Metabolic Cooperation Between Di€Erent Oncogenes During

Metabolic cooperation between dierent oncogenes during cell transformation: interaction between activated ras and HPV-16 E7

Sybille Mazurek*,1,4, Werner Zwerschke2,3,4, Pidder Jansen-DuÈ rr2 and Erich Eigenbrodt1

1Institute for Biochemistry and Endocrinology, Veterinary Faculty, University of Giessen, Frankfurter Strasse 100, 35392 Giessen, Germany; 2Institute for Biomedical Ageing Research, Austrian Academy of Sciences, Rennweg 10, A-6020 Innsbruck, Austria; 3Tyrolean Cancer Research Institute, Innrain 66, A-6020 Innsbruck, Austria

The metabolism of tumor cells (tumor metabolome) is linked to the disadvantage that those cells undergo characterized by a high concentration of glycolytic apoptosis under glucose limitation (Carmeliet et al., enzymes including pyruvate kinase isoenzyme type M2 1998; Shim et al., 1998). Glucose limitation is also (M2-PK), a high glutaminolytic capacity, high fructose found in solid tumors. To circumvent this disadvan- 1,6-bisphosphate (FBP) levels and a low (ATP+ tage, tumor cells use the amino acids glutamine and GTP):(CTP+UTP) ratio. The sequence of events serine as alternative sources for energy production. In required for the establishment of the tumor metabolome analogy to glycolysis the conversion of glutamine to is presently unknown. In non-transformed rat kidney lactate has been termed glutaminolysis; the ¯ow of (NRK) cells we observed a high glutaminolytic ¯ux rate serine to lactate has been termed serinolysis (Eigen- and a low (ATP+GTP):(CTP+UTP) ratio, whereas brodt et al., 1998a; Mazurek et al., 2001a,b). FBP levels and M2-PK activity are still extremely low. To coordinate glycolysis, glutaminolysis and serino- After stable expression of oncogenic ras in NRK cells a lysis, proliferating cells express a certain isoenzyme of strong upregulation of FBP levels and of M2-PK activity pyruvate kinase, referred to as pyruvate kinase type was observed. Elevated FBP levels induce a tetrameriza- M2 (M2-PK) (Eigenbrodt et al., 1985). M2-PK occurs tion of M2-PK and its migration into the glycolytic in a tetrameric form with a high anity for its enzyme complex. AMP levels increase whereas UTP and substrate phosphoenolpyruvate (PEP) and in a dimeric CTP levels strongly decrease. Thus, ras expression form with a low PEP anity. The glycolytic completes the glycolytic part of tumor metabolism phosphometabolite fructose 1,6-bisphosphate (FBP) leading to the inhibition of nucleic acid synthesis and induces the reassociation of the dimeric to the cell proliferation. The HPV-16 E7 oncoprotein, which tetrameric form (Eigenbrodt et al., 1992). When cooperates with ras in cell transformation, directly binds M2-PK is dimeric, glucose carbons are channelled to to M2-PK, induces its dimerization and restores nucleic synthetic processes, preferentially catalyzed by trans- acid synthesis as well as cell proliferation. Apparently, ketolase to nucleic acids (Glossmann et al., 1981; the combination of the dierent metabolic eects of ras Eigenbrodt et al., 1992; Boros et al., 1997, 2000; and E7 constructs the perfect tumor metabolome as Zwerschke et al., 1999; Cascante et al., 2000; Mazurek generally found in tumor cells. Oncogene (2001) 20, et al., 2001b) The ¯ow of glutamine to lactate increases 6891 ± 6898. to ensure energy production when M2-PK is inhibited (Mazurek et al., 1999, 2001b). Aside from mitochon- Keywords: metabolome; glycolysis; glutaminolysis; M2- drial glutaminolytic energy production, the glycolytic PK; adenylate kinase; nucleotides pyruvate kinase reaction is the only way to produce energy in tumor cells. Therefore, pyruvate kinase controls the ATP:ADP and GTP:GDP ratio and, in Introduction cooperation with adenylate kinase and 6-phosphofruc- to 1-kinase, the AMP levels (Mazurek et al., 1998; Tumorigenesis is generally characterized by an increase Thomas and Fell, 1998). An increase in AMP levels in the glycolytic enzyme system (Eigenbrodt et al., inhibits DNA synthesis and cell proliferation by a 1985, 1992; Board et al., 1990; Dang and Semenza, reduction of NAD(H) as well as UTP levels and CTP 1999). The increased glycolytic capacity allows tumor synthesis (Henderson and Scott, 1981; Mazurek et al., cells to migrate and survive under hypoxic conditions 1997a,b). By this mechanism M2-PK activity adapts as occur in solid tumors (Eigenbrodt et al., 1985; Kim cell proliferation to the nutrient supply. The metabolic et al., 1997; Reshkin et al., 2000). This advantage is network that is controlled by M2-PK has been termed the metabolome (Mazurek et al., 2001a,b). All tumor cells analysed to date express a high amount of the low PEP ane dimeric form of M2-PK (Eigenbrodt et al., *Correspondence: S Mazurek; 1992). M2-PK was shown to be directly targeted by E-mail: [email protected] 4These authors contributed equally to this paper certain oncoproteins, such as pp60v-src kinase and the Received 23 May 2001; revised 3 July 2001; accepted 5 July 2001 E7 oncoprotein of the human papillomavirus type 16 Interaction between ras and HPV-16 E7 S Mazurek et al 6892 (Presek et al., 1988; Eigenbrodt et al., 1998b; Results Zwerschke et al., 1999). The pp60v-src kinase phosphor- ylates M2-PK in tyrosine; the HPV 16-E7 oncoprotein Increased glutaminolysis and serinolysis in ras-expressing directly binds to M2-PK. Both oncoproteins trigger the rat kidney cells dimerization of M2-PK (Eigenbrodt et al., 1998b; Zwerschke et al., 1999). Due to the central regulatory A cell line (referred to as 14/2 cells) was derived from role of M2-PK, the interaction of the E7 oncoprotein normal rat kidney (NRK) cells by stable transfection with this key glycolytic enzyme dierentially in¯uences with a constitutive expression vector for EJras and a glycolysis, glutaminolysis, serinolysis, phosphometabo- hormone-inducible expression vector for HPV-16 E7 lite levels and nucleic acid metabolism in a way that (Crook et al., 1989). Expression of oncogenic ras leads depends on the individual enzyme features of the host to growth arrest which is overcome by expression of cell line (Mazurek et al., 2001b). E7 (Crook et al., 1989; Serrano et al., 1997). This A detailed metabolome analysis has been performed system allows to compare the phenotype of parental for the investigation of the metabolic eect of E7 rat kidney cells with the properties of ras-expressing transformation on two dierent NIH3T3 cell strains (referred to as ras-expressing RK cells below), and ras/ (high and low glycolytic) (Mazurek et al., 2001b). E7-coexpressing rat kidney cells (referred to as ras/E7- NIH3T3 cells, which are already immortal, are expressing RK cells below), and hence reveals transformed by E7 expression alone whereas transfor- consequences of the sequential expression of both mation of primary normal rat kidney cells (NRK cells) oncogenes. Parental NRK cells were characterized by requires the expression of two cooperating oncopro- low rates of glucose consumption and lactate produc- teins such as ras and E7 (Bedell et al., 1987; Crook et tion. Constitutive ras-expression, as in 14/2 cells, led to al., 1989; Zwerschke and Jansen-DuÈ rr, 2000). Metabo- a drastic increase of both (Table 1). The ratio between lome analysis has not yet been used to investigate the lactate production and glucose consumption decreased metabolic cooperation of two oncoproteins during from 1.9 in NRK cells to 1.4 in ras-expressing RK transformation. To obtain insight in metabolic altera- cells (Table 2). In glycolysis, 1 mole of glucose is tions involved in cell transformation by ras and E7, we converted into 2 moles lactate. A ratio between lactate analysed the metabolome of primary NRK, a ras- production and glucose consumption of nearly two expressing cell line derived from NRK as well as NRK- indicates that all lactate produced must be derived derived cells expressing both ras and E7. In NRK cells, from glucose. Therefore, the decrease in the ratio ras-expression leads to an inhibition of proliferation between glucose consumption and lactate production that is restored by the additional expression of E7. means that in ras-expressing RK cells either more Therefore, oncogene-expressing rat kidney cells are a glucose was channeled to synthetic processes or that good model to investigate links between metabolic more lactate derived from other sources than glucose. control and the regulation of cell proliferation. This question can be clari®ed by the comparison of the

