WO2015011222A1 - Activation of ras by a fructose-1,6-bisphosphate / guanine nucleotide exchange protein complex - Google Patents

Abstract

The present invention relates to an isolated fructose-1,6-bisphosphate / SOS complex. The invention relates further to the use of the complex to screen for compounds that modulate the activation of Ras, and the use of yeast strains for said screening. The invention relates also to compounds, disrupting the fructose-1,6-bisphosphate / SOS complex and the use of such compounds in the treatment of cancer.

Description

ACTIVATION OF RAS BY A FRUCTOSE-1 ,6-BISPHOSPHATE / GUANINE NUCLEOTIDE EXCHANGE PROTEIN COMPLEX

The present invention relates to an isolated fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex. The invention relates further to the use of the complex to screen compounds that modulate the activation of Ras, and to use of yeast strains for said screening. The invention relates also to compounds, disrupting the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex and the use of such compounds in the treatment of cancer.

Fermentation originated in evolution much earlier than respiration. The unusually high fermentative activity of cancer cells constitutes a reversal to an ancient state of energy generation. This suggests that the metabolic control that is defective in cancer cells may well be conserved in evolution. Yeast is one of the most prominent examples of eukaryotic cells, which displays in the presence of glucose and under aerobic conditions high fermentative activity as well as rapid cell proliferation, just like cancer cells. Yeast may thus represent an excellent model system to elucidate the connection between the high activity of the glycolytic pathway responsible for rapid fermentation and its possible role in stimulating cell proliferation (Diaz-Ruiz et al., 2009; Fiechter & Gmunder, 1989). Mutant RAS genes, encoding overactive Ras proteins, are the most common oncogenes found in naturally occurring and artificially induced cancer cells (Barbacid, 1987; Bos, 1989). High activity of Ras is thus well established as a trigger of rapid and uncontrolled cell proliferation, which makes Ras an attractive candidate as putative target for stimulation by high glycolytic activity. Ras is strongly conserved and also serves as an important regulator of cell proliferation in yeast, but acts through activation of cAMP synthesis and the protein kinase A (PKA) pathway (Broach & Deschenes, 1990; Thevelein & de Winde, 1999). Moreover, previous work has shown that glucose triggers rapid Ras-mediated cAMP signaling in yeast and that active glucose catabolism in glycolysis is required for stimulation of Ras activity (Colombo et al., 2004; Mbonyi et al., 1988; Van Aelst et al., 1990; van Aelst et al., 1991 ). Although Ras activation by glucose catabolism has not been described in mammalian cells, the inverse relationship has been well documented. Ras activates, by means of the PI 3-kinase/Akt pathway, the hypoxia-inducible factor 1 (HIF-1 ), which stimulates transcription of genes coding for glycolytic enzymes and glucose transporters. Constitutive activation of this signaling pathway has been documented in many types of cancer cells (Blum et al., 2005; lyer ei a/., 1998; Mazure et al., 1997; Semenza, 2003). The strong conservation of Ras and its regulators, Cdc25/Sos and Ira1 ,2/NF1 in yeast and mammalian cells, suggests that the unknown mechanism responsible for glycolytic activation of Ras may also be conserved between the two cell types. There is a striking correlation between the rate of uncontrolled cell proliferation and the aggressive metastasis character of cancers and the extent of the 'Warburg effect' (Lo et al., 1968; Mathupala et al., 1997; Weber & Lea, 1966). This observation has fueled the idea that the enhanced glycolytic rate may be a cause or at least may be important for induction or maintenance of cancerous growth. In nearly all cells, glucose is phosphorylated by hexokinase or glucokinase to glucose-6-phosphate (Glu6P) after its transport into the cells, then converted to fructose-6-phosphate by phosphoglucoisomerase and subsequently phosphorylated to Fructose-1 ,6-bisphosphate (Fru1 ,6bisP) by phosphofructokinase 1. Fru1 ,6bisP is by far the most elaborately controlled glycolytic metabolite since its biosynthetic and hydrolytic enzymes are post-translationally controlled by a specific allosteric regulator, Fru2,6bisP, which is synthesized and hydrolyzed in a parallel regulatory pathway (Hers, 1984). Fru1 ,6bisP is split into the triose phosphates dihydroxyacetone-phosphate (DHAP) and glyceraldehyde-3- phosphate (GAP), which are ultimately converted into pyruvate, and further into a fermentation product, either ethanol in yeast or lactic acid in mammalian cells. To identify a possible molecular connection between high glycolytic activity and overactive cell proliferation, much attention has focused on effects of oncogenes on glycolysis regulation (Semenza, 2003). It has been observed that not only the amount of the glycolytic enzymes is increased in cancer cells, but also that changes in the predominant isoenzymes with particular kinetic properties contribute to the distinctive high glycolytic rate (Herling et al., 201 1 ). Especially changes in the main rate-controlling steps, i.e. glucose transport (overexpression of the high affinity transporter GLUT1 (Macheda et al., 2005)), glucose phosphorylation (overexpression of HK-2 and enhancement of its association with mitochondria (Herling et al., 201 1 ; Mathupala et al., 1997)) and the rate of phosphofructokinase type 1 (increased enzymatic activity and concentration of the main allosteric activator Fru2,6bisP (Yalcin et al., 2009)) have been identified as contributing factors. Interestingly, the effects on the rate of phosphofructokinase type 1 are dependent on the Ras oncogene (Kole et al., 1991 ; Yalcin et al., 2009).

Hexokinase activity has received particular attention as one of the major determinants of the Warburg effect. In certain types of cancer cells, type II hexokinase is strongly associated with the mitochondrial ADP/ATP carrier, losing its feedback inhibition by glucose-6P and allowing much higher catalytic activity due to the efficient coupling to ATP provision at the inner mitochondrial membrane (Mathupala et al., 1997). When such cancer cells were grown on galactose, a sugar whose metabolism does not involve hexokinase, the high fermentation rate typical of the Warburg effect, was strongly reduced (Bustamante & Pedersen, 1977). Subsequent work showed that this alteration of type II hexokinase was sufficient to enhance the glycolytic rate of a normal cell up that observed in cancer cells (Bustamante et al., 1981 ).

Extensive additional evidence has highlighted the critical role of type II hexokinase in aberrant glycolytic metabolism of cancer cells (Mathupala et al., 1997). Since then, hexokinase has become a focus of interest as possible new target for anticancer drugs (Chen et al., 2009).

Whereas mammalian hexokinase is feedback-inhibited by its product Glu6P (Wilson, 2003), yeast hexokinase is feedback-inhibited by trehalose-6-phosphate (Tre6P) (Blazquez et al., 1993). Tre6P is synthesized from Glu6P and UDPG, and is the only intermediate in trehalose biosynthesis. Because they lack Tre6P, tpsIA cells have overactive hexokinase activity and therefore display uncontrolled influx of glucose into glycolysis. This is characterized by hyperaccumulation of metabolites in the upper part of glycolysis, especially Fru1 ,6bisP, rapid depletion of ATP and phosphate, and causes strong inhibition of growth by even low amounts of glucose (Neves et al., 1995; Van Aelst et al., 1993). Deletion of hexokinase 2, the most active hexokinase isoenzyme in yeast, completely suppresses aberrant glucose catabolism and restores growth on glucose of the tpsIA strain (Hohmann et al., 1993). Since galactose metabolism bypasses the hexokinase step for entry into glycolysis, tpsIA cells grow normally on galactose (and on non-fermentable carbon sources like glycerol and ethanol).

Surprisingly, we found that glucose addition to galactose-grown tpsIA cells triggers rapid and dramatic activation of Ras, as opposed to the modest activation observed in wild type cells. The hyperactivation of Ras is causing programmed cell death. Even more surprisingly, the hyperactivation of Ras is mediated by a Fru1 ,6bisP / guanine nucleotide exchange protein complex, more specifically a Fru1 ,6bisP / Sos complex; disruption of the complex blocks the hyperactivation of Ras.

A first aspect of the invention is an isolated fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex. A guanine nucleotide exchange protein, as used here is a protein that comprises at least the conserved regions 598-606 and 950-973 of HsSosl , or the equivalent regions of HsSos2, ScCdc25 or ScSdc25, as indicated in Figure 7. Preferably, said protein comprises the region 598-973 of hsSosl (numbering according to Genbank accession NP_005624, version NP_005624.2; Gl: 15529996), or the equivalent region in a variant guanine nucleotide exchange protein such as the region 596-971 of HsSos2 (numbering according to Genbank accession number NP_008870, version NP_008870.2, Gl: 154689780), 1 1 18-1501 of ScCdc25 (numbering according to Genbank accession number NP_013413, version NP_013413.1 ; Gl: 6323341 ) or 579-954 of ScScd25 (numbering according to Genbank accession number CAA97461 , version CAA97461 .1 ; Gl:1360187). Even more preferably, said guanine nucleotide exchange protein is selected from the group of variants consisting of HsSosl , HsSos2, ScCdc25 and ScSdc25. HsSosl , HsSos2, ScCdc25 and ScSdc25 are known to the person skilled in the art, and include but are not limited to the sequences above.

