WO2009100760A1 - Cellular screening assay for modulators of mevalonate pathway - Google Patents

Abstract

The present invention relates to a cellular assay for screening modulators of the mevalonate pathway. More specifically the cellular assay comprises the following steps: a. culturing eukaryotic cells expressing a protein with a prenylation signal in the presence and in the absence of a test compound; and, b. comparing the subcellular localization of said protein with a prenylation signal in the presence of a test compound with said subcellular localization in the absence of the test compound; wherein a change in the subcellular localization of said protein with a prenylation signal is indicative that said test compound is a modulator of the mevalonate pathway.

Description

CELLULAR SCREENING ASSAY FOR MODULATORS OF MEVALONATE PATHWAY

The present invention relates to a cellular assay for screening modulators of the mevalonate pathway. More specifically the cellular assay comprises the following steps: a. culturing eukaryotic cells expressing a protein comprising a prenylation signal in the presence or in the absence of a test compound; and, b. comparing the subcellular localization of said protein with a prenylation signal in the presence of test compound with said subcellular localization in the absence of the test compound; wherein a change in the subcellular localization of said protein with a prenylation signal is indicative that said test compound is a modulator of the mevalonate pathway.

BACKGROUND

The mevalonate pathway leads to synthesis of cholesterol and isoprenoid lipids (Figure 1 ). Mevalonate is the first product of the pathway and functions as the precursor for isoprenoid intermediates, lsoprenyl intermediates with progressive length are synthesized by sequential condensation of isopentyl diphosphate (IPP) with an allylic diphosphate. IPP is condensated with dimethylallyl diphosphate to produce geranyl pyrophosphate (GPP). Farnesyl pyrophosphate (FPP) is then produced from GPP and IPP by farnesyl pyrophosphate synthase (FPPS). FPP can be further metabolized via squalene to cholesterol or to geranylgeranyl pyrophosphate (GGPP). Cholesterol functions as a precursor for several steroid hormones, whereas FPP and GGPP are attached to the C-terminus of several small GTPases. Proteins like Ras, HDJ2 and lamins get farnesylated, whereas Rab, Rap1A, RhoA, Rac1 and Cdc42 get geranylgeranylated (Roskoski, Jr., 2003). A so-called CaaX motif functions as a prenylation signal for the enzymes farnesyl transferase and geranylgeranyl transferase. The C in the motif refers to cysteine, the a to an aliphatic amino acid and the X is typically a methionine, serine, alanine, glutamine or leucine (Roskoski, Jr., 2003). After attachment of the isoprenoids the three most C-terminal amino acid residues are removed by a CaaX endoprotease (Roskoski, Jr., 2003). Thereafter additional modification with other lipid groups can occur. The isoprenoid groups are essential for the biological activity and proper localization of the small GTP-binding proteins. One of the first human oncogenes identified was ras (Der et al., 1982). The Ras proteins transduce information from cell surface to cytoplasmic components of the cells. The ras genes are mutated or overexpressed in 90% of pancreatic carcinomas, 50% of thyroid and colon cancers, 30% of lung cancers and myeloid leukemias (Scharovsky et al., 2000). In order to carry out their biological function the Ras proteins have to be associated with plasma membrane (PM). The PM association is achieved by addition of farnesyl and palmitoyl groups to the C-terminus of the protein. In the absence of the lipid moieties the Ras proteins cannot perform their biological function (Tamanoi et al., 2001 ). Prevention of the activity of Ras proteins by inhibiting their prenylation has been extensively studied as a potential treatment for cancer (Perrin and Hill, 2000;Scharovsky et al., 2000;Tamanoi et al., 2001 ). The enzyme protein farnesyl transferase (PFT) as well as HMG CoA reductase and farnesyl pyrophosphate synthase (FPPS) have been the major targets (Cohen et al., 2000;Perrin and Hill, 2000;Scharovsky et al., 2000;Tamanoi et al., 2001 ).

Statins, like mevastatin and lovastatin, are inhibitors of the HMG CoA reductase, and they have been already long used to reduce serum cholesterol levels. More lately the potential of statins in cancer therapy has been discovered, and a few statins are in clinical trials for various cancers (Swanson and Hohl, 2006).

Bisphosphonates (BPs) are drugs that are currently used for osteoporosis and metastatic bone cancers. Zoledronate and several other nitrogen-containing BPs inhibit the enzyme FPPS (Bergstrom et al., 2000;Dunford et al., 2001 ;Keller and Fliesler, 1999;van Beek et al., 1999). However, the bisphosphonates have high affinity for bone mineral, poor cell permeability and poor oral availability (Conte and Guarneri, 2004;van Beek et al., 1998). Therefore new types of FPPS inhibitors with better cell permeability and oral availability would be desirable.

Inhibitors of protein farnesyltransferase (PFT) are effective in inhibiting growth of several cell lines in culture. They have proven to be effective also in mouse tumor models, and several of them are already in clinical trials (Swanson and Hohl, 2006). The potential of inhibitors of geranylgeranyl transferase as anti-cancer agents is being studied as well. The studies with these inhibitors, however, have not yet progressed as far as those with inhibitors of farnesyltransferase (Swanson and Hohl, 2006). Most inhibitors of the mevalonate pathway prevent the activity of several prenylated proteins, like Rho, Rac and the Rab protein family. In spite of the broad range of proteins and functions affected, inhibitors of the mevalonate pathway, like lovastatin and nitrogen- containing bisphosphonates (NCBP) are relatively well tolerated by patients and several of them have been already for long in clinical use (Cohen et al., 2000;Green, 2004;Perrin and Hill, 2000). Inhibitors of the mevalonate pathway have beneficial effects in treatment of cancer, osteoporosis, restenosis and they can be used to lower serum cholesterol levels.