Table 1 Nutrient ¯uxes in NRK cells, ras-expressing RK cells as well as ras/E7-expressing RK cells NRK RK+ras RK+ras+E7 Metabolites x+S.E.M. (nmoles/(h×105 cells))

Glucose consumption 43.6+3.7 84.5+3.3*** 120.9+2.5*** Lactate production 83.5+2.0 148.7+2.7*** 218.5+5.7*** Pyruvate consumption 3.8+0.1 9.9+0.5*** 7.3+0.5* Glutamine consumption 11.3+1.5 9.4+3.8n.s. 8.4+3.9n.s. Glutamate production 6.5+0.3 6.5+0.3n.s. 5.7+0.5n.s. Serine consumption 8.5+1.2 30.9+3.7*** 23.2+5.2n.s. Alanine production 3.5+0.3 1.7+0.2*** 1.8+0.2n.s.

The ¯ux rates were measured at an average cell density of 256105 cells/dish. NRK cells were compared with ras- expressing RK cells and ras-expressing RK cells were compared with ras/E7-expressing RK cells. *=P50.05; **P50.01; ***P50.001. n.s.=not signi®cant. nNRK=6, nRK+ras=5, nRK+ras+E7=5

Table 2 Correlations between dierent metabolite conversion rates in NRK, ras-expressing RK and ras/E7-expressing RK cells NRK RK+ras RK+ras=E7 Abscissa Ordinate Slope Intercept r Slope Intercept r Slope Intercept r

Glucose Lactate 1.9 53 0.974 1.4 558 0.992 1.7 221 0.997 Pyruvate Lactate 23.6 ±294 0.991 10.4 1181 0.973 17.8 1901 0.973 Glutamine Lactate 2.9 1515 0.375 3.3 2622 0.601 5.7 3479 0.678 Alanine Lactate 24.9 130 0.955 28.6 2346 0.960 42.9 3310 0.892 Glucose Pyruvate 0.08 16 0.963 34 ±55 0.961 0.09 ±81 0.980

For these calculations the ®rst metabolite (abscissa (nmoles/(h dish)) was plotted versus the second metabolite (ordinate (nmoles/h dish)). n=15

Oncogene Interaction between ras and HPV-16 E7 S Mazurek et al 6893 intercepts that re¯ect lactate production without Ras dependent reduction in the activity of hydrogen glucose consumption. In comparison to the parental shuttle enzymes NRK cells, in ras-expressing RK cells an approxi- mately 10-fold higher amount of lactate derived from The consumption of extracellular pyruvate is necessary other sources than glucose (Table 2). This lactate can to recycle NAD for an active glyceraldehyde 3- derive from the degradation of the amino acids phosphate dehydrogenase reaction. When more glutamine (glutaminolysis) and serine (serinolysis) or NADH is produced by the glyceraldehyde 3-phosphate from direct reduction of extracellular pyruvate (Figure dehydrogenase reaction than pyruvate is produced by 1). All three pathways increased in ras-expressing RK the glycolytic sequence, the ratio between glucose cells. Serine consumption increased 3.6-fold in ras- consumption and extracellular pyruvate consumption expressing RK cells (Table 1). Total glutamine increases greatly (Mazurek et al., 2001b). Indeed, when consumption calculated in (nmoles/(h×105 cells)) was glucose consumption was plotted versus pyruvate not signi®cantly altered after ras expression but the consumption the slope of the regression line dramati- increase of the linkage between glutamine consumption cally increased from 0.08 in NRK to 34 in ras- and lactate production from r=0.375 in NRK cells to expressing RK cells (Table 2). A factor that in¯uences r=0.601 in ras-expressing RK cells re¯ected an the ratio between glucose consumption and pyruvate increased conversion of glutamine to lactate (Tables consumption is the capacity of the shuttles that 1 and 2). Extracellular pyruvate consumption increased transport hydrogen from the cytosol into the mito- 2.6-fold in ras-expressing RK cells (Table 1). Since one chondria (Mazurek et al., 1997b). Investigation of the mole of pyruvate is converted into one mole of lactate enzymes involved in the malate aspartate shuttle the slope of the regression line when pyruvate revealed a strong inactivation of the cytosolic gluta- consumption (abscissa) is plotted versus lactate mate oxaloacetate transaminase (GOT) activity in ras- production (ordinate) is one, when the lactate expressing RK cells (Figure 2). Down regulation of the produced is only derived from extracellular pyruvate. cytosolic GOT (cGOT) is accompanied by an upregu- A slope steeper than 1.0 indicates that more lactate is lation of the mitochondrial GOT isoenzyme (mGOT). derived from sources other than pyruvate and that less However, mGOT cannot functionally replace cGOT. pyruvate is converted to lactate. In NRK cells the ratio between pyruvate consumption and lactate Ras-dependent changes in the glycolytic enzyme complex production decreased from 23.6 (without ras-expression) to 10.4 (with ras-expression), indicating that As determined by gel permeation experiments, the more extracellular pyruvate was converted to lactate in expression of ras in NRK cells led to a large increase in ras-expressing RK cells than in the parental NRK cells the total activity of M2-PK and a signi®cant shift (Table 2). towards the tetrameric form of M2-PK (Figure 3). In

Figure 1 Metabolic scheme of the interaction between glycolysis, glutaminolysis, serinolysis and nucleic acid metabolism. Green=glycolysis; blue=glutaminolysis; yellow=serinolysis; red=nucleic acid metabolism. M2-PK=pyruvate kinase type M2; AK 2=adenylate kinase type 2