Other variants are, as a non-limiting example ScSdc25 genbank accession number

AAA16565, version AAA16565.1 ; Gl: 457494 and HsSos2 accession number AAI43368, version AAI43368.1 ; Gl: 219519967. The binding between Fru1 ,6bisP and the guanine nucleotide exchange protein may be reversible or irreversible.

Preferably, said complex between the guanine nucleotide exchange protein and Fru1 ,6bisP is mediated by the amino acids R962, K963 and/or K602 of HsSosl (numbering according to Genbank accession NP_005624, version NP_005624.2; Gl: 15529996) of the equivalent amino acid residue in a variant guanine nucleotide exchange protein.

Preferably, the complex according to the invention is further comprising a Ras protein. Ras proteins are known to the person skilled in the art and belong to a protein family of small GTPases (pfam PF00071 ), which are all related in 3D structure and function. As a non-limiting example, Ras proteins in human are HRAS, KRAS, NRAS, DIRAS1 , DIRAS2, DIRAS3, ERAS, GEM, MRAS, NKIRAS1 , NKIRAS2, NRAS, RALA, RALB, RAP1A, RAP1 B, RAP2A, RAP2B, RAP2C, RASD1 , RASD2, RASL10A, RASL10B, RASL1 1A, RASL1 1 B, RASL12, REM1 , REM2, RERG, RERGL, RRAD, RRAS and RRAS2; Ras proteins in yeast are Rasl p and Ras2p.

Another aspect of the invention is the use of a complex according to the invention to screen for compounds to modulate Ras activation. Preferably, said complex is a fructose-1 ,6- bisphosphate / guanine nucleotide exchange protein / Ras complex. Indeed, as the Fru1 ,6bisP binding to a guanine nucleotide exchange factor is essential for Ras activation, i.e. for induction or maintenance of the Ras-GTP bound state, compounds that interfere with this binding will modulate the Ras activity, either by inhibiting the activation, or by replacing Fru1 ,6bisP and stimulating the activation, e.g. by irreversible binding and constitutive activation. Preferably, said modulation is an inhibition.

Still another aspect of the invention is a modulator of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation for use as a medicament. In one preferred embodiment, said modulator is a Fru1 ,6bisP mimicking compound that can be used to induce programmed cell death. In another preferred embodiment, said modulator is an inhibitor of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation. Even more preferably said modulator of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation is for use in treatment of cancer. Both inhibition of the complex formation (inhibiting the Ras activation) as well as the hyperactivation of Ras, e.g. by irreversible binding of a compound on the guanine nucleotide exchange protein can be used in treatment of cancer. Indeed, hyperactivation will lead to programmed cell death of the cancer cells, as indicated in the examples. Preferably, said modulator of the fructose-1 ,6- bisphosphate / guanine nucleotide exchange protein complex formation is an inhibitor of the complex formation.

Modulators of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation are, as a non-limiting example peptides and peptidomimetics comprising one or more of the sequences given in Figure 7a, antibodies, nanobodies, alphabodies or other binding proteins that bind on the interaction interface, especially those binding to the sequences given in figure 7a - all those modulators will act as inhibitors. Other modulators of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation are analogs of fructose-1 ,6-bisphosphate, such as, but not limited pyridoxal-5-phosphate, mannitol-1 ,6-bisphosphate, D-ribulose-1 ,5-diphosphate, tagatose-1 ,6-diphosphate, L-sorbose- 1 -phosphate and gluconic acid 6-phosphate. Preferably, said modulator is a monophosphate compound. Examples of irreversible binders are 2-keto-3-butenyl-phosphate and 2-keto-4,4,4- trifluorobutyl-phosphate.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 : Glucose triggers activation of Ras and apoptosis in tpslA cells, a, Ras-GTP level before (gal) and 10 min after addition of glucose (glu) to cells grown on galactose (gal). The strain expressing Ras2^val19 was grown on glucose. Total Ras = Ras-GDP + Ras-GTP. b, Specific cytochrome c release from the mitochondria after addition of glucose to wild type and tpslA strain. M, mitochondrial fraction; SN, post-mitochondrial supernatant; Anti-CyC, anti- cytochrome c antibody; Anti-Cox2, cytochrome oxidase 2 antibody; C: control cytochrome c or cytochrome oxidase 2. c, Annexin-V-FITC/propidium iodide staining 2h after addition of glucose to cells grown on galactose. Green colored fluorescence originates from annexin staining, red colored fluorescence (less bright in the annexin panels) represents propidium iodide binding to DNA of death cells, d, Reactive oxygen species (ROS) accumulation 2 h after addition of glucose to cells grown on galactose. ROS was stained with the dye dihydrorhodamine 123 (indicated as DHR 123). DIC, differential interference contrast image. Figure 2: Restoration of growth on glucose and glycolytic metabolite levels in the tpslA strain with downregulation of the Ras-cAMP-PKA pathway, a, Growth of tpslA and suppressor strains on galactose and different concentrations of glucose. Five-fold dilutions were spotted, b, Intracellular Glu6P, Fru1 ,6bisP and ATP concentration before and after addition of 100 mM glucose. The short-term measurements (0-10 min) were performed with non-growing cells incubated in buffer, while the long-term measurements (0-30h) were performed with cells growing in minimal medium.

Figure3: Growth of the tpslA strain and tpslA suppressor strains in rich liquid medium with 100 mM galactose or 100 mM glucose.

Figure 4: Activation of Ras by Fru1 ,6bisP. a, Ras-GTP level 10 min after addition of 5 mM of different glycolytic metabolites to permeabilized yeast spheroplasts. Glu6P, glucose-6- phosphate; Tre6P, trehalose-6-phosphate; Fru6P, fructose-6-phosphate; Fru1 ,6bisP, fructose- 1 ,6-bisphosphate; DHAP, dihydroxyacetone phosphate; GAP, glyceraldehyde-3-phosphate; 3PG, 3-phosphoglycerate; PEP, phosphoenolpyruvate. b, Activation of Ras in permeabilized spheroplasts with different concentrations of GTP, Fru1 ,6bisP, DHAP and GAP. c, Ras-GTP level before and 10 min after addition of 100 mM glucose (glu) to cells grown on glycerol + 2.5 mM glucose (gly). d, Intracellular ATP, Glu6P, and Fru1 ,6bisP concentration before and after addition of 100 mM glucose to cells grown in glycerol + 2.5 mM glucose.

Figure 5: Evaluation of requirement for the Ras regulators Cdc25, Sdc25 and Ira1 , Ira2. a, Growth on galactose and glucose. Five-fold dilutions were spotted, b, Ras-GTP level before and 10 min after addition of glucose (glu) to cells grown on galactose (gal).

Figure 6: Growth of the tpsIA strain and tpsIA strains with additional deletions in Ras-GEF or Ras-GAP factors, in rich liquid medium with 100 mM galactose or 100 mM glucose.

Figure 7: Conserved residues in the C-terminus of Cdc25 and Sos1 are required for Fru1 ,6bisP activation of Ras. a, Alignment of the conserved region in the Ras GEF factors, HsSosl and HsSos2 from Homo sapiens and ScCdc25 and ScSdc25 from Saccharomyces cerevisiae. b,d, Ras-GTP level 3 min after addition of 5 mM Fru1 ,6bisP or 1 mM GTP as positive control, to permeabilized yeast spheroplasts from strains expressing mutant ScCdc25 (b) or HsSosl (d) alleles. Western blots were quantified and the Ras-GTP levels are shown relative to the level observed after addition of GTP. Dunnett's test was used for evaluation of statistical significance, c, Recovery of growth on glucose in the tpsI A sdc25A strain expressing Cdc25R1 122D,K1491 E instead of wild type Cdc25. e, Crystal structure of the complex between human Sos1 (middle, left) and human Ras (middle, right) (PD entry: 1 BKD; Boriack- Sjodin et al., 1998). The arrow points to the hypothetical binding cleft for Fru1 ,6bisP, which contains the conserved region shown in (a). Enlarged view on the 3D-structure surrounding the relevant residues in Sos1 and Cdc25 located in the hypothetical binding cleft. Putative docking of Fru1 ,6bisP using the 3D structure (1 BKD) and the SwissDock program (Grosdidier et al., 201 1 ).

Figure 8: Phenotypes of a tpsIA strain expressing Cdc25T1490P instead of wild type Cdc25. a, Absence of recovery of growth on glucose, b, Reduced trehalose and glycogen levels in cells grown on glycerol into stationary phase. EXAMPLES

Materials and methods to the examples

Yeast strains and plasmids

In this work, all yeast strains used share the same genetic background, i.e. W303. All strains are listed in Table 1 . The YEplac195 expression vector contains the promoter and terminator sequences of the 5' and 3' regions of the PDE2 gene. The YCpRas2^val19 plasmid contains promoter and terminator sequences of the 5' and 3' regions of the RAS2^val19 gene.

