation of 3T3 by SV40 results in a two-fold increase in VmiX with no significant alteration in Km; (3) the tumour cell line, SV-3T3-BT, has a significantly lower Km (and Vmax) than the SV3T3-B cell line from which it was derived. These data support the observations of other workers (Isselbacher, 1972; Plagemann, 1973; Weber, 1973; Kletzien & Perdue, 19746, 1975; Royer-Pokora et al. 1978) that viral transformation by both RNA and DNA tumour viruses is

4-4-4 541E 541M CG04 CGL3

1Acn2 lAcnlTG 2BlColl 5Amc3

S1814 X HeLa D98 F908A3

Hybrid

GM1604 x HeLa D98 AH2

Hybrid parents

+ +

-

+ + +

-

-

Tumorigenicity

6.6 X 8.8 x 6.9 x 1.0 X

8.9 x 7.7 X 7.5 X 7.8 x 7.7 x 10' lo4 lo4 105

lo4 10' 10' lo4 lo4


5 5 5 6

6 6 6 6 6


(m~)

+

2.047 f 0.101 0.899 f 0.061 3.484 0.241 1.451 f 0.070

1.882 f 0.120 4.1 13 f 0.208 1.491 k 0.049 1.175 f 0.078 1.228 f 0.079

K m

Table 10. Gnetic constants of hexose uptake in human carcinoma xfibroblast hybnbIs

v4.3 9.3 3.9 6.4 5.2 335.4 f 11.5 116.5 f 3.3 265.5 f 15.1 124.2 k 4.3

86.5 f 217.4 f 183.1 f 158.5 f 173.8 f

(nmol/106 cells Per h)

h.

3

0

2

3

4

0 3

CC

S %

#

b

?

9 -' %

5;

is


67

Table 11. Effect of viral transformation on the kinetic constants of hexose uptake

Cell type NRK TG3 3.B77.Sc4

3T3A31clone9 SV-3T3-B SV-3T3-BT

Transformed morphology + + +

Cell density (cells/cm2) 7-7 x 1-3 x l-7x 2-6 x

104 105 105 105

3-3 xlO 4


6 6 5 S 6

Km (min)

V™ (nmol/10 6 cells per h)

2-056 ±0-266 1-67010-080 1-691 ±0-092 1-966 ±0-092 1-334 ±0082

120-318-8 323-818-6 122-6 ±2-8 248-2 ±9-0 101-8±4-2

associated with an increase in Vmax for hexose uptake, but with no alteration in Km . These findings emphasize the now well-documented difference between transformation in vitro and tumorigenicity (Boone, 1975; Stiles, Desmond Jr, Sato & Saier, 1975; Gee & Harris, 1979; Straus e* al. 1977; Stanbridge & Wilkinson, 1978). Uptake of 3-O-methyl-D-glucose The uptake of 2-deoxy-D-glucose is determined by a coupled reaction in which the transport of the hexose is linked to its subsequent phosphorylation. In any given situation, the rate of uptake may be influenced predominantly by either the transport step or the phosphorylation reaction (Waley, 1963; Wohlhueter & Plagemann, 1980; Pasternak et al. 1982). The purpose of the experiments described in this section was to measure transport independently of the phosphorylation reaction. This can be done with the analogue 3-O-methyl-D-glucose, which is transported into cells but is not metabolized further (Renner et al. 1972; Weber, 1973). Uptake of 3-O-methyl-Dglucose by most of the cell types investigated was linear for 2 min at 20 °C. Time points measured at 3 min or longer deviated from linearity. Over the linear portion of the uptake curves, reciprocal plots obeyed Michaelis—Menten kinetics. The kinetic constants of the 12 cell types investigated are given in Table 12. Vman values range from 107 to 413 nmol/106 cells per h and show no correlation with tumorigenicity. There is, however, a clear relationship betweenKm and tumorigenicity. Malignant cell types have/Cm values between 1-7 and 3-8 min while non-malignant cell types range from 4-4 to 8-5 miw. The hybrid in which malignancy is suppressed, clone 1G8, has aKm that is two-fold higher than the malignant segregant clone 1G1T1 and five-fold higher than the malignant segregant clone 1G1T2 (Fig. 6). The suppressed hybrid, lAcn2, has a Km that is three-fold higher than the malignant segregant lAcnlTG, and the suppressed hybrid, 2BlColl, has zKm that is two-fold higher than the malignant segregant 5Amc3. The non-malignant WGA-resistant melanoma derivative PG19WGARC2 has zKm that is 3-5-fold higher than its malignant parent PG19. These data are therefore in accord with those obtained with 2-deoxy-D-glucose. Since 3-O-methyl-D-glucose is not metabolized, the reduction in A'm seen in malignant cells must be due predominantly to an alteration in the transport of hexose across the cell membrane.