Biochemical assays for identification of FPPS inhibitors that are used at present include 1 ) western blotting to detect non-prenylated Rapi a (Thompson et al., 2006) and 2) mevalonolactone assay where radioactive mevalonolactone is given to the cells and after a certain incubation time the presence of radioactivity in the membrane fraction of the cells is measured. Alternatively the presence of radioactivity in proteins can be detected using SDS gels and phosphoimaging or autoradiography (Luckman et al., 1998). Both are tedious low- throughput assays.

To the best of Applicant's knowledge, there is no suitable readout available to screen and study for inhibitors of the mevalonate pathway in a quantifiable manner which can be used for cellular screening.

The present invention fulfills this need by providing the first cellular screening assays for identifying modulators of the mevalonate pathway. Indeed, the present invention discloses that fluorescent GFP tagged with a prenylation signal is localized to the plasma membrane when the mevalonate pathway is active. Inhibitors of the pathway prevent prenylation thereby leading to cytosolic or nuclear localization of the tagged GFP. Among the several advantages of the present methods, it should be noted that this change in the subcellular localization of tagged GFP can be detected either qualitatively or quantitatively by appropriate means. These findings enabled the conception of simple and efficient cellular screening assays for modulators of the mevalonate pathway, as well as cell lines for use in such assays, as further described in detail hereafter.

DETAILED DESCRIPTION OF THE INVENTION - A -

One aspect of the invention relates to a method of screening for modulators of the mevalonate pathway, said method comprising: a. culturing eukaryotic cells expressing a protein comprising a prenylation signal in the presence or in the absence of a test compound; and, b. comparing the subcellular localization of said protein with a prenylation signal in the presence of test compound with said subcellular localization in the absence of the test compound; wherein a change in the subcellular localization of said protein with a prenylation signal is indicative that said test compound is a modulator of the mevalonate pathway.

The mevalonate pathway is defined herein as any biological pathway which leads to the synthesis of farnesyl pyrophosphate and/or geranylgeranyl pyrophosphate. Some of the components of the mevalonate pathway are represented in figure 1. Components of the mevalonate pathway include, without limitation, (Hydroxymethyl-glutaryl coenzyme A reductase) HMG CoA reductase, farnesyl pyrophosphate synthase (FPPS), Protein farnesyltransferase (PFT), Geranylgeranyl pyrophosphate synthase (GGPPS) and Protein geranylgeranyltransferase (PGGT-1 ).

As used herein, "method of screening for modulators" includes any assay for identifying a compound that modulates the mevalonate pathway, when incubated with an appropriate test cell.

"Inhibitors", "activators" and "modulators" of the mevalonate pathway are used to refer to inhibitory, activating, or modulating molecules identified by using the assays based on the principles of the invention. Inhibitors can be for example compounds that partially or totally block, decrease, prevent, delay, inactivate, desensitize or down regulate the mevalonate pathway, e.g., inhibitors of HMG CoA reductase or FPPS. Activators are compounds that, e.g., stimulate, increase, activate, the mevalonate pathway, e.g., activator of HMG CoA reductase or FPPS activity. Modulators are either inhibitors or activators. Candidate compounds may thus either interfere by blocking, preventing or stimulating the mevalonate pathway. Activity of the mevalonate pathway is detected by the localization of the protein with the prenylation signal predominantly to the plasma membrane. Alternatively, inhibition of the mevalonate pathway is detected by cytosolic and/or nuclear localization of the protein with a prenylation signal. A "eukaryotic cell" which can be used in the assay can be a naturally occurring cell or a transformed cell, provided that said eukaryotic cell is capable of expressing a protein with a prenylation signal with a sufficient level of expression so that the subcellular localization of said protein with a prenylation signal change upon inhibition of the mevalonate pathway. In a preferred embodiment, a eukaryotic cell which can be used in the assay is selected among those where the subcellular localization of said protein with a prenylation changes from the plasma membrane to the cytosol upon inhibition of the mevalonate pathway.

Eukaryotic cells may be cultured cells, explants, cells in vivo, and the like. Eukaryotic cells may include yeast, insect, amphibian, or mammalian cells and the like. Preferably, eukaryotic cells used in the assay are mammalian cell lines, such as HaCaT, HeLa, MDCK, U-2 OS, CHO-K1 , or primary cells. The cells may be cultured under standard conditions, which will be determined easily by the one skilled in the Art.

As used herein, the term "prenylation signal" refers to a peptide comprising at least the CaaX motif, wherein "C" stands for cystein, "a" for an aliphatic amino acid and the X is preferably a methionine, serine, alanine, glutamine or leucine, e.g. the CVLS motif, and, preferably, further includes a palmitoylation motif.

Said prenylation signal is required for in vivo prenylation of a protein carrying such prenylation signal and expressed in a eukaryotic cells under appropriate conditions.

As used herein, the term "palmitoylation motif" refers to a peptide which contains one or more cysteine residues and is required for in vivo palmitoylation of a protein carrying such peptide and expressed in a eukaryotic cell under appropriate conditions.

In a specific embodiment, said prenylation signal is a 4-50 amino acids polypeptide. In another specific embodiment, such peptide is expressed as a C-terminal fusion with another protein and recombinantly expressed by the host cell. In yet another preferred embodiment, said prenylation signal combined with palmitoylation signal is a peptide consisting of a C- terminal fragment of human H-Ras, e.g. the peptide consisting of SEQ ID NO:1 GCMSCKCVLS as well as any variant or mutant version which retains the same subcellular localization as the wild type corresponding prenylation signal, hereafter referred as a "suitable variant of prenylation signal". In a specific embodiment, a suitable variant of said prenylation signal preferably shares at least 50%, preferably at least 60%, 70% or 80% identity, and even more preferably, at least 90% identity with the peptide sequence of SEQ ID NO:1 or any other known prenylation signal in the art, when performing optimal alignment. Optimal alignment of sequences for determining a comparison window may be conducted by the local homology algorithm of Smith and Waterman (Smith and Waterman, 1981 ), by the homology alignment algorithm of Needleman and Wunsch, (Needleman and Wunsch, 1970), by the search for similarity via the method of Pearson and Lipman, (Pearson and Lipman, 1988) or by computerized implementations of these algorithms (GAP, BESTFIT, FASTA and TFASTA in the Wisconsin Genetics Software Package Release 7.0, Genetic Computer Group, 575, Science Drive, Madison, Wisconsin). The best alignment (Ae., resulting in the highest percentage of identity over the comparison window) generated by the various methods is selected for determining percentage identity.