Table 1 : list of strains

Yeast strains Relevant genotype Reference

W303-1A (wild type) Mata Ieu2-3, 112 ura3-1 trp1-1 his3- (Thomas & Rothstein,

11, 15 ade2-1 can1-100 GAL SUC 1989)

YSH290 tps1::TRP1 (Hohmann et al., 1993)

YSH312 tps1::TRP1 hxk2::LEU2 (Hohmann et al., 1993)

JT22287 tps1::TRP1 ras2::KanMX This work

KEN9 ira 1 :: Kan MX ira2::KanMX This work

KEN22 tps1::TRP1 pfk1::KanMX This work

KEN24 tps1::TRP1 pfk2::KanMX This work

KEN25 tps1::TRP1 pfkl:: KanMX This work

pfk2::KanMX

KEN27 tps1::TRP1 iral:: KanMX This work

ira2::KanMX

KEN29 ira 1 :: KanMX ira2::KanMX This work

cdc25::LEU2

KEN31 tps1::TRP1 iral:: KanMX This work

ira2::KanMX cdc25::LEU2

KEN33 ira 1 :: KanMX ira2::KanMX This work

sdc25::HIS3

KEN35 tps1::TRP1 iral:: KanMX This work

ira2::KanMX sdc25::HIS3

KEN37 ira 1 :: KanMX ira2::KanMX This work

cdc25::LEU2 sdc25::HIS3

KEN39 tps1::TRP1 iral:: KanMX This work

ira2::KanMX cdc25::LEU2

sdc25::HIS3 Yeast strains Relevant genotype Reference

KEN46 cdc25::CDC25^T1490P This work

KEN50 cdc25::CDC25^K1491E This work

KEN52 cdc25::CDC25^R1122D This work

KEN53 cdc25::CDC25^K1491ER1122D This work

KEN58 cdc25::SOS1553-1024 This work

KEN59 cdc25::SOS1₅53-io24^R962T This work

KEN61 cdc25::SOS1₅53-io24^R962P This work

KEN64 cdc25::SOS1₅53-io24^K963E This work

KEN67 cdc25::SOS1₅53-io24^K602E This work

KEN68 cdc25::SOS1₅53-io24^{K602E K963E} This work

Growth conditions

When no selection was required, yeast cells were grown in rich YP (1 % (w/v) yeast extract, 2% (w/v) bactopeptone) medium, supplemented with 100 mg/l adenine and 2% (w/v) glucose, 2% (w/v) galactose, 1 % (v/v) ethanol or 3% (v/v) glycerol. In case of selection for the presence of an auxotrophic marker either on a plasmid or in the genome, cells were grown in synthetic medium containing 0.17% (w/v) yeast nitrogen base without amino acids, 0.5% (w/v) ammonium sulfate, the appropriate amount of the required synthetic 'drop out' amino acid and nucleotide mixture and supplemented with 2% (w/v) glucose, 2% (w/v) galactose, 1 % (v/v) ethanol or 3% (v/v) glycerol. The pH was adjusted with 4 M KOH to 5.5 for liquid medium and 6.5 for solid medium. For solid medium, 1.75% (w/v) agar was added after pH adjustment. Strains were grown at 30°C. Liquid cultures were grown under continuous shaking (180 rpm).

Expression and purification of RBD-GST from E. coli

Ras-binding domain of the Raf1 -kinase (RBD) GST-fusion protein was constructed in the pGEX2T-1 expression vector and expressed in the E. coli BL21 strain. Cells were grown for 1 h at 37°C prior to addition of IPTG to a final concentration of 0.3 mM to induce gene expression from the pGEX2T-1 plasmid. Induction lasted for 3 h at 30°C. Cells were harvested and washed with ice-cold PBS buffer. Cells were resuspended in 5 ml per I culture lysis buffer containing 50mM Tris-HCI pH 7.5, 50mM NaCI, 2mM EDTA pH 8, 1 mM EGTA pH 8, 20% sucrose and 5 mg/ml lysozyme at a concentration of 10 ml per I of culture and incubated for at least 15 min. 3 volumes of ice-cold buffer containing 50 mM Tris-HCI pH 7.5, 150 mM NaCI, 1 % Triton X-100 and protease inhibitor cocktail (complete, EDTA-free, Roche) was added. Cells were lysed by sonicating three times for 15 s. Lysates were clarified by centrifugation at 20,000 rpm for 25 min at 12,000 x g. The resulting supernatant fraction was incubated with 2 ml per I of initial culture of a 50-50 slurry of glutathione-sepharose beads (GE Healthcare) prewashed with 1 x TBST (50 mM Tris-HCI pH 7.5, 150 mM NaCI, 1 % Triton X-100 and 1 mM DTT) and incubated for 1 h in a rollerdrum at 4°C. Beads were collected by centrifugation at 1800 rpm for 1 min and washed 4 times with TBST. Beads were resuspended in TBS (TBST minus Triton) at a final concentration of 10 %. The addition of 0.02% NaN₃ allowed keeping this solution for up to 6 months.

Preparation of a yeast crude cell extract

Cells were grown to an OD of 3-4 and collected on a microfilter connected to a vacuum pump. After washing with ice-cold water, 200 mg cells were collected in a screw cap tube. 0.2 g of glass beads were added together with 500 μΙ of lysis buffer containing 50 mM Tris-HCI pH 7.5, 10% glycerol, 2.5 mM MgCI₂, 1 % Igepal (NP-40) and 200 mM NaCI. The cells were broken by shaking three times 20 s in a fast prep apparatus. A clarified crude protein extract was obtained after 5 min centrifugation at 8000 rpm. Protein concentration was measured by mixing 20 μΙ of sample together with 300 μΙ of the PIERCE 660 nm protein measurement buffer and absorbance was measured at 660nm using a bovine serum albumin (BSA) standard curve. Generation of permeabilized spheroplasts

The protocol for the generation of spheroplasts was adapted from a previous report(Averet et al., 1998). For Ras activation in permeabilized spheroplasts, 200 mg cells were used per sample. Strains were grown in a culture volume based on the number of samples to be measured. The cells were washed with water and resuspended in digestion buffer containing 1 .2 M sorbitol, 60 mM potassium phosphate buffer pH 7.5 (K₂HP0₄ + KH₂P0₄), 1 mM EDTA pH 8, 10 μΙ/ml β-mercaptoethanol and 100 units/ml lyticase at a concentration of 5 mg/ml. Digestion of the cell wall was achieved by incubation in a water bath at 30°C. The digestion was followed by the decrease in OD at 660 nm when spheroplasts are broken in 5% SDS (10 μΙ spheroplast suspension + 990 μΙ SDS). When the OD decreased with 80% in this assay, the spheroplasts were considered appropriate for use. Spheroplasts were washed two times with 1 .2 M sorbitol and resuspended in spheroplast buffer containing 1 .2 M sorbitol, 0.75 mM EDTA pH 8, 2 mM MgS0₄, 1.8 mM NaCI and 10 mM potassium phosphate buffer pH 6.8. A sample was taken of the spheroplast suspension to measure protein concentration. For that purpose, 10 μΙ was diluted and broken by vortexing in 90 μΙ of water. 20 μΙ of the resulting solution was mixed with 300 μΙ of the PIERCE 660 nm protein measurement buffer and the protein concentration was measured. The spheroplasts were then diluted with spheroplast buffer till a protein concentration of 2 mg/ml. To permeabilize the spheroplasts, they were incubated for 10 min at 30°C with 20 μΙ/ml nystatin (from a 2 mg/ml DMSO stock solution). Ras activation assay

Two different Ras activation assays were used. Ras activation was either studied in vivo upon addition of glucose to galactose-grown cells, or in situ, upon addition of different glycolytic intermediates, which cannot be taken up by intact cells, to permeabilized spheroplasts.

Ras activation as determined with samples taken in vivo

For the in vivo Ras activation assay, glucose was added to galactose-grown cells and the cells were incubated further for 10 min. Crude cell extracts were prepared and a volume containing 2 mg of total protein was added to the beads bound to RBD-GST and incubated for 1 h at 4°C on a rollerdrum. After incubation, the beads were washed three times with the lysis buffer used to make the crude extracts. After washing, the beads were dried by removing all liquid with a narrow orifice 30g needle. To the dried beads, 20 μΙ of 2x loading buffer was added. This buffer contains 100 mM Tris-HCI pH 8, 20 mM β-mercaptoethanol, 4% SDS, 0.2% bromophenol-blue and 20% glycerol. This solution was boiled for 5 min at 96°C. A volume with 20 μg of total protein was used for the control in which the total Ras content was determined and an equal amount of 2x loading dye was added and boiled. The samples were then ready for western blotting.