1.3 x 10' 1.2 x 105 1.6 X 105 1.3 X 10'

-

+ -

+

S1814 x HeLa D98 1Acn2

S1814 x HeLa D98 lAcnlTG

S1814 X HeLa D98 2BlColl

S1814 X HeLa D98 5Amc3

1.5 X 105 1-9 x 1@ 7.4 x 10'

+ +

-

YACIR x CBAT6T6 fibroblasts: Clone 1G8 Clone l G l T l Clone 1 G l T 2

6

5

4

5

6

1.2 x 105 1.9 x 105

+

5 4 5 5


105 105 10' 10'

-

1-7 X 1.4 X 6.6 X 9.6 x

A/Sn fibroblasts W

-

+


A9HT

CBAT6T6 fibroblasts: P3 P4

PG19

Cell type

Tumorigenicity

+

(m)

3.500 f 0.554

6.301 f 1.170

2.771 f 0.241

3.674 f 0.228

6.564 f 1.000

+

2.114 0.172 2.221 f 0.460 4.419 f 0.520 4.868 1.214

K m

Table 12. Kinetic constants of 3- 0-methylglucose uptake

v-

155.8 f 17.8

206.0 f 25.5

107.6 f 7.3

202.1 f 8.2

300.4 f 39.7

+

1 0 6 - 6 f 5.1 116.6 f 18.2 325.1 f 26.8 420.3 90.8

(nmol/ lo6 cells per h)

i. 4 '

a

m

t?

is

s!

is a


69

(rriM (^mol/10 6 cells per h)" 1

1G8

1G1T2

9

-8

-7

-6

-5

-4

-3

-2

-1

Fig. 6. Reciprocal plots of 3-0-methyl-D-glucose uptake in the hybrids YACIR X CBAT6T6 fibroblast clones 1G8 and 1G1T2.

MSWBS (mM (pmol/nl per min)

Daudi

-5

-4

-3

-2

-1

Fig. 7. Reciprocal plots of D-glucose uptake in the cell lines MSWBS, Daudi and SEWA.

70


The Km for 3-O-methyl-D-glucose in any one cell type is higher than the Km for 2-deoxy-D-glucose. This has been reported by other investigators in other cell types (Renneretal. 1972; Weber, 1973; Kletzien & Perdue, 1974a,b,c; Graff, Wohlhueter & Plagemann, 1978). The presence of substituents on the 3' position oxygen of glucose has been shown to decrease its affinity for the erythrocyte glucose transporter (Barnett, Holman & Munday, 1973). This might also account for the higher Km for 3-O-methyl-D-glucose seen in cells in culture. The Vmax values for the two hexoses, 2-deoxy-D-glucose and 3-O-methyl-D-glucose, lie in the same range and show some degree of correlation in any one cell type. Indeed, parallel cultures of PG19 assayed for the uptake of both hexoses on the same day gave a V ^ value for 2-deoxy-D-glucose of 132-5 ± 10-8 (Table 1) and a VnViX value for 3-O-methyl-D-glucose of 116-6 ± 18-2 (Table 12). These values are not significantly different. Hexokinase activity If, as our results show, the kinetics of hexose transport in the cells we have studied are determined predominantly by the transport component of the coupled reaction and not the phosphorylation component, one would expect that the rate at which the cell is able to phosphorylate the transported sugar would substantially exceed the rate at which the sugar is transported across the cell membrane. In this section we report measurements we have made on the hexokinase activities of homogenates of most of the cell types in which we have studied 2-deoxy-D-glucose uptake (Kletzien & Perdue, 1974a). These activities are expressed as V ^ values of the enzyme per 106 cells and Vmax values per milligram of protein. Since there is some variation between different cell types in the amount of protein per cell, the two sets of values are not interchangeable. Table 13. Hexokinase activities of malignant and non-malignant mouse cells Cell type

Hexokinase activity (nmol/10 6 cells per h) (/imol/mg protein per h) 954 1302

516

YACIR

470

2-57

SEWA

2052

A9HT

1518 1080

11-24 5-18 4-72

TA3HaB

497

5-48

PG19-G

2108

7-44

CBAT6T6 fibroblasts P2

1631

3-29

A/Sn fibroblasts P6

1374

2-65

A/Sw fibroblasts P2

882

3-41

1096

4-39

960

5-93

PG19

A/Sn X BALB/c fibroblasts P7 PG19WGAR clone C2

5-04


71

Mouse cells. The Vmax for hexokinase in mouse tumour cells and mouse fibroblasts are given in Table 13. (Where replicate values for a cell type are given, determinations were done on samples derived from different cell cultures.) Comparison of these values with the Vmax values for deoxyglucose uptake (Tables 1 and 2) shows that the VWx for hexokinase is between 5-5 and 42-fold higher than for deoxyglucose uptake. (TA3HaB is exceptional in having a hexokinase Vmax only 1-7-fold higher.) Furthermore, there is no correlation between hexokinase activity and tumorigenicity. Values for malignant cells vary between 470 and 2108nmol/l0 6 cells per h while values for normal cells vary between 882 and 1631 nmol/106 cells per h. The non-malignant WGA-resistant PG19 derivative, PG19WGAR clone C2, has a similar hexokinase activity to its malignant parent, PG19. Hybrid mouse cells and lectin-resistant derivatives of malignant mouse cell lines. Table 14 gives the hexokinase Vmax values for some of the hybrid mouse cells and the lectin-resistant mouse cell lines in which 2-deoxy-D-glucose uptake has been studied. These values are much higher than the respective Vmax values for 2-deoxyD-glucose uptake. There is no correlation between hexokinase activity and tumorigenicity. Human cells. The hexokinase activity of five tumorigenic human cell lines and three types of non-tumorigenic human fibroblast were assayed (Table 15). There is no correlation between hexokinase activity and tumorigenicity. The hexokinase Vmax in the human cell types is 2 to 26-fold higher than the Vmax for deoxyglucose uptake in the same cells (Table 9). Human carcinoma X fibroblast hybrids. As described above, the kinetics of 2-deoxy-D-glucose uptake have been investigated in nine hybrids between the HeLa Table 14. Hexokinase activities of hybrid mouse cells and lectin-resistant mouse cell derivatives V™,