Oncogene Interaction between ras and HPV-16 E7 S Mazurek et al 6894 ras-expressing RK cells, the mass of M2-PK was Mazurek et al., 1996, 1998; Nagy et al., 2000). The associated within the glycolytic enzyme complex glycolytic enzyme complex can be demonstrated by (Figure 4). The glycolytic enzyme complex is a labile isoelectric focusing. Proteins associated within the structure, which is formed by a variety of glycolytic glycolytic enzyme complex focus at a common isoenzymes and AU rich mRNA (Nagy and Rigby, 1995; electric point that is dierent from the isoelectric point of the puri®ed individual proteins. Migration of proteins inside or outside the glycolytic enzyme complex are re¯ected by shifts in their isoelectric points (Mazurek et al., 1996, 1998, 1999, 2001b). In the case of M2-PK only the tetrameric form is associated within the glycolytic enzyme complex. The tetramer:dimer ratio of M2-PK is primarily regulated by the FBP levels (Eigenbrodt et al., 1992). This was also shown by in vitro incubation of NRK- cell extracts with 200 mM FBP which led to a tetramerization of M2-PK and a migration of M2-PK into the glycolytic enzyme complex (data not shown). In ras-expressing RK cells the determination of FBP levels revealed a more than 30-fold increase compared to the parental cells (Table 3). In ras-expressing RK cells but not in the parental NRK cells adenylate kinase (AK) is associated with M2- PK within the glycolytic enzyme complex (Figure 4). Adenylate kinase catalyzes the conversion of 2 moles of ADP to 1 mole ATP and 1 mole AMP (Figure 1). NRK cells express the AK isoenzymes type 2 and 3. AK2 is Figure 2 Isoelectric focusing of glutamate oxaloacetate transa- generally found in proliferating tissues and is associated minase isoenzymes in NRK (*), ras-expressing RK (&) and ras/ within the glycolytic enzyme complex and with mito- E7-expressing RK cells (~) chondria (Inouye et al., 1999; Kohler et al., 1999; Noma

Figure 3 Quaternary structure of M2-PK in NRK cells (*), ras-expressing RK (&) cells and ras/E7-expressing RK cells (!). Results of gel permeations of cytosolic cell extracts. As internal molecular weight markers the activity of GAPDH (160 kDa), enolase (100 kDa), and PGM (60 kDa) were measured in every gel permeation experiment

Oncogene Interaction between ras and HPV-16 E7 S Mazurek et al 6895 et al., 1999). In ras-expressing RK cells the migration of consumption and lactate production from 1.4 to 1.7 M2-PK into the glycolytic enzyme complex in close (Table 2). The intercept decreased from 558 ± proximity to AK2 correlated with the decrease of ATP 221 nmoles/(h dish) indicating that E7-expression pre- levels and the increase of AMP levels (Figure 4 and Table dominantly reduced the ¯ux from serine to lactate. The 3). In the parental NRK cells AK was not associated glutaminolytic ¯ux was not altered whereas that of with M2-PK and AMP was not measurable (Table 3). pyruvate consumption decreased and less pyruvate was The ATP:ADP ratio is regulated by the activity of M2- channelled to lactate (Tables 1 and 2). In accordance, PK whereas AMP levels are regulated by the activity of the ratio between pyruvate consumption and lactate AK (Figure 1). In ras-expressing RK cells but not in the production increased from 10.4 to 17.8 upon E7- parental NRK cells the ATP:ADP ratio dropped with expression (Table 2). When pyruvate consumption was increasing AMP levels (slope of the regression line, when plotted versus glucose consumption the slope of the ATP:ADP (ordinate) was plotted versus AMP levels regression line dropped from 34 in ras-expressing RK (abscissa)=72.1, r=0.633). cells to 0.09 in ras/E7-expressing RK cells similar to the value that was found in the parental NRK cells (Table 2). Therefore, the expression of E7 reactivates Metabolic cooperation of E7 with activated ras oncogene the capacities of the malate aspartate shuttle to values Expression of HPV-16 E7 in ras-expressing RK cells found in normal proliferating cells. In accordance, led to an increase in the ratio between glucose investigation of the enzymes involved in the malate aspartate shuttle revealed a strong reactivation of the cytosolic GOT activity in ras/E7-expressing RK cells (Figure 2). Furthermore, E7-expression led to a shift to the dimeric form of M2-PK and to a release of M2-PK out of the glycolytic enzyme complex (Figure 3 and data not shown). This correlated with a further increase in the levels of FBP (Table 3). However, this increase was not sucient to stabilize the tetrameric form of M2-PK under these conditions, probably due to the direct ability of E7 to shift the M2-PK equilibrium to the dimeric form, as a result of direct binding (Zwerschke et al., 1999; Mazurek et al., 2001a). As a consequence of the E7 induced dimerization of M2-PK, PEP levels increased and the ATP:ADP ratio as well as the GTP:GDP ratio decreased (Table 3). Furthermore, the ¯ux of serine to lactate was reduced (Table 1). Serine is an essential precursor of purine and pyrimidine synthesis. In accordance, in ras/E7-expressing RK cells IMP, the synthetic product of purine de novo synthesis, accumulated and AMP levels consisted Figure 4 Association between pyruvate kinase type M2 (M2- PK) and adenylate kinase type 2 (AK 2) in ras-expressing RK at high levels (Table 3). The inhibition of P-ribose-PP cells. Result of an isoelectric focusing experiment with cytosolic synthetase by AMP was overcome by the increased cell extracts conversion of glucose to ribose 5-P which led to an