In a preferred embodiment, a suitable variant of prenylation signal of SEQ ID NO:1 is any C-terminal fragment of the H-Ras protein which retains substantially the same subcellular localization as wild type corresponding SEQ ID NO:1. In a more preferred embodiment, a suitable fragment is a fragment of SEQ ID NO:1 which consists of or contains at least the CVLS domain.

In one specific embodiment of the method, the eukaryotic cells stably or transiently express the gene encoding the protein with a prenylation signal as defined by SEQ ID NO:1 or a suitable variant thereof. Preferably, the eukaryotic cells stably express recombinant protein fusion with a prenylation signal under a promoter which enables sufficient expression, such as the CMV promoter. An alternative method uses eukaryotic cells that express endogenous protein with a prenylation signal.

The method of the invention is based on the principle that a protein comprising a prenylation signal is unable to localize to the plasma membrane upon inhibition of the mevalonate pathway, e.g., in the presence of FPPS inhibitors.

The term "subcellular localization" should be understood as the specific localization of a gene product within the volume of the cell, and more specifically, how it is disposed relative to the cytosol and other identifiable cellular features or compartments, or organelles such as the plasma membrane, the Golgi membranes, endosomal vesicles, nucleus, endoplasmic reticulum, mitochondria and so on.

Detection of the protein localization can be monitored by any appropriate means available in the art. In the method of the invention, the cells are mechanically intact and alive throughout the experiment. Alternatively, the cells are fixed at a point in time after the application of the influence at which the response has been predetermined to be significant, and the recording is made at an arbitrary later time.

Any change in the subcellular localization of the protein comprising the prenylation signal is detected either qualitatively and/or quantitatively by appropriate means. A significant change in localization from the plasma membrane to the cytosol in the presence of the test compound is indicative that said test compound is capable of inhibiting the mevalonate pathway. Preferably, "significant change in localization" is obtained when a measurable amount of the detected proteins with a prenylation signal have their subcellular localization changed in the presence of the test compound as compared with cells not incubated with the test compounds. For example, significant changes in the cellular localization can be determined by the use of IN Cell Analyzer 3000 with the membrane trafficking module, as described in the "methods" paragraph of the experimental part below.

Examples of detection methods are:

- immunodetection, wherein a specific and mass-produced available antibody raised against the protein with a prenylation signal is used to detect endogenous or recombinantly expressed protein with prenylation signal present in mammalian cells. Alternatively, an antigenic tag is conjugated to recombinantly expressed protein comprising a prenylation signal such as the flag or myc tags which are foreign and therefore unique antigens in eukaryotic cells and for which mass-produced antibodies are available so that the need to develop an antibody against said protein comprising a prenylation signal is avoided,

- direct detection wherein a probe comprising the prenylation signal is engineered with a fluorophore or luminophore tag, or using a protein tag which can be labeled in living cells using e.g. a colored substrate and thereby directly revealing its own cellular distribution.

In one possible embodiment, subcellular localization of recombinant protein with a prenylation signal is directly detected by microscopy to detect the prenylation signal tagged with a luminophore or a fluorophore. Those detection means are described for example in Thastrup et al., US6, 518,021. In another embodiment, subcellular localization of the protein with a prenylation signal can be indirectly detected by using an extracellular quencher for the fluorescent label, such as Acid Red 88 (Sigma). The quencher quenches the fluorescence of the protein tag if said protein is localized at the plasma membrane, but not if said protein is localized in the cytosol, and fluorescence intensity can be determined on a fluorescence reader. Those detection means are described for GFP-tagged Akt (Lundholt et al., 2005). In another embodiment, subcellular localization of protein with a prenylation signal is detected by using a specific primary antibody binding to protein with prenylation signal, and e.g. a secondary antibody specific for the primary antibody, which is labeled with e.g. fluorophore or luminophore tag which can be detected by microscopy.

In the described embodiments, said eukaryotic cells express a recombinant protein having a prenylation signal, coupled to a luminophore or fluorophore. Alternatively, said cells express endogenous protein having a prenylation signal, which protein is detected by an antibody coupled directly or indirectly to a luminophore or fluorophore.

As used herein, a luminophore or fluorophore can be any fluorescent or luminescent molecule which enables to detect the subcellular localization of a protein which is conjugated to it.

In a further preferred embodiment, the luminophore is a fluorophore. It is preferably a fluorescent protein, such as the green fluorescent protein (GFP), or a non-fluorescent protein, which can be labeled in living cells by e.g. using a labeled enzyme substrate.

In the present context, the term "green fluorescent protein" (GFP) is intended to indicate a protein which, when expressed by a cell, emits fluorescence upon exposure to light of the correct excitation wavelength (e.g. as described by Chalfie, M. ef al.. (Chalfie et al., 1994). Such a fluorescent protein in which one or more amino acids have been substituted, inserted or deleted is also termed "GFP". "GFP" as used herein includes wildtype GFP derived from the jelly fish Aequorea victoria, or from other members of the Coelenterata, such as the red fluorescent protein from Discosoma sp. (Matz et al., 1999) or fluorescent proteins from other animals, fungi or plants, and modifications of GFP, such as the blue fluorescent variant of GFP disclosed by Heim ef al. (Heim et al., 1994), and other modifications that change the spectral properties of the GFP fluorescence, or modifications that exhibit increased fluorescence when expressed in cells at a temperature above 30⁰C as described in WO971 1094. Preferred GFP variants are F64L-GFP, F64L-Y66H-GFP, F64L- S65T-GFP, F64L-E222G-GFP. One especially preferred variant of GFP for use in the methods of the invention is EGFP, which is a F64L-S65T variant with codons optimized for expression in mammalian cells is available from Clontech Laboratories, Inc., and other vendors. Alternatively, GFP can be cloned, and inserted into e.g. the pcDNA3 vector or its derivatives (Invitrogen Inc.). Other protein tags, which are non-fluorescent/ luminescent, and can be labeled in living cells, are e.g. the SNAP-tag from Covalys AG, or the HaloTag from Promega.