Tumorigenicity

(nmol/10 6 cells per h)

—

+ + + + + +

966 1206 1356 1854 1386 2106 2238

2-81 613 3-67 614 2-63 1-84 3-56

A9HT x C57B1 lymphocyte Clone 3 Clone 4 Clone 2T1

— +

858 1584 1080

4-76 3-22 5-20

A9HTWC

-

528

3-02

A9HTWD

+

1296

5-45

Cell type YACIR X CBA6T6 fibroblasts Clone 1G8 Clone 1G1T1 Clone 1G1T2 Clone 1G1T3 Clone lG8a Clone lG8b Clone lG8c

(jUmol/mg protein per h)

72

M. K. White, M. E. Bramivell and H. Harris Table 15. Hexokinase activities of human cells

Cell type

Tumorigenicity

H.Ep.2 H29/219 RT112/84 HeLa flow HeLaD98AH2 HT115 DS1 P5 MRC5 P36 GM1604

Hexokinase activity (/imol/mg protein per h) (nmol/106 cells per h)

+ + + + + ND* -

616 756 990 1302 1020 396 678 2490 786

2-24 1-98 4-46 3-58 2-93 1-27 3-74 5-03 1-31

• Not determined.

Table 16. Hexokinase activities of human carcinoma X fibroblast hybrids

Hybrid parents GM1604xHeLaD98AH2

S1814xHeLaD98 F908 A3

Hybrid 4-4-4 541E 541M CG04 CGL3 lAcn2 lAcnlTG 2BlColl 5Amc3

Tumorigenicity

+ + + + +

Hexokinase activity (jimol/mg (nmol/106 cells per h) protein per h)

1032 768 708 936 960 529 599 840 786

2-31 1-60 3-15 3-04 2-90 1-26 1-71 1-96 2-79

carcinoma cell line and human diploid fibroblasts (Table 10). The hexokinase activity of each of these hybrids was measured and is given in Table 16. Again, the hexokinase Vma-x value of each cell type was larger than that for the uptake of 2-deoxy-D-glucose. There is an apparent association between tumorigenicity and hexokinase activity in this set of cells when hexokinase activity is expressed in nmol/mg protein per h, but not when it is expressed in nmol/106 cells per h. This disparity arises because the suppressed hybrids have a higher protein content per cell (0-42-0-48 mg protein/ 106 cells) than the malignant segregants (0-22-0-35 mg protein/106 cells). The higher protein content of the hybrids in which malignancy was suppressed was peculiar to this set of hybrids. While hexokinase measurements in homogenates may not reflect hexokinase activities in the intact cell, it is, in any case, clear that hexokinase activity as measured does not correlate with tumorigenicity and that, in general, the amount of hexokinase available greatly exceeds what would be required to phosphorylate the hexose transported across the cell membrane.

+

+ -

PG19

MSWBS

Daudi

C57BI fibroblasts P6

A9HT

SEWA

YACIR

Tumorigenicity

+ + +

Cell type

1.950 f 0.340 2.332 f 0.237

+ 0.683 + 0.083

0.490 f 0.026 1.250 0.380

0.756 f 0.127

2.48 f 0.50 14.80 f 1.34

0.84 k 0.05

0.63 f 0.04 2.84 f 0.60

1.12 f 0.18

6.06 f 0.37

+ 0.096

1.030

17.90 f 2.23

(pmol/nl per min)

1.060 f 0.205

Km ( m ~ )

v-

7.4

4.5 96.0 f 13.4 1084.4f 98-2

78.3 f

49.1+ 3.1 212.5 f 46.4

70.6f

6.6

+ 110.9

109.1 f

890.5

(nmol/l@ cells per h)

Table 17. ki'netic conslants of glucose uptake by dtfferent cell t-ypes

4 4

5

4 4

5

5


B 2

a

3-

s3.