Ras activation as determined with samples taken in situ (in permeabilized spheroplasts) Glycolytic metabolites and GDP were dissolved in H₂0 to a final concentration of 100 mM, except glyceraldehyde-3-phosphate, which was dissolved in a potassium phosphate buffer (because it is only available as a very acidic liquid solution). GTP was dissolved in H₂0 to a final concentration of 20 mM. 50 μΙ of the stock solutions was added to 1 ml permeabilized spheroplasts to obtain a final concentration of 5 mM for the metabolites and GDP and a final concentration of 1 mM for GTP. The cells were incubated for 3 min at 30°C and after incubation snap frozen in liquid nitrogen. 330 μΙ of a 4x lysis buffer was added to the frozen spheroplasts. Spheroplast lysis was achieved by simple vortexing and the supernatant was cleared by centrifugation at 8000 rpm for 5 min. Crude extracts were incubated with the RBD- GST beads.

Western blotting

All samples were boiled for 5 min at 96°C and briefly centrifuged before loading on the electrophoresis gel. Proteins were separated via SDS-PAGE (NuPAGE 4-12% Bis-Tris gel,

Invitrogen) in NuPAGE MOPS SDS running buffer (Invitrogen) at a constant voltage of 150-

200V. After electrophoresis, proteins were transferred onto nitrocellulose membranes (HybondC extra, GE Healthcare) by blotting for 60 min at 400 mA in blotting buffer (NuPAGE MOPS SDS running buffer, with 20% (v/v) methanol). Aspecific antibody binding was prevented by incubating the membrane in blocking buffer [(5% (w/v) BSA in TBS-Tween buffer (25 mM Tris-HCI pH 8, 150 mM NaCI, 0.05% (v/v) Tween-20)] for 1 h at room temperature. Subsequently, the blots were incubated overnight at 4°C in blocking buffer with the appropriate anti-Ras2 antibody (Sc-6759, Santa Cruz). Subsequently, the blots were washed three times with TBS-Tween buffer and incubated with the appropriate secondary antibody (Sc-2020, Santa Cruz) in TBS-Tween. After washing three times with TBS-Tween, the membranes were incubated with Supersignal chemiluminescence substrate (Pierce). Proteins were visualized by exposure of the membrane in the LAS4000 mini digital system (Fujifilm). Signals were quantified with Aida software.

Metabolite determinations

For short-term (up to 10 min) measurement of the intracellular concentration of metabolites, cells were grown in YP glycerol supplemented with 0.1 % glucose till an OD of 2-3. Cells were harvested by centrifugation and washed two times with 0.5 M β-morpholinoethanesulfonic acid (MES) pH 6. After the last wash step, the cells were dried and weighed. The cells were then suspended at a concentration of 75 mg/ml MES buffer and incubated in a water bath at 30°C. Samples were taken by quenching 2 ml of cells in 10 ml of 60% methanol at - 40°C. For metabolite determination after glucose addition, 0.5 M glucose was added to the cells and samples were taken as a function of time by taking 2.5 ml cells and quenching them in ice-cold methanol. The cells were centrifuged at -20°C, at 3000 rpm for 3 min. The supernatant was removed completely by the use of a vacuum pump and resuspended in 500 μΙ of 1 M perchloric acid (HCI0₄). The cells were transferred to a screw cap tube and broken with glass beads in the fast prep apparatus (two times 20 s at speed setting of 6). 500 μΙ of 1 M HCI0₄ was added again and the mixture spun down for 4 min at 12,000 rpm, at 4°C. 250 μΙ was mixed with 10 μΙ thymol blue (25 mg/100 ml) and 50 μΙ of 5 M K₂C0₃. The tubes were left open for at least 15 min to make sure all C0₂ was released. 200 μΙ was transferred to a new test tube and mixed with 100 μΙ of 1 M HCI and 10 μΙ of 2 M Tris-HCI pH 7.5 (de Koning & van Dam, 1992).

For long-term (up to 30h) measurement of the intracellular concentration of different metabolites, cells were grown in YP glycerol supplemented with 0.1 % glucose till an OD of 2-3. 2% glucose was added and samples were taken by filtration and the dry weight was determined. 500 μΙ of 1 M perchloric acid was added to the dried cells and the extract was treated as described above. Glucose-6-phosphate and ATP

50 μΙ of the sample was mixed with 150 μΙ of reaction buffer containing 100 mM Hepes pH 7.6, 0.8 mg/ml NADP and 10 mM MgCI₂. To measure glucose-6-phosphate the enzyme glucose-6- phosphate dehydrogenase (G6PDH) was added in a concentration of 50 μg ml. The NADPH formed in this reaction was measured at a wavelength of 340 nm. The amount of NADPH determined is directly proportional to the amount of glucose-6-phosphate present in the sample. For ATP determination, hexokinase at a concentration of 100 μg ml was added together with 0.5 mM glucose. Hexokinase will produce glucose-6-phosphate with the glucose added and the ATP present in the sample and the glucose-6-phosphate will be converted again by G6PDH with the generation of NADPH, that is measured and is directly proportional to the ATP concentration present in the sample.

Fructose- 1, 6-bisphosphate

50 μΙ of the sample was mixed with 150 μΙ of reaction buffer containing 100 mM Hepes pH 7.6 and 0.3 mM NADH, 2.5 μg ml glyceraldehyde-3-phosphate dehydrogenase and 2.5 μg ml triose phosphate isomerase. For measurement of fructose-1 , 6-bisphosphate, aldolase was added at a concentration of 100 μg ml and the consumption of NADH was measured at a wavelength of 340 nm.

Docking experiments

Docking calculations were executed with SwissDock, a protein-small molecule docking web service based on EADock DSS (Fast docking using the CHARMM force field with EADock DSS) (Grosdidier et al., 201 1 ). SwissDock, whose algorithm consists of the following four steps: i) many binding modes are generated either in a box (local docking) or in the vicinity of all target cavities (blind docking); ii) simultaneously, their CHARMM energies are estimated on a grid; iii) the binding modes with the most favorable energies are evaluated with FACTS, and clustered; iv) the most favorable clusters can be visualized online and downloaded. UCSF Chimera (a visualization system for exploratory research and analysis) was used for molecular structure viewing (Pettersen et al., 2004). The crystal structure was obtained from the PDB with the accession code 1 BKD(Boriack-Sjodin et al., 1998). Fructose-1 , 6-bisphosphate structure was obtained by the Zinc Data base (Irwin & Shoichet, 2005) (ZINC3869916).

Reactive oxygen species

The production of reactive oxygen species (ROS) was assessed using dihydrorhodamine123 (DHR)(Sigma-Aldrich). Cells were grown in YPGal medium until mid-exponential phase, a sample was collected and 2% glucose was added after which a new sample was taken. 5 μg ml DHR was added to the cells and incubated for 2 h. ROS production was visualized using fluorescence microscopy (Axioplan 2 imaging, Zeiss).

Annexin staining of phosphatidylserine exposure

Cells were stained using the Annexin-V-FLUOS Staining Kit (Roche). After growing cells till mid exponential phase in YPGal medium, 2% glucose was added. After 16 h the cells were collected, washed with PBS buffer and centrifuged for 5 min at 200 x g. Annexin-V-FLUOS labeling solution was prepared according to the instructions of the manufacturer: 1 ml of incubation buffer and 20 μΙ of Annexin-V-FLUOS labeling reagent and 10 μΙ of propidium iodide solution. Washed cells were incubated in 100 μΙ of this labeling solution for 15 min after which the cells were observed in the fluorescence microscope. Annexin-V-FLUOS excitation was performed at 488 nm and emission was measured at 518 nm. Propidium iodide excitation was performed at 488-540 nm and emission was measured at 617 nm.

Isolation of mitochondria for determination of cytochrome c release

The procedure was adapted from a previous report (Herrmann et al., 1994). A 1 I culture was used for each sample. 2% glucose was added to a 1 I culture after which the culture was incubated for the indicated time. The cells were harvested by centrifugation at 3000 rpm for 10 min and washed two times with 1.2M sorbitol. The cells were then resuspended in 5 ml of digestion buffer per g cells. The digestion buffer contained 1 .2 M sorbitol, 60 mM potassium phosphate buffer pH 7.5, 1 mM EDTA pH 8, 1 % β-mercaptoethanol (w/v) and 1 % zymolyase 20 T. The mixture was incubated for 1 h at 30°C. The spheroplasts formed were collected by centrifugation for 5 min at 1500 x g, washed two additional times with 1 .2 M sorbitol and resuspended in 3 ml of ST-Buffer, containing 0.25 M sorbitol, 20 mM Tris-HCI pH 7.5, 1 mM EDTA and protease inhibitor cocktail (Roche), per g of cells. The suspension was incubated with agitation at 4°C until the spheroplasts were osmotically broken. Cell debris were removed by centrifugation at 2000 x g for 5 min and the supernatant was again centrifuged at 12,000 x g for 15 min to precipitate mitochondria. Both the post-mitochondrial supernatant (PMS) and the pellet (the mitochondrial pellet was resuspended in 0.5 M sorbitol) were used to detect cytochrome c by Western blot analysis.