Table 3 Intracellular nucleotide and metabolite levels in NRK cells, ras-expressing RK cells and ras/E7-expressing RK cells Intracellular NRK RK+ras RK+ras+E7 nucleotides/metabolites x+S.E.M. (nmoles/mg protein)

ATP 175.0+18.0 101.6+9.1** 73.3+8.3* ADP 35.2+4.7 16.5+2.5** 20.8+1.6n.s. ATP:ADP 5.1+0.4 5.2+0.6n.s. 3.5+0.3** AMP n.d. 1.6+0.4* 1.8+0.4n.s. IMP n.d. 0.2+0.4 2.0+1.7* GTP 15.8+2.6 34.0+3.9** 22.0+3.3* GTP:GDP 2.4+0.3 8.6+0.8*** 4.5+0.6** CTP 36.2+3.9 21.2+2.8* 20.9+1.5n.s. UTP 51.1+2.1 13.2+3.3*** 30.4+4.2* UDP-glucnac 24.1+4.5 54.4+5.8** 56.7+6.4n.s. (ATP+GTP): 2.2+0.1 5.3+0.8** 1.5+0.3** (CTP+UTP) FBP 0.4+0.1 (5) 13.4+0.6 (5)*** 28.3+5.9 (5)* PEP 2.4+0.5 (5) 0.9+0.1 (5)* 1.7+0.2 (5)* n=5

Oncogene Interaction between ras and HPV-16 E7 S Mazurek et al 6896 increase in UTP levels in ras/E7-expressing RK cells Moyer et al., 1982; Linke et al., 1996; Downes et al., (Figure 1 and Table 3). The (ATP+GTP): 2000). Low ATP levels and low ATP:ADP ratio, as (UTP+CTP) ratio drastically dropped to a value of found in ras/E7-expressing RK cells, are generally 1.5 (Table 3). correlated with high malignancy (Weber et al., 1971; Jackson et al., 1980).