.a

m

8

22

74


Uptake of D-glucose D-Glucose is the physiological substrate of the hexose transport system of animal cells and therefore the ideal compound for the measurement of uptake rates. However, it is rapidly metabolized to compounds such as lactate and carbon dioxide, which may leave the cell, so that measured uptake rates may be severe underestimates of the true rates (see Plagemann & Richey, 1974). However, this problem can be in large part circumvented by the use of very short incubation times (less than 30 s). This can be achieved by the silicone oil filtration centrifugation technique of Werdan et al. (1980). This method was used to measure the kinetics of D-glucose uptake in several cell types. In most experiments, linear uptake was maintained for up to 30 s. The uptake rates at different D-glucose concentrations were used to determine the Km and Vmax values, as described above for 2-deoxy-D-glucose and 3-O-methyl-D-glucose uptake. The results are presented in Table 17. Comparison of the Km values obtained for D-glucose with those for 2-deoxy-Dglucose shows that there is good agreement. The tumorigenic cell types have Km values in the range 0-49 to 1-25 ITIM while the non-tumorigenic cells have Km values in the region of 2 mM. Werdan et al. (1980) obtained a Km of 2mM for human diploid fibroblasts by this method. The Vmax values obtained by this method are of the same order as those obtained with 2-deoxy-D-glucose. It thus appears that, in our assay, D-glucose and 2-deoxy-D-glucose behave similarly. Indeed, with YACIR cells, the time course of uptake of 2-deoxy-D-[l-14C]glucose was found to be identical to that obtained with D-[U-14C]glucose at the same concentration (data not shown). Furthermore, since 2-deoxy-D-glucose is not metabolized beyond the phosphorylation step, these data indicate that metabolic conversion does not cause D-glucose uptake rates to be underestimated in the present assay. This conclusion is reinforced by experiments in which glycolysis was inhibited with sodium fluoride. YACIR cells, preincubated with fluoride under conditions in which glycolysis was inhibited by 90%, gave uptake time courses identical to those given by untreated cells. The silicone oil filtration centrifugation method was also used to measure the uptake of L-[l-14C]glucose, which is not a substrate for the hexose transport system and is thus only able to enter cells by the much slower process of simple diffusion. The uptake of this sugar was so slow that incubations of 1—3 h were necessary to achieve significant uptake. Table 18 shows the results obtained and, for comparison, the rate Table 18. Relative rates of uptake of L-glucose and D-glucose Cell type

Rate of L-glucose uptake (fmol/nl per min)

Rate of D-glucose uptake (fmol/nl per min)

SEWA PG19 MSWBS A9HT Daudi YACIR

0-73 ±0-18 1-93 ±0-18 0-43 ±0-64 0-46 ±0-06 0-20 + 0-05 0-20 ±0-05

1600 ±62 220 ±24 110 ± 14 124 ± 9 123 ± 7 546 ±44


75

-

of D-glucose uptake at the same concentration (0 1 mM) with the same cell suspension. It is clear that L-glucose enters cells at a rate that is less than 1 % of that of D-glucose. Simple diffusion does not therefore contribute significantly to hexose uptake. Furthermore, these data demonstrate the stereospecificity of the glucose transporter of the cell membrane and the suitability of L-glucose as a control for trapping of extracellular fluid.

DISCUSSION

In the present study the kinetics of hexose uptake were studied in 86 different cell lines, and, without exception, the ability of cells to grow progressively in vivo was linked to a reduction in the Km for hexose transport. This change in Km reflects a change in the transport system itself and is not determined by the subsequent metabolism of the hexose. Previous studies have shown a similar inverse correlation between tumorigenicity and the Km of glucose uptake, but these studies did not dissociate transport from metabolism (Burk, Woods & Hunter, 1967; Hatanaka, 1971). How might the alteration in Km that we have described be brought about? One possibility is that the hexose transport protein has undergone mutation in a way that increases its affinity for hexose. Whitfield et al. (1982) have isolated a mutant from CHO cells that is resistant to the cytotoxic effects of 3-0-methyl-D-glucose and has a twofold lower Km for hexose transport. This result demonstrates that a lower Km can be generated by a mutational event, but it remains unlikely that simple mutations in the hexose transport protein would generate a wide range of molecules with different degrees of increased affinity for hexose, as evidenced by the wide range of malignant cells examined. Another possibility is that there might be polymorphic forms of the hexose carrier protein with different Km values for hexose transport. Malignancy would then be associated with an increased expression of low/Cm variants of the hexose transporter. A shift towards low Km isoenzymes has been observed for several glycolytic enzymes in transformed cells (G. Weber, 1974). Changes in Km values might also be produced by secondary changes in the hexose transporter protein, for example, by glycosylation (Bramwell & Harris, 1978a; Warren, Buck & Tuszynski, 1978) or phosphorylation. And finally, changes in the hexose transport Km might be produced by more general modifications of the cell membrane, which might affect not only the transport of hexose but also of other molecules. Such modifications might, for example, be brought about by alterations in the lipid composition of the membrane. Itaya, Hakamori & Klein (1976) have described changes in membrane lipid composition associated with malignancy in somatic cell hybrids; Yuli, Wilbrandt & Shinitzky (1981) have shown that the hexose transport system is sensitive to the lipid composition of the cell membrane; Pilch, Thompson & Czech (1980) have produced evidence that, in the rat adipocyte, changes in the phospholipid microenvironment of the glucose transporter might be involved in its response to insulin. Of these various possibilities we are most attracted by the notion that changes in the glycosylation of cell membrane proteins might induce changes in the Km of the hexose transport process. Numerous authors have shown that the pattern of glycosylation of