The SNAP-tag is a tool for protein research, allowing the specific, covalent attachment of virtually any molecule to a protein of interest. There are two steps to the use of this system: sub-cloning and expression of the protein of interest as a SNAP-tag fusion, and labeling of the fusion in live cells with the SNAP-tag substrate of choice. SNAP-tag is a small protein based on mammalian O⁶-alkylguanine-DNA-alkyltransferase (AGT). SNAP-tag substrates are derivates of O⁶-benzylguanines. In the labeling reaction, the benzyl group of the substrate which carries the label is covalently attached to the SNAP-tag. As SNAP-tag is highly optimized for reaction with O⁶-benzyl guanines and related substrates, no background signal from endogenous AGT is observed (Keppler et al., 2004;Keppler et al., 2006;).

The HaloTag Interchangeable Labeling Technology is a tool for imaging live or fixed mammalian cells that express the HaloTag protein or protein fusions, analyzing e.g. posttranslational modification of labeled fusion proteins. The technology is based on efficient formation of a covalent bond between a specially designed reporter protein encoded by the HaloTag pHT2 Vector and a specific ligand such as in living cells, or in solution. The HaloTag pHT2 Vector contains the open reading frame for a genetically engineered derivative of a hydrolase gene. This protein is not endogenous to mammalian cells. The ligand can carry a variety of functionalities, including fluorescent labels. The covalent bond forms rapidly under general physiological conditions, is highly specific and essentially irreversible, yielding a complex that is stable even under stringent conditions. The open architecture of the technology enables use of different ligands. Promega offers HaloTag ligands that can readily cross the cell membrane with either red or green fluorophores or biotin. Additional ligands will be offered soon (Schindler et al., 1999;Los et ai. 2005).

The advantage of the use of a polypeptide molecule being e.g. a fluorophore or luminophore, or which can be labeled in living cells with a fluorophore or luminophore, is that it can be expressed as a protein fusion with a protein having a prenylation signal or directly with said prenylation signal and light emission of this protein localizes exactly with said protein fusion. In such fusion protein, the prenylation signal or the protein with the prenylation signal is preferably fused to the C-terminus of the fluorophore or luminophore, optionally via a peptide linker. The protein fusion is encoded by a nucleic acid construct and said nucleic acid construct is transfected to the eukaryotic cells. The nucleic acid construct contains an appropriate promoter for expression of the protein fusion, for example the cytomegalovirus (CMV) promoter. The nucleic acid construct can be transfected in the eukaryotic cells for transient expression or integrated into the genome for stable expression. Methods for transfecting mammalian cells and selecting cell lines stably expressing a gene construct are well known in the art.

In one possible embodiment, said nucleic acid construct comprises a gene encoding a the prenylation signal fused C-terminally to GFP protein, e.g. EGFP, said gene being under the control of a promoter, e.g., the CMV promoter. Alternatively, said nucleic acid construct comprises a gene encoding the prenylation signal fused C-terminally to an alternative protein tag such as a SNAP or HaloTag protein.

In another embodiment, subcellular localization of endogenous protein having a prenylation signal is detected by using a specific primary antibody, e.g. anti-H-RAS antibody which detects endogenously present human H-RAS protein, and using secondary antibody labeled e.g. with AlexaFluor488/633 (Molecular Probes). The secondary antibody can be e.g. a chicken anti-mouse or a rabbit anti-goat antibody, which recognizes the anti-H-RAS antibody, and the fluorophore or luminophore tag is detected by microscopy.

The cell may optionally express one or more non-native gene products, e.g. receptors, enzymes, enzyme substrate, prior to or in addition to the protein with the prenylation signal construct. In one specific embodiment, it may be appropriate to co-express both the protein with the prenylation signal construct with siRNA/shRNA capable of downregulating a component of the pathway in order to increase the sensitivity of the assay.

In a specific embodiment where GFP was used in fusion with a prenylation signal for subcellular localization, it was observed in the experiments described below that toxicity may lead to overlapping GFP and nuclear stain due to shrinking of the cells. It is also known that inhibitors of the mevalonate pathway lead to cell apoptosis. Therefore, toxic compounds tested with the method of the invention could be identified as false positive. However, it has been shown here that the change of subcellular localization always preceded apoptosis of the cells. Therefore, according to the present method, it is possible to discriminate between false positive corresponding to toxic compounds and true inhibitors of the mevalonate pathway. More specifically, in a specific embodiment of the method, the subcellular localization is detected at least in live or fixed cells, at a time point when the cells still have normal morphology and would have not started apoptosis if the compound was toxic in order to discriminate among an apparent cytosol localization due to apoptosis and prevention of plasma membrane localization due to inhibition of the mevalonate pathway. In another specific embodiment, the subcellular localization is detected at least in live or fixed cells, at a compound concentration with which the cells still have normal morphology and do not start apoptosis despite toxicity of the compounds at a higher concentration in order to discriminate among an apparent cytosol localization due to apoptosis and prevention of plasma membrane localization due to inhibition of the mevalonate pathway.

It is another object of the invention to provide a eukaryotic cell line, stably comprising a nucleic acid construct encoding a protein fusion between a green fluorescent protein and a prenylation signal, or a fusion of a protein having a prenylation signal and a non-fluorescent/ -luminescent protein tag, which can be labeled in living cells, such as the SNAP-tag or HaloTag.