Discussion Metabolic consequences of ras-induced tetramerization of M2-PK In the present study the eect of cooperation between ras and HPV-16 E7 on the metabolic properties of As in other cell lines, in NRK cells oncogenic ras NRK cells in culture was investigated. We found that led to an overproportional increase of glycolytic expression of ras in rat kidney cells led to a enzymes, especially in 6-phosphofructo 1-kinase signi®cant increase in the consumption of glucose (PFK 1) (Gregory et al., 1976; Kole et al., 1991). and in the production of lactate (Table 1). The increase of PFK 1 leads to an increase in FBP Furthermore, ras-expressing cells use at least three levels (Table 3). In ras-expressing RK cells the high additional pathways to generate lactate: degradation FBP levels induced a tetramerization of M2-PK that of glutamine (glutaminolysis), degradation of serine led to a decrease in the conversion of glucose (serinolysis) and reduction of extracellular pyruvate carbons to pyrimidine synthesis (Figure 3). In (Figure 1). All three substrates can be used for the accordance, in ras-expressing RK cells the GTP:GDP synthesis of cell building materials such as nucleic ratio increased whereas PEP, UTP and CTP levels acids and phospholipids or can be degraded to decreased (Table 3). The ATP:ADP ratio persisted at lactate. Ras expression led to a tetramerization and its levels in ras-expressing RK cells since ATP and activation of M2-PK as well as to a reduction of ADP levels simultaneously decreased (Table 3). The cytosolic GOT activity (Figures 2 and 3). Both selective decrease in ATP and ADP levels was interactions led to an increase of the ¯ow of caused by the migration of the tetrameric M2-PK glutamine, serine and pyruvate to lactate (Tables 1 into the glycolytic enzyme complex in close and 2). Therefore, the metabolome of ras-expressing proximity to AK 2 (Figures 1 and 4). AK 2 RK cells is characterized by a more catabolic state. catalyzes the conversion of 2 moles of ADP into E7-expression led to promotion of the dimeric form 1 mole ATP and 1 mole AMP (Figure 1). Therefore, of M2-PK which favors synthetic processes such as in ras-expressing RK cells, M2-PK and AK2 nucleic acid synthesis (Figures 1 and 3). Therefore, cooperate to decrease the ATP:ADP ratio and to E7-expression shifts ras-expressing RK cells to a increase AMP levels (Table 3). more anabolic state. Due to the tetramerization of M2-PK in ras- The metabolic eect of the interaction between ras expressing RK cells, the conversion of glucose carbons and pyruvate kinase has been conserved during to nucleic acid synthesis was reduced and serine was evolution (Chopade et al., 1997). In yeast cells, predominantly converted to lactate (Tables 1 and 2). over-expression of pyruvate kinase leads to an The NRK-14/2 cell model provides further evidence inhibition of cell proliferation (Brazill et al., 1997). that tumor formation is correlated with relatively high Comparable to ras-expressing RK cells the inhibition glycolytic and glutaminolytic capacities, high glycolytic of cell proliferation in yeast cells is generally phosphometabolite levels, such as FBP and a low correlated with an increase of catabolic processes, (ATP+GTP):(CTP+UTP) ratio (Figure 1 and Tables such as the conversion of glucose to ethanol and a 1 ± 3) (Eigenbrodt et al., 1985; Ryll and Wagner, 1992; decrease of anabolic processes, such as a decrease in Mazurek et al., 1997a,b, 2001b; Penso and Beitner, the conversion of glucose to synthetic processes (Aon 1998; Glass-Marmor and Beitner, 1999; Zwerschke et et al., 1995). al., 1999). In the parental NRK cells and NRK derived cell The upregulation of all glycolytic enzymes by ras lines, from all products synthesized from glucose, might either be induced by stabilization of the UTP levels correlated best with cell proliferation transcription factor HF1a or by modulation of (Table 3). In proliferating NRK cells the ratio of Sp1/Sp3 functions (SchaÈ fer et al., 1997; Carmeliet (ATP+GTP):(UTP+CTP) was 2.2 whereas in non- et al., 1998). This function of ras can partly be proliferating ras-expressing RK cells the ratio be- undertaken by other oncoproteins such as myc or tween the purine and pyrimidine nucleotides increased Ets (Dupriez et al., 1993; Shim et al., 1997). Myc to a value of 5.3 (Table 3). This is in accordance induces an increase in lactate dehydrogenase type A with several other investigations