76


membrane proteins is altered in transformed cells (see, for example, Warren et al. 1978), and Atkinson & Bramwell (1980a,6) have shown that cell surface sialic acid content and sialyltransferase activity co-segregate with malignancy in hybrids between malignant and non-malignant cells. A systematic change in the glycosylation of a membrane protein that appeared to be involved in some way in glucose transport has also been shown to co-segregate with malignancy in hybrid cells (Bramwell & Harris, 1978a,6; Bramwell, 1980; Bramwell & Atkinson, 1982). Some preliminary experiments that we have done with tunicamycin indicate that this antibiotic, in concentrations that largely inhibit the dolichol phosphate-mediated glycosylation of membrane proteins, produces changes in the Km for hexose transport. We have now to consider what relevance the reduction in Km for hexose transport might have to the biology of cancer cells. Since a reduction in Km entails that the difference between malignant and non-malignant cells becomes more pronounced at lower external hexose concentrations, it is not difficult to see how this might confer a selective advantage on malignant cells in vivo. The cell density in a primary tumour nodule is high and the blood supply precarious, so that the availability of hexose could easily be limiting. This has indeed been shown to be the case for some ascitic tumours (Del Monte & Rossi, 1963). A decrease in the A'm of the hexose transport system might thus make the malignant cell a more effective scavenger of whatever hexose is available than the normal cells with which it must compete. This does not necessarily mean that the change in hexose transport is the primary determinant of the malignant state, although some authors have argued that an increase in the activity of one or more transport proteins might control cell proliferation (Holley, 1972; Bhargava, 1977). We thank Dr Harold P. Klinger, Department of Genetics, Albert Einstein College of Medicine, Bronx, N.Y., U.S.A. and Dr Eric J. Stanbridge, Department of Microbiology, College of Medicine, University of California, Irvine, Calif., U.S.A. for the gift of human hybrid cell lines; Professor J. F. Watkins, Department of Medical Microbiology, The Welsh National School of Medicine, Heath Park, Cardiff; Dr L. M. Franks, Imperial Cancer Research Fund Laboratories, Lincoln's Inn Fields, London and Dr C. J. Marshall, Chester Beatty Research Institute, Institute of Cancer Research, Royal Cancer Hospital, Fulham Road, London for the gift of cell lines; Dr S. J. Goss for his advice on the statistical aspects of the work; and Mrs Ruth Hennion and Mr Stephen Greig for skilful technical assistance. The work was supported by the Cancer Research Campaign, of which M.E.B. is the James Hanson Fellow. M.K.W. was in receipt of a Medical Research Council Scholarship for Training in Research Methods.

REFERENCES ANTONIADES, H. N., STATHAKOS, D. & SCHER, C. D. (1975). Isolation of a cationic polypeptide

from human serum that stimulates proliferation of 3T3 cells. Proc. natn. Acad. Sci. U.SA. 72, 2635-2639. ATKINSON, M. A. L. & BRAMWELL, M. E. (1980a). Studies on the surface properties of hybrid cells. I. Sialyl-transferase activity in homogenates of malignant and non-malignant cells. J. Cell Sci. 46, 187-201. ATKINSON, M. A. L. & BRAMWELL, M. E. (19806). Studies on the surface properties of hybrid cells. II. Sialyl-transferase on the surface of malignant and non-malignant cells. J. Cell Sci. 46, 203-220. ATKINSON, M. A. L. & BRAMWELL, M. E. (1981). Studies on the surface properties of hybrid cells. III. A membrane glycoprotein found on the surface of a wide range of malignant cells. J. Cell Sci. 48, 147-170.


11

BADER, J. P. (1972). Temperature-dependent transformation of cells infected with a mutant of Bryan Rous sarcoma virus. J. Virol. 10, 267-276. BARNETT, J. E. G., HOLMAN, G. D. & MUNDAY, K. A. (1973). Structural requirements for

binding to the sugar-transport system of the human erythrocyte. Biochem.J. 131, 211-221. BHARGAVA, P. M. (1977). Regulation of cell division and malignant transformation: a new model for control by uptake of nutrients. J. theor. Biol. 68, 101-137. BOONE, C. W. (1975). Malignant hemangioendotheliomas produced by subcutaneous inoculation of Balb/3T3 cells attached to glass beads. Science, N.Y. 188, 68-70. BRADLEY, W. E. C. & CULP, L. A. (1974). Stimulation of 2-deoxyglucose uptake in growthinhibited BALB/c 3T3 and revertant SV40-transformed 3T3 cells. Expl Cell Res. 84, 335-350. BRAMWELL, M. E. (1980). A glycoprotein abnormality associated with malignancy. Biochem. Soc. Trans. 8, 697-698. BRAMWELL, M. E. & ATKINSON, M. A. L. (1982). The apparent inducibility of tumour marker glycoproteins in a melanoma cell line selected for growth in low levels of glucose. .7. Cell Sci. 54, 241-254. BRAMWELL, M. E. & HARRIS, H. (1978a). An abnormal membrane glycoprotein associated with malignancy in a wide range of different tumours. Proc. R. Soc. Lond. B 201, 87-106. BRAMWELL, M. E. & HARRIS, H. (19786). Some further information about the abnormal membrane glycoprotein associated with malignancy. Proc. R. Soc. Lond. B 203, 93-99. BURK, D., WOODS, M. & HUNTER, J. (1967). On the significance of glucolysis for cancer growth with special reference to Morris rat hepatomas. J. natn. Cancer Inst. 38, 839-863. CHAN, W.-S. (1980). Tumorigenicity of malignant cells selected for resistance to wheat germ agglutinin. Thesis, University of Oxford. CORNISH-BOWDEN, A. (1976). Principles of Enzyme Kinetics. London: Butterworths. CORNISH-BOWDEN, A., PORTER, W. R. & TRAGER, W. F. (1978). Evaluation of distribution-free

confidence limits for enzyme kinetic parameters. J. theor. Biol. 74, 163-175. DEL MONTE, U. & Rossi, C. B. (1963). Glucose supply by the living host and glycolysis of Yoshida ascites hepatoma in vivo. Cancer Res. 23, 363-367. DER, C. J. & STANBRIDGE, E. J. (1978). Lack of correlation between the decreased expression of cell surface LETS protein and tumorigenicity in human cell hybrids. Cell 15, 1241-1251. DER, C. J. & STANBRIDGE, E. J. (1981). A tumor-specific membrane phosphoprotein marker in human cell hybrids. Cell 26, 429-438. EVERSE, J. (1975). Enzymatic determination of lactic acid. Meth. Enzym. 41, 41-44. FENYO, E. M., KLEIN, E., KLEIN, G. & SWIECH, K. (1968). Selection of an immuno-resistant