The invention further concerns the use of the cell lines as defined above in a cellular screening assay for identifying modulators of the mevalonate pathway, and more particularly in the cellular assays described above.

The following examples are intended for purposes of illustration only and should not be construed to limit the scope of the invention, as defined in the claims appended thereto.

LEGENDS OF THE FIGURES

Figure 1 : The mevalonate pathway leading to synthesis of cholesterol and the isoprenoid lipids farnesyl pyrophosphate and geranylgeranyl pyrophosphate. Figure modified from (Cohen et al., 2000).

Figure 2: Distribution of HaCaT cells in the green channel, non-transfected (upper panel), transfected with pHE869 before sorting (middle panel), and transfected with pHE869 after cell sorting (lower panel). M1 in the middle panel shows the region selected for sorting.

Figure 3: Localization of GFP-CaaX (left) and GFP-SaaX (right) in transiently transfected CHO K1 cells expressed from pHE869 or pHE877, respectively.

Figure 4: Localization of GFP-CaaX (left) and GFP-SaaX (right) in transiently transfected HaCaT cells.

Figure 5: HaCaT cells were transiently transfected with pHE869. 24 hours after transfection mevastatin was added to a final concentration of 20 μM. The cells were further incubated for 19 (left) or 25 (right) hours before photographing. In the right image cells that are rounding up and detaching from culture dish are marked with red arrows.

Figure 6: The effect of mevastatin on localization of GFP-CaaX (Dpeak) in stably transfected HaCaT cells (on the left) and the separation of the high and low values (on the right).

Figure 7: Dose-response curves for various inhibitors of the mevalonate pathway in U- 2 OS clone 5. 1000 cells per well were seeded in 384-well format. Next day the inhibitors were added, and the cells were incubated with the drugs for 48 hours. Then the cells were fixed and the nuclei stained with DRAQ5. The cells were imaged with an IN Cell Analyzer 3000, and the Dpeak was analyzed with the Plasma Membrane Trafficking module. The obtained IC50 values are indicated below each curve. Figure 8: Example images of U-2 OS clone 5 cells treated with different concentrations of zoledronic acid or mevastatin, the corresponding dose-response curves are shown in Fig 7.

Figure 9: Example images of U-2 OS clone 5 cells treated with different concentrations of GGTI298 or FTI276, the corresponding dose-response curves are shown in Fig 7.

EXAMPLES

Methods

Plasmid construction

Plasmids pHE869 and pHE877 encoding eGFP with C-terminal tail for a wild-type or a mutated prenylation signal, respectively, were constructed as follows. First a GFP construct with signal for monopalmitoylation and prenylation was constructed by synthesizing GFP with PCR using Pwo polymerase, primers HO521 (SEQ ID NO: 2) and HO710 (SEQ ID NO: 3), and, as a template, a plasmid having non-tagged eGFP as a Hindlll-EcoRI fragment in pcDNA3 vector. Primer HO521 creates a Hindi 11 site immediately upstream of the start ATG codon of GFP. Primer HO710 omits the natural stop codon of GFP and elongates the cDNA with a BamHI site and a sequence coding for a CKCVLS peptide, and finally creates an EcoRI site downstream of the new stop codon. The more C-terminal C serves as the site for prenylation and the more N-terminal C for palmitoylation (Hancock et al., 1989). This PCR fragment and pcDNA3 were digested with Hindlll and EcoRI and ligated together. The resulting plasmid was named pHE853.

To obtain doubly palmitoylated GFP, pHE853 was digested with BamHI and EcoRI. Oligonucleotides HO714 (SEQ ID NO: 4) and HO715 (SEQ ID NO: 5) coding for the prenylation and double palmitoylation signals from human H-Ras (GCMSCKCVLS, the C- terminal C for prenylation, the two N-terminal Cs for palmitoylation (Hancock et al., 1989)) were annealed with each other and ligated to the BamHI-EcoRI-cleaved pHE853. Plasmid pHE877 coding for the mutated prenylation signal (GCMSCKSVLS) was constructed similarly as pHE869, but using the oligonucleotides HO716 (SEQ ID NO:6) and HO717 (SEQ ID NO:7). The GFP encoded by pHE869 will be called GFP-CaaX in the text below, and the GFP encoded by pHE877 is called GFP-SaaX. All plasmids were verified by sequencing at Solvias. The oligonucleotides are listed in Table 1.

Table 1 : Oligonucleotides used in plasmid construction.

The restriction sites are in underlined, The start codon and stop codon are marked in bold. HO714-HO717 were designed to have compatible ends for BamHI and EcoRI and were ordered as 5' phosphorylated from Microsynth GmbH, Switzerland.

Cell culture and transfection

CHO (Chinese hamster ovary) K1 cells were cultivated in RPMI 1640 medium with stable glutamine supplemented with 10% fetal calf serum (FCS) and 1% penicillin/streptomycin (P/S). HaCaT cells (human keratinocytes) were cultivated in DMEM medium with stable glutamine supplemented with 5% FCS, 1 % P/S, and 1 % Na-pyruvate. In cultivation of stably transfected HaCaT cells G418, 500 μg/ml, was used. Human osteosarcoma U-2 OS cells were cultivated in McCoy's 5A medium with stable glutamine, 10% FCS, 1 % P/S. For stably transfected cells 400 μg/ml G418 was additionally included. All cell lines were transfected using Lipofectamine 2000 (Invitrogen, cat no 1 1668-019) and OPTIMEM 1 (Invitrogen, cat no 31985-047) following the instructions of the manufacturer unless otherwise stated in the Results section.