Screening of Fru1,6bisP mimicking compounds

Precultures were prepared by growing a single colony of tpsIA strain under continuous shaking (250 rpm) at 30°C in 50mL liquid YP galactose medium, containing 1 % (w/v) yeast extract, 2% (w/v) bactopeptone, 100mM galactose and 100mg/L adenine.

Assay plates (clear flat bottom 96-well plates from BD Falcon) were prepared using a 40x dilution of the preculture described above. Cells were harvested (2000 rpm, 5 min) and the pellet was resuspended in YP 1 mM glucose (1 % (w/v) yeast extract, 2% (w/v) bactopeptone, 1 mM glucose and 100mg/L adenine). For the positive control, a fraction of the same preculture was also centrifuged and resuspended in YP 5mM glucose (1 % (w/v) yeast extract, 2% (w/v) bactopeptone, 5mM glucose and 100mg/L adenine).

Each 96-well assay plate harbors 8 growth reference controls (150μΙ_ of tpsIA strain cells in YP 1 mM glucose) and 8 positive controls (no growth; 150μΙ_ of tpsIA strain cells in YP 5mM glucose). The remaining 80 wells were filled with 150μΙ_ of tpsIA strain cells in YP 1 mM glucose, to which the compounds will be added. For this 96-well plate preparation, the Biomek 2000 liquid handling station (Beckman Coulter) was used.

Next, 1 .5μΙ_ of 5mM compound solutions (dissolved in DMSO) were added to each of the 80 sample wells, so that each assay plate was used to test 80 different compounds. To the 16 control conditions, 1.5μΙ_ of DMSO was added. For addition of the compounds and DMSO, the Tecan freedom evo 200 liquid handling platform was used. Following addition, the solution was mixed by pipetting (5x). Immediately after, OD600 was measured, which served as a blank measurement.

The plates were then incubated under continuous shaking at 30°C. After 24h, the cultures were resuspended by pipetting (15x) using the Tecan Freedom Evo 200 platform (Tecan). Immediately after the resuspension, OD600 was measured.

First the blank measurement of OD600 at Oh was subtracted from the OD600 measure at 24h. Next, the corrected value was normalized with the average of the growth reference control (1 mM glucose + DMSO) for each plate. Growth inhibition was scored; the cutoff for hit selection was set at a difference of 25% of control growth. Growth on galactose, in presence of the compounds is used as a control to exclude general (non-glucose induced) toxicity.

Screening of Fru1 ,6bisP complex inhibitors

The screening was carried out essentially in the same way as the one for the Fru1 ,6bisP mimicking compounds, but 5mM glucose was used instead of 1 mM glucose, to ensure complete growth arrest, and growth in presence of the compounds was scored as indication for complex inhibiting activity

Example 1 : Addition of glucose to galactose-grown tpsIA cells activates Ras and causes programmed cell death

Addition of glucose to galactose-grown tpsIA cells triggers rapid and dramatic activation of Ras, as opposed to the modest activation observed in wild type cells (Fig. 1 a). The GTP loading state on Ras in the tpsIA strain, 10 min after glucose addition, is similar to that in the constitutively active Ras2^val19 oncogene equivalent (Fig. 1 a). A Ras2^val19 strain, however, is not sensitive to glucose (Broek et al., 1985). Deletion of hexokinase 2 abolishes glucose-induced hyperactivation of Ras (Fig. 1 a), which is consistent with its suppression of the glucose growth defect of the tpsIA strain (Hohmann et al., 1993). Whereas the Ras2^val19 strain shows normal growth on glucose, it grows badly on poor carbon sources and dies in stationary phase (Broek et al., 1985). More recent work has shown that hyperactive Ras causes programmed cell death in non-growing yeast cells (Gourlay & Ayscough, 2006). Hence, we reasoned that the combination of aberrant growth initiation due to the glycolytic deregulation and hyperactivation of Ras in tpsIA cells may cause programmed cell death and that this actually may be the real cause for the unusually high glucose sensitivity of tpsIA cells (Neves et al., 1995) rather than the deregulation of glycolysis per se. Determination of classical read-outs for programmed cell death/apoptosis in yeast gave results consistent with this hypothesis. Glucose addition to tpsIA cells, as opposed to wild type cells, caused rapid release of cytochrome c from the mitochondria (Fig. 1 b), exposure of phosphatidylserine at the plasma membrane (Fig. 1 c) and generation of reactive oxygen species (ROS) (Fig. 1 d).

Deletion of Ras2 in the tpsIA strain caused partial recovery of growth on glucose (Fig. 2a). Overexpression of PDE2, which encodes the high-affinity cAMP phosphodiesterase and is known to counteract phenotypes caused by hyperactive Ras (Sass et al., 1986), restored growth on glucose even better (Fig. 2a). These results are consistent with overactive Ras playing a role in the glucose growth defect of the tpsIA strain. Interestingly, in both strains a very similar deregulation of glycolysis after addition of glucose was observed, including hyperaccumulation of Fru1 ,6bisP and depletion of ATP, but the metabolite profile started to recover after about 5 h to finally reach the same profile as observed in wild type cells (Fig. 2b). The difference was most striking in the ATP level, which stayed close to zero in the tpsIA strain, but recovered to the same level as in wild type cells in the tpsIA ras2A and tpsIA pPDE2 strains (Fig. 2b, results on solid medium; Fig.3 results in liquid medium). This observation reveals a remarkable robustness and flexibility of the glycolytic pathway in yeast. The results suggest that the induction of programmed cell death, caused by overactivation of the Ras-cAMP-PKA pathway, is the true cause of the inability of the tpsIA strain to grow on glucose, rather than the glycolytic deregulation.

Example 2: Hyperactivation of Ras is mediated by Fru1 ,6bisP

To investigate whether one or more of the hyperaccumulated glycolytic metabolites in the tpsIA strain acted as the trigger for hyperactivation of Ras, we made use of nystatin- permeabilized wild type yeast spheroplasts (Noubhani et al., 2000). Because of their fragility, they were used as such, i.e. the protein-containing permeabilized cell sacs were not washed. The small molecules leaked into the medium were therefore very strongly diluted. Addition of individual glycolytic metabolites to nystatin-permeabilized spheroplasts revealed strong activation of Ras by Fru1 ,6bisP at the higher physiological concentrations (4-8 mM) (Fig. 4a, b), similar to that observed with physiological concentrations of GTP (Fig. 4b). DHAP and GAP also caused activation but only in unphysiologically high concentrations (1 -5 mM) (Fig. 4a, b). Other glycolytic intermediates, glucose and Tre6P did not cause a significant activation (Fig. 4a).

To evaluate whether Fru1 ,6bisP is responsible for glucose-induced hyperactivation of Ras in the tpsIA strain, we constructed tpsIA strains with additional single or double deletion of the PFK1 and PFK2 genes, encoding phosphofructokinase 1. Glucose-induced activation of Ras was reduced in cells of the tpsIA pfkIA and tpsIA pfk2A strains, and abolished in cells of the tpsIA pfkIA pfk2A strain (Fig. 4c). This would be consistent with the hyperaccumulation of Fru1 ,6bisP in tpsIA cells being responsible for hyperactivation of Ras. However, upon measurement of the glycolytic metabolites, the tpsIA pfkIA and tpsIA pfk2A strains were already deficient in glucose-induced hyperaccumulation of Fru1 ,6bisP and consistently showed a much higher accumulation of Glu6P than the tpsIA strain (Fig. 4d). Glu6P was also higher than in the corresponding pfkIA and pfk2A strains (Fig. 4d).

The presence of partial Ras activation without Fru1 ,6bisP accumulation in the tpsIA pfkIA and tpsIA pfk2A strains suggests that there may be a second mechanism responsible for glucose- induced hyperactivation of Ras in the tpsIA strain and/or that the huge hyperaccumulation of Glu6P in these strains in some way triggers a residual activation. Although the complete absence of Ras hyperactivation in the tpsIA pfkIA pfk2A strain seems to contradict this possibility, the very poor growth of this strain, requiring feeding with both a respiratory carbon source and a low level of glucose, may have compromised the second activation mechanism.