Moloney lymphoma subline with decreased concentration of tumor-specific surface antigens. J.

natn. Cancer Inst. 40, 69-80. GAZDAR, A., HATANAKA, M., HERBERMAN, R., RUSSELL, E. & IKAWA, Y. (1972). Effect of

dibutyryl cyclic adenosine phosphate plus theophylline on murine sarcoma virus transformed non-producer cells. Proc. Soc. exp. Biol. Med. 141, 1044-1050. GEE, C. J. & HARRIS, H. (1979). Tumorigenicity of cells transformed by simian virus 40 and of hybrids between such cells and normal diploid cells. J. Cell Sci. 36, 223-240. GEY, G. O., COFFMAN, W. D. &KUBICEK, M . T . (1952). Tissue culture studies of the proliferative capacity of cervical carcinoma and normal epithelium. Cancer Res. 12, 264—265. GINGRICH, R. D., WOUTERS, M., BRAMWELL, M. E. & HARRIS, H. (1981a). Immunological

delineation in normal and malignant cells of a membrane protein involved in glucose transport. I. Preparation and properties of the antibody. J. Cell Sci. 52, 99-120. GINGRICH, R. D., WOUTERS, M., BRAMWELL, M. E. & HARRIS, H. (19816). Immunological

delineation in normal and malignant cells of a membrane protein involved in glucose transport. II. Function of the antigen. J. Cell Sci. 52, 121-135. GRAFF, J. C , WOHLHUETER, R. M. & PLAGEMANN, P. G. W. (1978). Deoxyglucose and 3-0-

methylglucose transport in untreated and ATP-depleted Novikoff rat hepatoma cells. Analysis by a rapid kinetic technique, relationship to phosphorylation and effects of inhibitors. J . cell. Physiol. 96, 171-188. HANES, C. S. (1932). Studies on plant amylases. I. The effect of starch concentration upon the velocity of hydrolysis by the amylase of germinated barley. Biochem. J. 26, 1406-1421. HARRIS, H. (1971). Cell fusion and the analysis of malignancy. The Croonian Lecture. Proc. R. Soc. Lond. B 179, 1-20.

78


HATANAKA, M. (1971). Biochemical markers of viral oncogenesis with special reference to glucose transport as a functional expression of sarcoma genes. Gann Monograph 10, 45-73. HATANAKA, M., AUGL, C. & GILDEN, R. V. (1970). Evidence fora functional change in the plasma membrane of murine sarcoma virus-infected mouse embryo cells. J. biol. Chem. 245, 714—717. HATANAKA, M., GILDEN, R. V. & KELLOFF, G. (1971). Induction of sugar uptake by a hamster pseudotype sarcoma virus. Virology 43, 734—736. HATANAKA, M. & HANAFUSA, H. (1970). Analysis of a functional change in membrane in the process of cell transformation by Rous sarcoma virus; alteration in the characteristics of sugar transport. Virology 41, 647-652. HATANAKA, M., HUEBNER, R. J. & GILDEN, R. V. (1969). Alterations in the characteristics of

sugar uptake by mouse cells transformed by murine sarcoma viruses. J . natn. Cancer Inst. 43, 1091-1096. HATANAKA, M., TODARO, G. J. & GILDEN, R. V. (1970). Altered glucose transport kinetics in murine sarcoma virus-transformed BALB/3T3 clones. lnt.J. Cancer 5, 224—228. HAUSCHKA, T. S. (1953). Cell population studies on mouse ascites tumours. Trans. N.Y.Acad. Sci. 16, 64-73. HAYFLICK, L. &MOORHEAD, P. S. (1961). The serial cultivation of human diploid cell strains. Expl Cell Res. 25, 585-621. HINO, S. & YAMAMOTO, T. (1971). Expression of the sarcoma genome in a Rous mouse tumor cell line, SR-C3H-2127. Gann 62, 539-544. HOLLEY, R. W. (1972). A unifying hypothesis concerning the nature of malignant growth. Proc.

natn. head. Sci. U.SA. 69, 2840-2841. ISSELBACHER, K. J. (1972). Increased uptake of amino acids and 2-deoxy-D-glucose by virustransformed cells in culture. Proc. natn. Acad. Sci. U.SA. 69, 585-589. ITAYA, K., HAKOMORI, S. I. & KLEIN, G. (1976). Long-chain neutral glycolipids and gangliosides of murine fibroblast lines and their low- and high-tumorigenic hybrids. Proc. natn. Acad. Sci. U.SA. 73, 1568-1571. JACOBS, J. P., JONES, C. M. & BAILLE, J. P. (1970). Characteristics of a human diploid cell designated MRC5. Nature, Land. 227, 168-170. JONASSON, J., POVEY, S. & HARRIS, H. (1977). The analysis of malignancy by cell fusion. VII. Cytogenetic analysis of hybrids between malignant and diploid cells and of tumours derived from them. J. Cell Sci. 24, 217-254. KLEIN, E., KLEIN, G., NADKARNI, J. S., NADKARNI, J. J., WIGZELL, H. & CLIFFORD, P. (1967).