Construction of cell lines and cell populations stably expressing GFP with the prenylation signal

HaCaT cells were seeded on a 6-well plate, 300.000 cells per well. The plasmid pHE869 was linearized with BgIII restriction enzyme. The completeness of the restriction digest was determined using agarose gel electrophoresis. Next day the cells were transfected with pHE869, 2 μg plasmid DNA per well using Lipofectamine 2000. 2 μl of Lipofectamine 2000 was used per 1 μg of plasmid DNA. Otherwise the protocol of the manufacturer of Lipofectamine 2000 was followed. Microscopic inspection of the cells 24 hours after transfection indicated that the transfection was successful. Using Nikon Eclipse TE2000 microscope with the C-FL EPI-FL filter block GFP-B (ex460-500 nm, em510-560 nm) the expression of GFP-CaaX could be confirmed. The medium was changed to a new one containing either 500 μg (half of the wells) or 700 μg (half of the wells) G418 to prevent growth of the non-transfected cells. The cells were cultivated further in the presence of G418. Wells having most green cells and cells where GFP fluorescence could be observed in the plasma membrane were chosen for further cultivation. Since only part of the G418-resistant cells expressed GFP, the GFP-expressing cells were enriched using fluorescence-activated cell sorting. Sorting was done using FACSCalibur (BD) as follows. The cells were detached from a culture dish with 0.25% trypsin/1 mM EDTA, counted with NucleoCounter, and the cell concentration was adjusted to a maximum of 1x107 cells/ml in PBS. For aseptic sorting the BD FACSsorter was rinsed 30 min with 70% ethanol followed by a 30 min PBS flow. For Acquisition instrument settings for Side Scatter, Forward Scatter and GFP were adjusted for the HaCaT cells and were then applied to the sort. To get only the fraction of transfected cells containing the GFP-CaaX plasmid, GFP-positive cells were gated as shown in the histogram in figure 2. Before sorting about 16% of the cells expressed GFP. Immediately after sorting this was increased to 87%. However, microscopic inspection indicated that fewer than 87% of the cells expressed GFP. Also, during cultivation of the cells the proportion of the GFP-expressing cells decreased to 6%. Therefore the sorting was repeated. After the second round of cell sorting about 70% of the cells expressed GFP according to the FACS analysis. Microscopic observation indicated again that less than half of the cells expressed GFP. However, this cell population could be used in experiments to study the localization of the GFP-CaaX.

The human osteosarcoma U-2 OS cells suit well for imaging, since the cells are large and do not grow on top of each other. Moreover, they would provide another cell background to study the effect of inhibitors of the mevalonate pathway. In addition, they grow faster than the HaCaT cells and therefore are more suitable for high-throughput screening than the HaCaT cells. Therefore we also transfected the U-2 OS cells with pHE869, and selected cell populations and clones stably expressing GFP-CaaX. The plasmid pHE869 was linearized again with BgIII and transfection and selection were performed as for HaCaT cells above. Only Fugene 6 (Roche) instead of Lipofectamine 2000 was now used as the transfection reagent. 24 hours after transfection G418 was added to a final concentration of 800 μg/ml to select for the successfully transfected cells. After nine days of selection with G418 the GFP- expressing U-2 OS cells were enriched with cell sorting similarly as the HaCaT cells above. After sorting the cells were counted, diluted and plated on 96-well plates so that in theory there should be one cell / well. After five days in culture 30 GFP-expressing clones were selected based on microscopic inspection. Six of these clones were analyzed in the GFP- CaaX localization assay using mevastatin and zoledronate.

Detection of localization of GFP with the CaaX motif

Screening and staining procedure

3000 stably transfected HaCaT cells or 1000 stably transfected U-2 OS per well were seeded on black 384-well Corning plates (cat no 3985) with transparent bottom. The culture volume was 20 μl per well. 24 hours later the compounds were added in a volume of 5 μl / well. The cells were incubated for 48 hours with the compounds. Then the cells were fixed and the nuclei stained by addition of 25 μl of 8% paraformaldehyde (PFA) plus 1 μM DRAQ5. The cells were incubated with the PFA and DRAQ5 (Biostatus Ltd, cat no BOS-889-001 ) for 15 minutes, washed twice with PBS, and left in 20 μl PBS per well. The plates were sealed with Silverseal aluminium-tape (Greiner, cat no 76090) and kept at 4°C in the dark until analysis with IN Cell Analyzer 3000 high content cell imager.

Imaging of cells with IN Cell Analyzer 3000

Images and data were automatically obtained with an IN Cell Analyzer 3000 imager using the 4Ox objective. Flat field correction was done using a plate, where various mixtures of Oregon Green, Cy5 calibration reagent and Alexa Fluor 350 carboxylic acid had been prepared. One image per well, 750 μm x 750 μm, was taken. Images for both channels (DRAQ5 and GFP) were taken simultaneously. DRAQ5 was excited at 647 nM and GFP at 488 nM. Emission for DRAQ5 was detected with 695/BP55 filter and that for GFP with 535/BP45. Usually 10% of the krypton laser power was used for DRAQ5 excitation and 25% of the argon-ion laser power for the GFP excitation. Exposure time was 1.7 ms.

The plasma membrane trafficking analysis module of the IN Cell Analyzer 3000 software was used for image analysis. Dpeak is the measure for the distance of the peak fluorescence intensity on the GFP channel from the center of the nucleus. When GFP-CaaX is membrane- bound the Dpeak value is high. When the mevalonate pathway is inhibited, and thereby the GFP-CaaX protein cannot become prenylated but is localized to the cytoplasm and nucleus, the Dpeak value is low. When analyzing the HaCaT cells where only part of the population is expressing the GFP-CaaX, we can exclude the non-expressing cells from the image analysis. The plasma membrane trafficking analysis module provides us additional information as shown in table 2. Table 2.

Calculation of the results

Data generated by the IN Cell Analyzer 3000 software were either exported to Excel or the data were analyzed with the Genedata Screener software. Averages and standard deviations of the parallel samples were calculated. The IC50 values were calculated either using XLFit, or the IC50 curves were fitted using the Condoseo module of the Genedata Screener software, z' value was calculated from the following formula:

3 x σhigh + 3 x σi_ow

Z' = 1 averagehigh - averagei_ow

where σ is standard deviation.