Example 3: Fru1 ,6bisP activates Ras via a Fru1 ,6bisP /guanine nucleotide exchange protein complex

We next investigated whether Fru1 ,6bisP caused direct activation of Ras or acted through one of its regulators, the guanine nucleotide exchange proteins Cdc25 and Sdc25, or the GTPase activating proteins Ira1 and Ira2. For that purpose we constructed several strains with combinations of tpsIA and deletions in the genes encoding these regulators. Deletion of

CDC25 and SDC25 is lethal (Boy-Marcotte et al., 1996), while deletion of IRA 1 and IRA2 causes hyperactivation of Ras (Tanaka et al., 1990). The iralA ira2A cdc25A sdc25A strain retains enough Ras activity for growth (Fig.5a). All quadruple tpsIA strains with at least one wild type GEF factor, Cdc25 or Sdc25, were unable to grow on glucose, whereas absence of both GEF factors in the quintuple deletion strain tpsIA cdc25A sdc25A iralA ira2A restored to some extent growth on glucose medium, (Fig. 5a, Fig. 6). This suggests that glucose-induced hyperactivation of Ras through the GEF factors is a major cause of the glucose sensitivity of the tpsIA strain. The tpsIA IralA ira2A strain showed already a high Ras-GTP level on galactose medium, which made it difficult to evaluate any further glucose-induced increase (Fig. 5b). On the other hand, in the quadruple deletion strains, tpsIA sdc25A iralA ira2A and tpsIA cdc25A iralA ira2A, the rapid glucose-induced hyperactivation of Ras was largely abolished, indicating requirement of both GEF factors. The absence of Ras hyperactivation in these strains may be restricted to the short-term response, since both strains were still unable to grow on glucose (Fig.5a). Unexpectedly, we found that the quintuple deletion strain, tpsIA cdc25A sdc25A iralA ira2A, showed a high basal activity of Ras already on galactose medium, which not increased further upon addition of glucose (Fig. 5b). Since galactose metabolism does not involve hexokinase activity, a high Ras-GTP level should not cause a growth defect on galactose medium in the tpsIA cdc25A sdc25A iralA ira2A strain and other tpsIA strains with high basal Ras activity. Although these experiments and the interpretation of the results are complicated because of the essential character of the GEF factors, the high basal Ras activity caused by deletion of the GAP proteins, and the unexpectedly high basal level of Ras- GTP in the quintuple deletion strain, the results are consistent with a role of the GEF factors in mediating glucose-induced hyperactivation of Ras in the tpsIA strain. This was confirmed by the observation that deletion of SDC25 and especially CDC25 in the tpsIA strain caused partial growth recovery on glucose. This result also suggests that Fru1 ,6bisP exerts its activating effect mainly through Cdc25.

Example 4: Fru1 ,6bisP interacts with the conserved C-terminus of the guanine nucleotide exchange protein

Since previous work had shown that the strongly-conserved C-terminus of Cdc25 was required for glucose activation of cAMP synthesis (Schomerus et al., 1990; Van Aelst et al., 1990), we concentrated on this part of Cdc25 and noticed that it contained a strongly-conserved region with several positively charged residues (Fig.7 a). This suggested that this area could be involved in the formation of a binding site for the negatively charged Fru1 ,6bisP molecule. Interestingly, another positively-charged residue, R1 122 in yeast Cdc25, which corresponds to K602 in human Sos, resides in the vicinity of this region in the 3D-structure of the human Sos catalytic domain (Boriack-Sjodin et al., 1998). R1 122/K602 is located itself in a small conserved region. We first mutagenized the positively charged residues R1 122 and K1491 in the putative

Fru1 ,6bisP binding site, to negatively charged aspartate and glutamate residues, respectively.

In addition, we constructed the T1490P allele in order to disturb the structure of the alpha helix in which the putative binding site is located (Schomerus et al., 1990). In permeabilized spheroplasts of strains expressing only Cdc25^{R1122D K1491E} or Cdc25^T1490P (and no wild type Cdc25), addition of Fru1 ,6bisP no longer caused activation of Ras as opposed to addition of GTP (Fig. 7b). This suggests that the residues R1 122 and K1491 may be involved in the interaction with Fru1 ,6bisP and that the proper structure of the alpha helix surrounding T1490 is important for proper interaction with Fru1 ,6bisP. The activity of the other yeast guanine nucleotide exchange protein Sdc25 was assumed negligible during exponential growth on rich media, since it has been reported to be expressed only during late stationary phase and on non-fermentable carbon sources (Boy-Marcotte et al., 1996). The strains carrying the Cdc25 mutated alleles in the absence of Sdc25 were viable, which together with the remaining activation by GTP, shows that the alleles were functional. Interestingly, a tpsIA sdc25A strain expressing Cdc25^{R1122D,K1491E} instead of wild type Cdc25 displayed strong recovery of growth on glucose, supporting that these Cdc25 residues also mediate Fru1 ,6bisP activation of Ras in vivo (Fig. 7c). We did not observe such recovery of growth on glucose with a strain expressing Cdc25^T1490P instead of wild type Cdc25 (Fig. 8a). However, such a strain displayed reduced trehalose and glycogen levels in stationary phase after growth on glycerol (Fig. 8b). This suggests the presence of constitutively activated PKA, which may be explained by Thr¹⁴⁹⁰ serving as a PKA target site for feedback inhibition of PKA on Cdc25. Next, we assessed whether human Sos expressed in yeast is also responsive to the same type of glycolytic stimulation as yeast Cdc25. Interestingly, addition of Fru1 ,6bisP to permeabilized yeast spheroplasts expressing only the guanine nucleotide exchange factor region of human Sos (residues 553-1024), caused a clear activation of Ras similar to that of GTP (Fig. 7d). Wild type Sos has the INFSKRR962K sequence. However, the yeast strain expressing the complete Sos protein grew poorly on galactose. The mutant protein R962T, which has the same INFSKRTK sequence as in Cdc25, also supported activation of Ras with Fru1 ,6bisP and GTP (Fig. 7d). The strain expressing this protein showed normal growth on galactose. Activation with Fru1 ,6bisP, but not GTP, was absent for the mutant alleles Sos^K602E' ^K963E and Sos^R962P (Fig. 7d), which have the equivalent mutations to those mentioned above for yeast Cdc25 (Boriack-Sjodin et al., 1998). With Sos^K963E, which is the equivalent of Cdc25^K1491E, significant activation with Fru1 ,6bisP could not be demonstrated (Fig. 7d). These results show that Sos is able to mediate activation of Ras by Fru1 ,6bisP and that the equivalent domain and corresponding amino acid residues to those in Cdc25 are important for the activation. Fig. 7e shows the 3D structure of Sos in association with Ras and the position in space of selected amino acid residues in the cleft formed by the alpha helix I of Sos and the switch 1 region of Ras, thus in the area where the two proteins interact. The amino acid residues K602, R962 and K963 of Sos, are located within this cleft (Fig. 7e). We also suggest a putative model for the positioning of Fru1 ,6bisP within this site, interacting with its negatively charged phosphate residues with the positively charged residues K963 and K602 in Sos, and K1491 and R1 122 in Cdc25. How interaction of Fru1 ,6bisP with Cdc25/Sos triggers activation of Ras is unclear at this moment. It might stimulate the binding of Cdc25/Sos with Ras, facilitating exchange of GDP for GTP or alternatively shielding Ras from the action of the GTPase activating proteins Ira1 ,2/NF1.

It has been reported that Ras activation stimulates glycolytic rate in tumor cells (Kole et al., 1991 ; Mazure et al., 1997; Telang et al., 2006). The opposite signaling effect that we describe here, i.e. Ras activation by high glycolytic rate through Fru1 ,6bisP, may close a reciprocal stimulatory signaling loop between cell proliferation and glycolysis. In this regard, increased cytosolic concentrations of Fru1 ,6bisP have been found in cancer cells (Marin-Hernandez et al., 201 1 ; Mazurek et al., 2001 ; Rodriguez-Enriquez et al., 2001 ). Moreover, generation of mitochondrial dysfunction, accumulation of ROS and switch from respiration to increased glycolysis has also been observed in a mammalian cell system in which the K-ras(G12V) oncogene was expressed from an inducible promoter (Hu et al., 2012). The reciprocal activation between glycolytic flux and Ras may lock cancer cells in a vicious cycle between stimulation of cell proliferation and overactive metabolism, from which they can no longer escape. This would explain the close correlation between the proliferation rate and aggressive character of cancer cells and the degree of hyperactivity of the fermentation pathway. It may also help explain other phenomena, for instance, the observation that wild type Sos and Ras are required for tumorigenesis in cells expressing oncogenic Ras (Jeng et al., 2012). The similarity between yeast and mammalian cells is further underscored by the finding that in both systems translocation of activated Ras to the mitochondria results in loss of respiratory activity and the appearance of apoptotic markers (Hu et al., 2012; Leadsham et al., 2009).