Surface IgM specificity on cells derived from a Burkitt's lymphoma. Lancet 2, 1068-1070. KLEIN, G., BREGULA, U., WIENER, F. & HARRIS, H. (1971). The analysis of malignancy by cell

fusion. I. Hybrids between tumour cells and L-cell derivatives. J. Cell Sci. 8, 659-672. KLEIN, G. & KLEIN, E. (1958). Histocompatibility changes in tumours. J . cell. comp. Plivsiol. 52, 125-168. KLETZIEN, R. F. & PERDUE, J. F. (1974a). Sugar transport in chick embryo fibroblasts. I. A functional change in the plasma membrane associated with the rate of cell growth. J'. biol. Chem. 249, 3366-3374. KLETZIEN, R. F. & PERDUE, J. F. (19746). Sugar transport in chick embryo fibroblasts. II. Alterations in transport following transformation by a temperature-sensitive mutant of the Rous sarcoma virus. J . biol. Chem. 249, 3375-3382. KLETZIEN, R. F. & PERDUE, J. F. (1974c). Sugar transport in chick embryo fibroblasts. III. Evidence for post-transcriptional and post-translational regulation of transport following scrum addition. J. biol. Chem. 249, 3383-3387. KLETZIEN, R. F. & PERDUE, J. F. (1975). Regulation of sugar transport in chick embryo fibroblasts infected with a temperature-sensitive mutant of RSV. Cell 6, 513-520. KLINGER, H. P. (1980). Suppression of tumongenicity in somatic cell hybrids. I. Suppression and re-expression of tumorigenicity in diploid human X D98AH2 hybrids and independent segregation of tumorigenicity from other cell phenotypes. Cytogenet. Cell Genet. 27, 254—266. KOHLER, G., HOWE, S. C. & MILSTEIN, C. (1976). Fusion between immunoglobulin-secretingand

non-secreting myeloma cell lines. Eur.J. Immun. 6, 292-295. LOWRY, O. H., ROSEBROUGH, N. J., FARR, A. L. & RANDALL, R. J. (1951). Protein measurement

with the folin phenol reagent. J . biol. Chem. 193, 265-275. MARSHALL, C. J. (1980). Suppression of the transformed phenotype with retention of the viral 'sre'


79

gene in cell hybrids between Rous sarcoma virus-transformed rat cells and untransformed mouse cells. Expl Cell Res. 127, 373-384. MARSHALL, C. J., FRANKS, L. M. & CARBONELL, A. W. (1977). Markers of neoplastic transforma-

tion in epithelial cell lines derived from human carcinomas. J. natn. Cancer Inst. 58, 1743-1751. MAY, J. T., SOMERS, K. D. & K I T , S. (1974). Temperature-dependent alterations in 2-deo.\yglucose uptake in rat cells transformed by a cold-sensitive murine sarcoma virus mutant.

Int. J. Cancer 11, 377-384. MOORE, A. E., SABACHEWSKY, L. & TOOLAN, H. W. (1955). Culture characteristics of four

permanent lines of human cancer cells. Cancer Res. 15, 598-602. MOORE, G. E. & SANDBERG, A. A. (1964). Studies of a human tumour cell line with a diploid karyotype. Cancer 17, 170-175. PASTERNAK, C. A., MICKLEM, K. J., ELLIOTT, K. R. F., PLAGEMANN, P. G. W. & WOHLHUETER,

R. M. (1982). Hexose transport in hybrids between malignant and normal cells. Nature, Land. 296, 588. PILCH, P. F., THOMPSON, P. A. & CZECH, M. P. (1980). Co-ordinate modulation of D-glucose

transport activity and bilayer fluidity in plasma membranes derived from control and insulintreated adipocytes. Proc. natn. Acad. Sri. U.SA. 77, 915-918. PLAGEMANN, P. G. W. (1973). Deoxyglucose transport by uninfected, murine sarcoma virustransformed and murine leukemia virus-infected mouse cells. J . cell. Physiol. 82, 421-434. PLAGEMANN, P. G. W. & RICHEY, D. P. (1974). Transport of nucleosides, nucleic acid bases, choline and glucose by animal cells in culture. Biochim. biophys. Ada 344, 263-305. PORTER, W. R. & TRACER, VV. F. (1977). Improved non-parametric statistical methods for the estimation of Michaelis-Menten kinetic parameters by the direct linear plot. Biochem. J. 161, 293-302. RENNER, E. D., PLAGEMANN, P. G. W. & BERNLOHR, R. W. (1972). Permeation of glucose by

simple and facilitated diffusion by Novikoff rat hepatoma cells in suspension culture and its relationship to glucose metabolism, jf. biol. Chetn. 247, 5765-5776. ROYER-POKORA, B., BEUG, H.,

CLAVIEZ, M., WINKHARDT, H.-J., FRIIS, R. R. & GRAF,

T.