Percent inhibition of the mevalonate pathway was calculated from the formula:

_M ■ _{u u}-..- Dpeak of compound x 100% _{i nn(}«

% inhibition = -£ ; — ~A_cn . , - 100%

Dpeak of DMSO control

Microscopy

Living and fixed cells were visually inspected and photographed using Nikon Eclipse TE2000 fluorescence microscope equipped with DXM 1200 digital camera. C-FL EPI-FL filter block (ex340-380 nm, em 435-485 nm) was used for DAPI staining and C-FL EPI-FL filter block GFP-B (ex460-500 nm, em 510-560 nm) for eGFP.

Results

Construction of plasmids expressing GFP with a wild-type and a mutated prenylation tag In order to follow the activity of FPPS and the prenylation status of proteins within cells, we constructed the plasmid pHE869 encoding GFP with a C-terminal tag for prenylation and double palmitoylation. A DNA sequence coding for the peptide GCMSCKCVLS (SEQ ID NO: 1 ) was added to the last codon of eGFP using a BamHI site which created a Gly-Ser linker between GFP and the peptide. Hereafter this constructed is referred as GFP-CaaX. This peptide is the C-terminus of human H-Ras protein (AF493916). The palmitoylation signals were included since prenylation alone causes only weak attachment of the protein to membranes and palmitoylation is required for the localization of the protein to the PM (Choy et al., 1999;Hancock et al., 1989;Willumsen et al., 1996). In the pHE877 control plasmid, GFP was tagged with a mutated palmitoylation and prenylation signal, GCMSCKSVLS (SEQ ID NO:8), where the critical cysteine for prenylation was replaced by a serine residue (hereafter referred as GFP-SaaX). The mutated tag is neither prenylated, nor palmitoylated, since palmitoylation follows prenylation and proteolysis of the last three residues of the tag (Choy et al., 1999). The tagged GFPs were under the CMV promoter in pcDNA3 vector. To confirm the expression of the two tagged GFPs and their subcellular localization, the plasmids pHE869 and pHE877 were transiently transfected in CHO K1 cells and analyzed with fluorescence microscopy. As expected GFP-CaaX localized mainly to the PM and GFP- SaaX to the cytoplasm (Fig. 3).

Choosing a cell line for the assay

To facilitate imaging and localization studies, cells that form a monolayer and do not grow on top of each other are preferably used. We tested for this purpose HaCaT cells, which are human keratinocytes (Boukamp et al., 1988). These cells were transiently transfected with pHE869 and pHE877 to analyze the localization of GFP-CaaX and GFP-SaaX. Fig 4 shows that GFP-CaaX was localized to membranes, mainly PM, but also probably to Golgi and nuclear membrane (Choy and Philips, 2001 ). GFP-SaaX appeared soluble.

To ascertain that localization of GFP-CaaX to membranes is prevented by inhibitors of the mevalonate pathway, we incubated HaCaT cells transiently transfected with pHE869 with 20 μM mevastatin for 19 h and 25 h and then analyzed the localization of GFP-CaaX. Mevastatin inhibits HMG CoA reductase, the first enzyme of the mevalonate pathway. Mevastatin caused cytosolic localization of GFP-CaaX (Fig. 5) and with 25 h incubation also apoptosis was observed. Change in localization of GFP-CaaX always preceded apoptosis, which was an important feature allowing us to discriminate between inhibition of FPPS and other toxic effects of compounds.

Cell lines expressing stably the GFP-CaaX would be desirable in order to use the assay in screening. Performing transient transfection in each screen is cumbersome, costs time and money. More importantly, with transient transfection only part of the cells is expressing the reporter protein, and the expression level varies from cell to cell. This causes a lot variation in the results.

Thus, we linearized the plasmid pHE869 using BgIII restriction enzyme, which cuts the vector backbone upstream of the CMV promoter. We used the linearized plasmid to transfect HaCaT cells, and used G418 as a means to select for the stably transfected cells. HaCaT cells do not grow well, when diluted too much. Therefore, it is difficult to obtain cellular clones by diluting the cells and plating only one cell / well. Instead we enriched the GFP-CaaX- expressing cells with cell sorting as described in Materials and Methods. After first round of sorting there were still a lot cells not expressing GFP-CaaX. Therefore we repeated the cell sorting for the cells already once sorted. With these means we got a population where less than 50% of cells expressed GFP-CaaX. An important factor was that expression was at appropriate level for imaging.

The response of the HaCaT cell population expressing GFP-CaaX to mevastatin and zoledronic acid was tested. The cells were incubated for 48 hours with different concentrations of mevastatin or zoledronic acid in quadruplicate samples, and the IC50 value and z' value were calculated. The results are shown in Tables 3 and 4. IC50 values 19 μM and 2.5 μM for mevastatin and zoledronic acid, respectively, were obtained, and Z' values were 0.37 and 0.45 for mevastatin and zoledronic acid, respectively. The Z' values are relatively good for an imaging assay, especially when only part of the cells are expressing the reporter construct. Moreover, we were unable to increase the cell number analyzed due to the inflexibility of the software, which does not allow us capture several images per well with this analysis module. Table 3. Raw data, averages and standard deviations, plus calculated % inhibition from the GFP-CaaX localization assay using mevastatin as an inhibitor of the mevalonate pathway.

Table 4. Raw data, averages and standard deviations, plus calculated % inhibition from the GFP-CaaX localization assay using zoledronate as an inhibitor of the mevalonate pathway.