Another interesting feature of Ras activation in yeast is the different response between high steady-state Ras activity and its rapid transition-dependent activation. When Ras is already highly activated before addition of glucose as in the quintuple deletion strain (Fig.5 b) or in strains expressing the Ras2^val19 allele (Fig.1 a), the lethal phenotype is not observed. By contrast, when Ras is highly activated during a sudden transition as observed with glucose addition to cells of the tpsIA strain (Fig. 1 a), the cells enter the pathway of programmed cell death. Hence, it appears that it is the sudden activation of Ras in cells experiencing a metabolic problem that is the trigger for apoptosis.

In conclusion, we have demonstrated that a yeast mutant with overactive hexokinase activity, is sensitive to glucose because its hyperaccumulation of Fru1 ,6bisP causes hyperactivation of Ras. In combination with its metabolic deregulation, this results in apoptosis and explains the extreme glucose sensitivity of this mutant. Using permeabilized yeast spheroplasts, we have demonstrated that Fru1 ,6bisP causes activation of Ras in physiological concentrations, that this is mediated by a conserved domain in its GEF Cdc25 and can be mimicked by its human homologue Sos. We have identified the positively charged residues K1491 and R1 122 in Cdc25 and K963 and K602 in Sos, located in a cleft at the interaction point between Cdc25/Sos and Ras, as possible binding partners for the negatively charged phosphates in Fru1 ,6bisP. Our results suggest that active glycolysis stimulates cell proliferation through activation of Ras and that the underlying regulatory mechanism is conserved from yeast to mammals. This finding may be highly relevant for cancer cell physiology since Ras is one of the most important oncogenes and overactive glycolysis is one of the hallmarks of cancer cells. Moreover, the synthesis of Fru1 ,6bisP by phosphofructokinase is a major target of metabolic control pathways, as well as oncogene products and tumor suppressor genes in mammalian cells (Kole et al., 1991 ; Telang et al., 2006; Vander Heiden et al., 2009; Yalcin et al., 2009). Hence, the activation of Ras by Fru1 ,6bisP might be a very ancient mechanism in the evolution of eukaryotic cells, coupling high flux through glycolysis to a high rate of cell proliferation.

Example 5: Use of the tpsIA strain for screening of Fru1 ,6bisP mimicking compounds, and Fru1 ,6bisP / guanine nucleotide exchange protein complex inhibitors

As described earlier, the yeast tpsI A shows an unlimited influx of glucose and a hyperaccumulation of Fru1 ,6bisP, which is binding on Cdc25 so activating Ras. The tpsI A mutant is extremely sensitive to glucose. On 1 mM glucose, there is a strongly limited growth, and the growth is completely blocked on 5mM glucose. This feature can be used to screen for compounds that either inhibit the growth on 1 mM glucose, by mimicking the Fru1 ,6bisP function, or for compounds that restore the growth on 5mM glucose.

The screen was carried out as described in the materials and methods. For the screening on 1 mM glucose, compounds that resulted in growth inhibition of 75% or more, as compared to the growth reference control, the top 640 compounds were selected for the confirmation screen, and the growth inhibition was confirmed.

Similarly, in the screen on 5mM glucose, several possible complex formation inhibitors have been identified.

Both kind of compounds have anti-cancer action, either by inducing programmed cell death (the Fru1 ,6bisP mimicking compounds), or by blocking the Ras hyperactivation (the complex inhibitors) REFERENCES

Averet, N., Fitton, V., Bunoust, O., Rigoulet, M., Guerin, B., 1998. Yeast mitochondrial metabolism: from in vitro to in situ quantitative study. Mol Cell Biochem. 184, 67-79.

Barbacid, M., 1987. ras genes. Annual review of biochemistry. 56, 779-827.

Beullens, M., Mbonyi, K., Geerts, L, Gladines, D., Detremerie, K., Jans, A. W., Thevelein, J.

M., 1988. Studies on the mechanism of the glucose-induced cAMP signal in glycolysis and glucose repression mutants of the yeast Saccharomyces cerevisiae. European journal of biochemistry / FEBS. 172, 227-31 .

Blazquez, M. A., Lagunas, R., Gancedo, C, Gancedo, J. M., 1993. Trehalose-6-phosphate, a new regulator of yeast glycolysis that inhibits hexokinases. FEBS letters. 329, 51 -4.

Blum, R., Jacob-Hirsch, J., Amariglio, N., Rechavi, G., Kloog, Y., 2005. Ras inhibition in glioblastoma down-regulates hypoxia-inducible factor-1 alpha, causing glycolysis shutdown and cell death. Cancer research. 65, 999-1006.

Boriack-Sjodin, P. A., Margarit, S. M., Bar-Sagi, D., Kuriyan, J., 1998. The structural basis of the activation of Ras by Sos. Nature. 394, 337-43.

Bos, J. L, 1989. ras oncogenes in human cancer: a review. Cancer research. 49, 4682-9. Boy-Marcotte, E., Ikonomi, P., Jacquet, M., 1996. SDC25, a dispensable Ras guanine nucleotide exchange factor of Saccharomyces cerevisiae differs from CDC25 by its regulation. Mol Biol Cell. 7, 529-39.

Broach, J. R., Deschenes, R. J., 1990. The function of ras genes in Saccharomyces cerevisiae. Adv Cancer Res. 54, 79-139.

Broek, D., Samiy, N., Fasano, O., Fujiyama, A., Tamanoi, F., Northup, J., Wigler, M., 1985.

Differential activation of yeast adenylate cyclase by wild-type and mutant RAS proteins.

Cell. 41 , 763-9.

Bustamante, E., Morris, H. P., Pedersen, P. L, 1981. Energy metabolism of tumor cells. Requirement for a form of hexokinase with a propensity for mitochondrial binding. The Journal of biological chemistry. 256, 8699-704.

Bustamante, E., Pedersen, P. L, 1977. High aerobic glycolysis of rat hepatoma cells in culture: role of mitochondrial hexokinase. Proceedings of the National Academy of Sciences of the United States of America. 74, 3735-9.

Chen, Z., Zhang, H., Lu, W., Huang, P., 2009. Role of mitochondria-associated hexokinase II in cancer cell death induced by 3-bromopyruvate. Biochimica et biophysica acta. 1787, 553- 60.

Colombo, S., Ronchetti, D., Thevelein, J. M., Winderickx, J., Martegani, E., 2004. Activation state of the Ras2 protein and glucose-induced signaling in Saccharomyces cerevisiae. The Journal of biological chemistry. 279, 46715-22. de Koning, W., van Dam, K., 1992. A method for the determination of changes of glycolytic metabolites in yeast on a subsecond time scale using extraction at neutral pH. Anal Biochem. 204, 1 18-23.

Diaz-Ruiz, R., Uribe-Carvajal, S., Devin, A., Rigoulet, M., 2009. Tumor cell energy metabolism and its common features with yeast metabolism. Biochimica et biophysica acta. 1796, 252- 65.

Fiechter, A., Gmunder, F. K., 1989. Metabolic control of glucose degradation in yeast and tumor cells. Advances in biochemical engineering/biotechnology. 39, 1 -28.

Gourlay, C. W., Ayscough, K. R., 2006. Actin-induced hyperactivation of the Ras signaling pathway leads to apoptosis in Saccharomyces cerevisiae. Mol Cell Biol. 26, 6487-501 .

Grosdidier, A., Zoete, V., Michielin, O., 201 1 . SwissDock, a protein-small molecule docking web service based on EADock DSS. Nucleic Acids Res. 39, W270-7.

Herling, A., Konig, M., Bulik, S., Holzhutter, H. G., 201 1. Enzymatic features of the glucose metabolism in tumor cells. FEBS J. 278, 2436-59.

Herrmann, J. M., Stuart, R. A., Craig, E. A., Neupert, W., 1994. Mitochondrial heat shock protein 70, a molecular chaperone for proteins encoded by mitochondrial DNA. J Cell Biol. 127, 893-902.

Hers, H. G., 1984. The discovery and the biological role of fructose 2,6-bisphosphate. Biochemical Society transactions. 12, 729-35.

Hohmann, S., Neves, M. J., de Koning, W., Alijo, R., Ramos, J., Thevelein, J. M., 1993. The growth and signalling defects of the ggsl (fdp1/byp1) deletion mutant on glucose are suppressed by a deletion of the gene encoding hexokinase Pll. Current genetics. 23, 281 -9. Hu, Y. et al., 2012 K-ras(G12V) transformation leads to mitochondrial dysfunction and a metabolic switch from oxidative phosphorylation to glycolysis. Cell Res 22, 399-412.

Irwin, J. J., Shoichet, B. K., 2005. ZINC-a free database of commercially available compounds for virtual screening. J Chem Inf Model. 45, 177-82.

Iyer, N. V., Kotch, L. E., Agani, F., Leung, S. W., Laughner, E., Wenger, R. H., Gassmann, M., Gearhart, J. D., Lawler, A. M., Yu, A. Y., Semenza, G. L, 1998. Cellular and developmental control of 02 homeostasis by hypoxia-inducible factor 1 alpha. Genes Dev. 12, 149-62. Jeng, H.H., Taylor, L.J. and Bar-Sagi, D. 2012. Sos-mediated cross-activation of wild type Ras by oncogenic Ras is essential for tumorigenesis. Nat. Commun. 3, 399-412.