(1978). Transformation parameters in chicken fibroblasts transformed by AEV and MC29 avian leukemia virus. Cell 13, 751-760. SACER, R. & KOVAC, P. E. (1978). Genetic analysis of tumorigenesis: I. Expression of tumorforming ability in hamster hybrid cell lines. Somat. Cell Genet. 4, 375-392. SALTER, D. W., BALDWIN, S. A., LIENHARD, G. E. & WEBER, M. J. (1982). Proteins antigenically

related to the human erythrocyte glucose transporter in normal and Rous sarcoma virustransformed chicken embryo fibroblasts. Proc. natn. Acad. Sri. U.SA. 79, 1540-1544. SALTER, D. W. & WEBER, M. J. (1979). Glucose-specific cytochalasin B binding is increased in chicken embryo fibroblasts transformed by Rous sarcoma virus, jf. biol. Chem. 254, 3554-3561. SJOGREN, H. O., HELLSTROM, I. & KLEIN, G. (1961). Transplantation of polyoma virus-induced

tumors in mice. Cancer Res. 21, 329-337. STANBRIDGE, E. J. (1976). Suppression of malignancy in human cells. Nature, Land. 260, 17-20. STANBRIDGE, E. J. & WILKINSON, J. (1978). Analysis of malignancy in human cells: malignancy and transformed phenotypesare under separate genetic control. Proc. natn. Acad. Sri. U.SA. 75, 1466-1469. STEINBERG, B. M. & POLLACK, R. (1979). Anchorage independence: analysis of factors affecting growth and colony formation of wild-type and dl 54/59 mutant SV40-transformed lines. Virology 99,302-311. STILES, C. D., DESMOND, W. JR, SATO, G. & SAIER, M. H. (1975). Failure of human cells

transformed by simian virus 40 to form tumours in athymic nude mice. Proc. natn. Acad. Sci. U.SA. 72, 4971-4975. STRAUS, D. S., JONASSON, J. & HARRIS, H. (1977). Growth in vitro of tumour cell X fibroblast hybrids in which malignancy is suppressed..7. Cell Set. 25, 73-86. TODARO, G. J., HABEL, K. & GREEN, H. (1965). Antigenic and cultural properties of cells doubly transformed by polyoma virus and SV40. Virology 27, 179-185. VENUTA, S. & RUBIN, H. (1973). Sugar transport in normal and Rous sarcoma virus-transformed chick embryo fibroblasts. Proc. natn. Acad. Sri. U.SA. 70, 653-657. WALEY, S. G. (1964). A note on the kinetics of multi-enzyme systems. Biochem. J. 91, 514—517. WARREN, L., BUCK, C. A. & TUSZYNSKI, G. P. (1978). Glycoprotein changes and malignant

80


transformation. A possible role for carbohydrate in malignant behaviour. Biochim. biophys.Acta 516, 97-127. WATT, F. M., HARRIS, H., WEBER, K. & OSBORN, M. (1978). The distribution of actin cables and

microtubules in hybrids between malignant and non-malignant cells, and in tumours derived from them. J. Cell Sci. 32, 419-432. WEBER, G. (1974). Molecular correlation concept. In The Molecular Biology of Cancer (ed. H. Busch), pp. 487-521. New York: Academic Press. WEBER, M. J. (1973). Hexose transport in normal and in Rous sarcoma virus-transformed cells. J. biol. Chem. 248, 2978-2983. WERDAN, K., LEHNER, K., CREMER, T., STEVENSON, A. F. G. & MESSERSCHMIDT, O. (1980).

D-glucose transport in suspended human fibroblasts. Rapid measurement of uptake by silicone oil filtration centrifugation and comparison of different cell detachment procedures. HoppeSeyler's Z. physiol. Chem. 361, 91-104. WHITE, M. K., BRAMWELL, M. E. & HARRIS, H. (1981). Hexose transport in hybrids between malignant and non-malignant cells. Nature, Land. 294, 232-235. WHITE, M. K., BRAMWELL, M. E. & HARRIS, H. (1982). Hexose transport in hybrids between malignant and non-malignant cells. Nature, hand. 296, 588. WHITFIELD, C. D . , HUPE, L. M., NUGENT, C , URBANI, K. E. & WHITFIELD, H. J. (1982).

Increased hexose transport in CHO cells resistant to 3-O-methylglucose. J. biol. Chem. 257, 4902-4906. WIENER, F., KLEIN, G. & HARRIS, H. (1971). The analysis of malignancy by cell fusion. III. Hybrids between diploid fibroblasts and other tumour cells. J . Cell Sci. 8, 681-692. WIENER, F., KLEIN, G. & HARRIS, H. (1973). The analysis of malignancy by cell fusion. IV. Hybrids between tumour cells and a malignant L cell derivative. J. Cell Sci. 12, 253-261. WIENER, F., KLEIN, G. & HARRIS, H. (1974a). The analysis of malignancy by cell fusion. V. Further evidence of the ability of normal diploid cells to suppress malignancy. J. Cell Sci. 15, 177-183. WIENER, F., KLEIN, G. & HARRIS, H. (19746). The analysis of malignancy by cell fusion. VI. Hybrids between different tumour cells. J . Cell Sci. 16, 189-198. WOHLHUETER, R. M. & PLAGEMANN, P. G. W. (1980). The roles of transport and phosphorylation in nutrient uptake in cultured animal cells. Int. Rev. Cytol. 64, 171-240. YULI, I., WILBRANDT, W. & SHINITZKY, M. (1981). Glucose transport through cell membranes of modified lipid fluidity. Biochemistry 20, 4250-4256.

(Received 20 December 1982 -Accepted 20 December 1982)