in bacteria, plants and enolase type a activities. Ets upregulates fructose and mammalian cells that have shown that cell 2,6-bisphosphate kinase. Interestingly, the overexpres- proliferation is correlated with a decrease in ATP sion of lactate dehydrogenase type A is required for and GTP levels as well as with an increase in UTP myc-mediated transformation of rat 1A ®broblasts and CTP levels (Ryll and Wagner, 1992; Mazurek et (Shim et al., 1997). Therefore, it is conceivable that al., 2001b). Inhibition of cell proliferation is caused the overexpression of glycolytic enzymes can partly by a decrease of UTP and CTP levels and their complete the ras speci®c metabolic phenotype in correspondent dNTP levels (Hunting et al., 1981; NRK cells.

Oncogene Interaction between ras and HPV-16 E7 S Mazurek et al 6897 middle and high) in the presence of dexamethasone. There- Materials and methods after, dexamethasone was removed to eliminate E7-expression and to synchronize the cells. Flux measurements were Cell lines performed from mock-treated cells (ras-expressing RK) and Normal rat kidney cells (NRK) were purchased from ATCC from cells treated with dexamethasone for 4 h (ras/E7- and cultivated in Dulbecco's modi®ed Eagle's medium transformed cells). Cell culture supernatants were collected (DMEM), supplemented with 2 mM glutamine, 100 units at dierent cell densities and the metabolites were measured penicillin/ml, 100 mg streptomycin/ml and 10% fetal calf as described in Mazurek (1997b). For the determination of serum (FCS), all from SIGMA (Deisenhofen, Germany). For the ¯ux rates, two dierent forms of calculation were chosen. the ¯ux measurements the cells were arrested in DMEM The ®rst calculation is in (nmoles/(h 105 cells)) and describes medium containing 0% (v/v) CS for 72 h, washed twice in the consumption and production of a certain metabolite by DMEM and then released in DMEM medium containing each cell (Table 1). The second calculation in (nmoles/(h 10% (v/v) FCS. The supernatants were analysed 4 h after dish)) allows the detection of possible correlations between release. For all other experiments DMEM medium contain- the conversion rates of dierent metabolites (Table 2). The ing 10% (v/v) FCS was used. The 14/2 cell line has been combination of both calculations provides an individual derived from rat kidney cells by stable expression of an metabolic characterization of each cell strain (Mazurek et al., activated ras oncogene, combined with glucocorticoid- 1997b, 1999, 2001b). inducible expression of the HPV-16 E7 gene. 14/2 cells were grown in DMEM with 10% FCS and dexamethasone, as Nucleotide and metabolite measurements described (Crook et al., 1989). The cells were extracted with 0.6 N HClO4. FBP and PEP were determined enzymatically as described in Mazurek et al. Gel permeation (1997b). Nucleotides were measured via reversed-phase-ion Cells were extracted in a lysis buer containing 100 mM pair-liquid chromatography as described in Mazurek et al. Na2HPO4/NaH2PO4,1mM DTT, 1 mM NaF, 1 mM mer- (2001b). captoethanole, 1 mM e-aminocaproic acid, 0.2 mM PMSF and 10% glycerol, pH 7.4). The extracts were passed over a Statistical analysis gel permeation column (Pharmacia) and the activities of the dierent enzymes were determined in the fractions as For the glycolytic, glutaminolytic and serinolytic ¯ux described (Mazurek et al., 1997b). measurements statistical analyses were performed using the statistical program package BMDP. For the comparison of the dierent cell groups two-way analysis of covariance with Isoelectric focusing cell density was performed. In all other cases Student's t-test Cells were extracted with a homogenisation buer containing was employed. 10 mM Tris, 1 mM NaF and 1 mM mercaptoethanole, pH 7.4. Isoelectric focusing was carried out with a linear gradient of glycerine (50 to 0% (v/v)) and ampholines (pI 3.5 ± 10.5) as described previously (Mazurek et al., 1996). Acknowledgments This work was supported by the Land Hessen (S Mazurek: HSP III), the Deutsche Forschungsgemeinschaft (S Mazur- Flux measurements ek: Ma 1760 1-2 and 2-1) and in P Jansen-DuÈ rr's Metabolic ¯ux rates strongly depend on cell density. To laboratory by the Austrian Science Funds (FWF), the measure the ¯ux rates at dierent cell densities ras/E7- European Union (Biomed 2) and the Austrian Ministry of expressing RK cells were grown to a certain cell density (low, Science and Trac.