Kole, H. K., Resnick, R. J., Van Doren, M., Racker, E., 1991 . Regulation of 6-phosphofructo-1 - kinase activity in ras-transformed rat-1 fibroblasts. Arch Biochem Biophys. 286, 586-90. Leadsham, J.E., Miller, K., Ayscough, K.R., Colombo, S., Martegani, E., Sudbery, P. and Gourlay, C.W. 2009. Whi2p links nutritional sensing to actin-dependent Ras-cAMP-PKA regulation and apoptosis in yeast. J. Cell. Sci. 122, 706-715. Lo, C. H., Farina, F., Morris, H. P., Weinhouse, S., 1968. Glycolytic regulation in rat liver and hepatomas. Advances in enzyme regulation. 6, 453-64.

Macheda, M. L, Rogers, S., Best, J. D., 2005. Molecular and cellular regulation of glucose transporter (GLUT) proteins in cancer. J Cell Physiol. 202, 654-62.

Marin-Hernandez, A., Gallardo-Perez, J. C, Rodriguez-Enriquez, S., Encalada, R., Moreno- Sanchez, R., Saavedra, E., 201 1. Modeling cancer glycolysis. Biochimica et biophysica acta. 1807, 755-67.

Mathupala, S. P., Rempel, A., Pedersen, P. L, 1997. Aberrant glycolytic metabolism of cancer cells: a remarkable coordination of genetic, transcriptional, post-translational, and mutational events that lead to a critical role for type II hexokinase. Journal of bioenergetics and biomembranes. 29, 339-43.

Mazure, N. M., Chen, E. Y., Laderoute, K. R., Giaccia, A. J., 1997. Induction of vascular endothelial growth factor by hypoxia is modulated by a phosphatidylinositol 3-kinase/Akt signaling pathway in Ha-ras-transformed cells through a hypoxia inducible factor-1 transcriptional element. Blood. 90, 3322-31 .

Mazurek, S., Zwerschke, W., Jansen-Durr, P., Eigenbrodt, E., 2001. Metabolic cooperation between different oncogenes during cell transformation: interaction between activated ras and HPV-16 E7. Oncogene. 20, 6891 -8.

Mbonyi, K., Beullens, M., Detremerie, K., Geerts, L, Thevelein, J. M., 1988. Requirement of one functional RAS gene and inability of an oncogenic ras variant to mediate the glucose- induced cyclic AMP signal in the yeast Saccharomyces cerevisiae. Mol Cell Biol. 8, 3051 -7.

Neves, M. J., Hohmann, S., Bell, W., Dumortier, F., Luyten, K., Ramos, J., Cobbaert, P., de Koning, W., Kaneva, Z., Thevelein, J. M., 1995. Control of glucose influx into glycolysis and pleiotropic effects studied in different isogenic sets of Saccharomyces cerevisiae mutants in trehalose biosynthesis. Current genetics. 27, 1 10-22.

Noubhani, A., Bunoust, O., Rigoulet, M., Thevelein, J. M., 2000. Reconstitution of ethanolic fermentation in permeabilized spheroplasts of wild-type and trehalose-6-phosphate synthase mutants of the yeast Saccharomyces cerevisiae. European journal of biochemistry / FEBS. 267, 4566-76.

Pettersen, E. F., Goddard, T. D., Huang, C. C, Couch, G. S., Greenblatt, D. M., Meng, E. C, Ferrin, T. E., 2004. UCSF Chimera-a visualization system for exploratory research and analysis. J Comput Chem. 25, 1605-12.

Rodriguez-Enriquez, S., Juarez, O., Rodriguez-Zavala, J. S., Moreno-Sanchez, R., 2001. Multisite control of the Crabtree effect in ascites hepatoma cells. European journal of biochemistry / FEBS. 268, 2512-9. Sass, P., Field, J., Nikawa, J., Toda, T., Wigler, M., 1986. Cloning and characterization of the high-affinity cAMP phosphodiesterase of Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences of the United States of America. 83, 9303-7.

Schomerus, C, Munder, T., Kuntzel, H., 1990. Site-directed mutagenesis of the Saccharomyces cerevisiae CDC25 gene: effects on mitotic growth and cAMP signalling. Mol Gen Genet. 223, 426-32.

Semenza, G. L, 2003. Targeting HIF-1 for cancer therapy. Nat Rev Cancer. 3, 721 -32.

Tanaka, K., Nakafuku, M., Satoh, T., Marshall, M. S., Gibbs, J. B., Matsumoto, K., Kaziro, Y., Toh-e, A., 1990. S. cerevisiae genes IRA 1 and IRA2 encode proteins that may be functionally equivalent to mammalian ras GTPase activating protein. Cell. 60, 803-7.

Telang, S., Yalcin, A., Clem, A. L, Bucala, R., Lane, A. N., Eaton, J. W., Chesney, J., 2006.

Ras transformation requires metabolic control by 6-phosphofructo-2-kinase. Oncogene. 25, 7225-34.

Thevelein, J. M., de Winde, J. H., 1999. Novel sensing mechanisms and targets for the cAMP- protein kinase A pathway in the yeast Saccharomyces cerevisiae. Molecular microbiology. 33, 904-18.

Thomas, B. J., Rothstein, R., 1989. Elevated recombination rates in transcriptionally active DNA. Cell. 56, 619-30.

Van Aelst, L, Boy-Marcotte, E., Camonis, J. H., Thevelein, J. M., Jacquet, M., 1990. The C- terminal part of the CDC25 gene product plays a key role in signal transduction in the glucose-induced modulation of cAMP level in Saccharomyces cerevisiae. European journal of biochemistry / FEBS. 193, 675-80.

Van Aelst, L, Hohmann, S., Bulaya, B., de Koning, W., Sierkstra, L, Neves, M. J., Luyten, K., Alijo, R., Ramos, J., Coccetti, P., 1993. Molecular cloning of a gene involved in glucose sensing in the yeast Saccharomyces cerevisiae. Molecular microbiology. 8, 927-43.

van Aelst, L, Jans, A. W., Thevelein, J. M., 1991 . Involvement of the CDC25 gene product in the signal transmission pathway of the glucose-induced RAS-mediated cAMP signal in the yeast Saccharomyces cerevisiae. J Gen Microbiol. 137, 341 -9.

Vander Heiden, M. G., Cantley, L. C, Thompson, C. B., 2009. Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science. 324, 1029-33.

Weber, G., Lea, M. A., 1966. The molecular correlation concept of neoplasia. Advances in enzyme regulation. 4, 1 15-45.

Wilson, J. E., 2003. Isozymes of mammalian hexokinase: structure, subcellular localization and metabolic function. The Journal of experimental biology. 206, 2049-57.

Yalcin, A., Telang, S., Clem, B., Chesney, J., 2009. Regulation of glucose metabolism by 6- phosphofructo-2-kinase/fructose-2,6-bisphosphatases in cancer. Exp Mol Pathol. 86, 174-9.

Claims

1 . An isolated fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex.

2. The isolated fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex according to claim 1 , wherein said guanine nucleotide exchange protein is selected from the group consisting of cdc25, sdc25, hSosl and hSos2.

3. The isolated complex according to claim 1 or 2, further comprising a Ras protein.

4. The use of a fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex to screen for compounds to modulate Ras activation.

5. The use of a fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex complex according to claim 4, wherein said complex is further comprising a Ras protein.

6. The use of a complex according to claim 4 or 5, wherein said modulation of Ras activation is inhibition of Ras activation.

7. A modulator of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation for use as a medicament.

8. The modulator of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation for use as a medicament according to claim 7, wherein said modulator is an inhibitor of the complex formation.

9. A modulator of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation for use in treatment of cancer.

10. The modulator of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation for use in treatment of cancer, according to claim 9, wherein said modulator is an inhibitor of the complex formation.

1 1 . The modulator of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation according to any of the claims 7 to 10, wherein said guanine nucleotide exchange protein is selected from the group consisting of cdc25, sdc25, hSosl and hSos2.

12. An inhibitor of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation for use in treatment of cancer, according to claim 10, wherein said inhibitor is selected from the group consisting of pyridoxal-5-phosphate, mannitol-1 ,6- bisphosphate, D-ribulose-1 ,5-diphosphate, tagatose-1 ,6-diphosphate, L-sorbose-1 - phosphate and gluconic acid 6-phosphate.

13. The inhibitor of the fructose-1 ,6-bisphosphate / guanine nucleotide exchange protein complex formation for use in treatment of cancer according to claim 12, wherein said guanine nucleotide exchange protein is selected from the group consisting of cdc25, sdc25, hSosl and hSos2.