References

Aon MA, Monaco ME and Cortassa S. (1995). Exp. Cell Cascante M, Centelles JJ, Veech RL, Lee W-NP and Boros Res., 217, 42 ± 51. LG. (2000). Nutr. Cancer, 36, 150 ± 154. Bedell MA, Jones KH and Laimins LA. (1987). J. Virol., 61, Chopade BA, Shankar S, Sundin GW, Mukhopadhyay S and 3635 ± 3640. Chakrabarty AM. (1997). J. Bacteriol., 179, 2181 ± 2188. Board M, Humm S and Newsholme EA. (1990). Biochem. J., Crook T, Morgenstern JP, Crawford L and Banks L. (1989). 265, 503 ± 509. EMBO J., 8, 513 ± 519. BorosLG,PuigjanerJ,CascanteM,LeeW-N,BrandesJL, Dang CV and Semenza GL. (1999). TIBS, 24, 68 ± 72. Bassilian S, Yusuf FI, Williams RD, Muscarella P, Melvin Downes CS, Bachrati CZ, Devlin SJ, Tommasino M, Cutts WS and Schirmer WJ. (1997). Cancer Res., 57, 4242 ± TJR, Watson JV, Rasko I and Johnson RT. (2000). J. Cell 4248. Sci., 113, 1089 ± 1096. Boros LG, Torday JS, Lim S, Bassilian S, Cascante M and Dupriez VJ, Darville MI, Antoine IV, Gegonne A, Ghysdael Lee W-NP. (2000). Cancer Res., 60, 1183 ± 1185. J and Rousseau GG. (1993). Proc.Natl.Acad.Sci.USA, Brazill DT, Thorner J and Martin GS. (1997). J. Bacteriol., 90, 8224 ± 8228. 179, 4415 ± 4418. Eigenbrodt E, Fister P and Reinacher M. (1985). Regulation Carmeliet P, Dor Y, Herbert J-M, Fukumura D, Brussel- of Carbohydrate Metabolism. Beitner R. (ed.). CRC Press mans K, Dewerchin M, Neeman M, Bono F, Abramovitch Inc.: Boca Raton, Fl, pp. 141 ± 179. R, Maxwell P, Koch CJ, Ratclie P, Moons L, Jain RK, Eigenbrodt E, Kallinowski F, Ott M, Mazurek S and Vaupel Collen D and Keshet E. (1998). Nature, 394, 485 ± 490. P. (1998a). Anticancer Res., 18, 3267 ± 3274.

Oncogene Interaction between ras and HPV-16 E7 S Mazurek et al 6898 Eigenbrodt E, Mazurek S and Friis RR. (1998b). Cell Growth Mazurek S, Zwerschke W, Jansen-DuÈ rr P and Eigenbrodt E. and Oncogenesis. Bannasch P, Kanduc D, Papa S and (2001b). Biochem. J., 356, 247 ± 256. Tager JM. (eds). BirkhaÈ user Verlag: Basel, pp. 15 ± 30. Moyer JD, Smith PA, Levy EJ and Handschumacher RE. Eigenbrodt E, Reinacher M, Scheefers-Borchel U, Scheefers (1982). Cancer Res., 42, 4525 ± 4531. H and Friis R. (1992). Crit. Rev. Oncol., 3, 91 ± 115. Nagy E, Henics T, Eckert M, Miseta A, Lightowlers RN and Glass-Marmor L and Beitner R. (1999). Eur. J. Pharmacol., Kellermayer M. (2000). Biochem. Biophys. Res. Commun., 370, 195 ± 199. 275, 253 ± 260. Glossmann H, Presek P and Eigenbrodt E. (1981). Mol. Cell. Nagy E and Rigby FC. (1995). J. Biol. Chem., 270, 2755 ± Endocrinol., 23, 49 ± 63. 2763. Gregory SH, Kumari HL, Lakshimi MV and Bose SK. Noma T, Adachi N and Nakazawa A. (1999). Biochem. (1976). Arch. Biochem. Biophys., 175, 644 ± 653. Biophys. Res. Commun., 264, 990 ± 997. Henderson JF and Scott FW. (1981). Pharma. Ther., 8, 539 ± Penso J and Beitner R. (1998). Eur. J. Pharmacol., 342, 113 ± 571. 117. Hunting D, Hordern J and Henderson JF. (1981). Can. J. Presek P, Reinacher M and Eigenbrodt E. (1988). FEBS Biochem., 59, 821 ± 829. Lett., 242, 194 ± 198. Inouye S, Yamada Y, Miura K, Suzuki H, Kawata K, Reshkin SJ, Bellizzi A, Caldeira S, Albarani V, Malanchi I, Shinoda K and Nakazawa A. (1999). Biochem. Biophys. Poignee M, Alunni-Fabbroni M, Casavola V and Res. Commun., 254, 618 ± 622. Tommasino M. (2000). FASEB J., 14, 2185 ± 2197. Jackson RC, Lui MS, Boritzki TJ, Morris HP and Weber G. Ryll T and Wagner R. (1992). Biotechnol. Bioeng., 40, 934 ± (1980). Cancer Res., 40, 1286 ± 1291. 946. KimCY,TsaiMH,OsmanianC,GraeberTG,LeeJE, SchaÈ fer D, Hamm-KuÈ nzelmann B and Brand K. (1997). Giard RG, DiPaolo JA, Peehl DM and Giaccia AJ. FEBS Lett., 417, 325 ± 328. (1997). Cancer Res., 57, 4200 ± 4204. Serrano M, Lin AW, McCurrach ME, Beach D and Lowe Kohler C, Gahm A, Noma T, Nakazawa A, Orrenius S and SW. (1997). Cell, 88, 593 ± 602. Zhivotovsky B. (1999). FEBS Lett., 447, 10 ± 12. Shim H, Chun YS, Lewis BC and Dang CV. (1998). Proc. Kole HK, Resnick RJ, van Doren M and Racker E. (1991). Natl. Acad. Sci. USA, 95, 1511 ± 1516. Arch. Biochem. Biophys., 286, 586 ± 590. ShimH,DoldeC,LewisBC,WuC-S,DangG,Jungmann Linke SP, Clarkin KC, Di Leonardo A, Tsou A and Wahl RA, Dalla-Favera R and Dang CV. (1997). Proc. Natl. GM. (1996). Genes. Dev., 10, 934 ± 947. Acad.Sci.USA,94, 6658 ± 6663. Mazurek S, Boschek CB and Eigenbrodt E. (1997a). J. Thomas S and Fell DA. (1998). Eur. J. Biochem., 258, 956 ± Bioenerg. Biomemb., 29, 315 ± 330. 967. Mazurek S, Eigenbrodt E, Failing K and Steinberg P. (1999). Weber G, Stubbs M and Morris HP. (1971). Cancer Res., 31, J. Cell. Physiol., 181, 136 ± 146. 2177 ± 2183. Mazurek S, Grimm H, Boschek CB, Vaupel P and Zwerschke W and Jansen-DuÈ rr P. (2000). Adv. Cancer Res., Eigenbrodt E. (2001a). Br.J.Nutr.,In press. 78, 1 ± 29. Mazurek S, Grimm H, Wilker S, Leib S and Eigenbrodt E. Zwerschke W, Mazurek S, Massimi P, Banks , Eigenbrodt E (1998). Anticancer Res., 18, 3275 ± 3282. and Jansen-DuÈ rr P. (1999). Proc. Natl. Acad. Sci. USA, Mazurek S, Hugo F, Failing K and Eigenbrodt E. (1996). 96, 1291 ± 1296. J. Cell. Physiol., 167, 238 ± 250. Mazurek S, Michel A and Eigenbrodt E. (1997b). J. Biol. Chem., 272, 4941 ± 4952.

Oncogene