Localization of GFP-CaaX in U-2 OS cells

In order to test the applicability of the GFP-CaaX localization assay in another cell line we chose the U-2 OS human osteosarcoma cells, since they are well suited for imaging assays. Moreover, they grow much faster than HaCaT cells. First we tested the effect of mevastatin and zoledronic acid on localization of GFP-CaaX in transiently transfected cells. Since these results were promising (not shown), we decided to make U-2 OS cells expressing stably GFP-CaaX. The plasmid pHE869 was linearized and transfected to the U-2 OS cells. From G418-resistant cells the GFP-expressing ones were enriched with cell sorting, and plated on 96-well plate to achieve one cell per well. From 30 GFP-CaaX-expressing cultures six were chosen to be analyzed in the localization assay using mevastatin and zoledronic acid as inhibitors of the mevalonate pathway. From each clone 4000 cells were seeded per well in 384-well format. Next day mevastatin or zoledronic acid was added. Two plates per clone were prepared, one to be incubated for 24 hours with the drugs and the other to be incubated for 48 hours with the drugs. Dose-response curves for mevastatin and zoledronic acid of the different clones have been assessed. The IC50 values obtained are summarized in table 5. Table 5. IC50 values (in μM) for mevastatin and zoledronic acid obtained with the different U-2 OS cell clones.

We chose clone 5 for further studies, since it had reasonable IC50 values for mevastatin and zoledronic acid. Moreover, the expression of GFP-CaaX was relatively uniform in the cells and it was localized mainly in the plasma membrane in untreated cells. The localization changed to cytosolic and nuclear by treatment with inhibitors of the mevalonate pathway. In clone 1 GFP-CaaX was only partially localized to the membranes. In clones 1 , 3, 4, 11 and 24 there was also much more variation in the expression level of GFP-CaaX between individual cells than in clone 5. High variation in expression level hampers the image analysis, since in overexpressing cells GFP-CaaX is localized to the cytoplasm in addition to the plasma membrane.

Testing inhibitors of the mevalonate pathway with the U-2 OS cell clone 5 We tested in the GFP-CaaX localization assay inhibitors of HMG CoA reductase (mevastatin), FPPS (zoledronic acid), farnesyltransferase (FTI276), and geranylgeranyltransferase (GGTI298). All compounds were tested in eight different concentrations, FTI276, GGTI298 and mevastatin having the highest concentration 100 μM. Zoledronic acid and mevastatin having the highest concentration 50 μM. Dilution factor was 3.16. All samples were in quadruplicate. DMSO, 0.5%, was used as a control. The cells were seeded on black 384-well plates with clear bottom, 1000 cells per well. Next day the compounds were added in a volume of 5 μl. Zoledronic acid was diluted in PBS, the other inhibitors in DMSO so that the end concentration of DMSO in the wells was 0.5%. The cells were incubated with the inhibitors for 48 hours. Then the cells were fixed and the nuclei stained with DRAQ5. The cells were imaged with IN Cell Analyzer 3000 and the images analyzed with the Plasma Membrane Trafficking module to get information about the localization of GFP-CaaX in the cells. Since inhibition of the mevalonate pathway can be proapoptotic and/or prevent cell proliferation we analyzed the images also with the Morphology Analysis module MPHO. dll to get information about the effect of the drugs on cell number and nuclear intensity. We have found cell number is one of the most sensitive measures for toxicity in a comparison of different toxicity assays (own unpublished results). When cells are undergoing apoptosis the nuclei become condensed, and this can be detected as an increased intensity of nuclear staining.

All the tested inhibitors of the mevalonate pathway prevented localization of GFP-CaaX to membranes (Figures 7-9). Mevastatin, zoledronic acid and GGTI298 were also toxic for the cells. They all reduced cell number and increased nuclear intensity. In all cases, however, the IC50 values for GFP-CaaX localization were slightly lower than for cell number or nuclear intensity. FTI276 did not show toxicity in our experiment (Table 6).

Table 6. The effect of different inhibitors of the mevalonate pathway on localization of GFP- CaaX, on cell number and nuclear intensity in U-2 OS clone 5 cells. The IC50 values are in μM. Mevastatin was analyzed in two different dilution series, starting from 50 μM or 100 μM. All samples were in quadruplicate.

The experiment was repeated by another person six weeks later. Now the effect of the drugs on GFP-CaaX localization was analyzed after 24 and 48 hour incubation. Also, an additional farnesyltransferase inhibitor, FTI277, was included. Dose-response curves for the 48 hour results have been assessed. The IC50 values for zoledronic acid and mevastatin were higher than in the first experiment, but the IC50 value for FTI276 was slightly lower. FTI276 and FTI277 had a clear effect already at the lowest concentration used. Therefore no IC50 value could be determined. Table 7 summarizes the results from the second experiment.

Table 7. IC50 values obtained for various inhibitors of the mevalonate pathway in the repetition of the experiment summarized in table 6.

Discussion

The mevalonate pathway leads to synthesis of cholesterol, farnesylpyrophosphate and geranylgeranylpyrophosphate. Inhibition or modulation of the mevalonate pathway can be used in treatment of high serum cholesterol levels, cancer, osteoporosis or restenosis. The lipids farnesylpyrophosphate and geranylgeranylpyrophosphate are attached to the CaaX motif of several small GTPases. These lipids are essential for the biological activity of the GTPases. Thus, by preventing the synthesis of the lipids via inhibition of the mevalonate pathway the activity of the small GTPases can be decreased.

The activity of the mevalonate pathway in cells has been studied until now with tedious low- throughput methods. We have developed a cellular imaging-based assay to monitor the status of the mevalonate pathway. We have fused the C-terminal peptide from human H-Ras to GFP. This peptide contains signals for prenylation and palmitoylation, which are critical for the targeting of the reporter protein to the plasma membrane. We have shown that with compounds inhibiting the mevalonate pathway at different stages the membrane-association of our reporter protein is inhibited, and the reporter becomes cytosolic/nuclear. We can quantitatively detect this using automated microscopy and automated image analysis.

Most inhibitors of the mevalonate pathway are expected to induce apoptosis or to prevent proliferation. This is due to inhibition of the activity of several small GTPases. Mevastatin and zoledronic acid, which block both synthesis of cholesterol and isoprenoid lipids indeed lead to apoptosis in our cellular assay. The farnesyltransferase inhibitors FTI276 and FTI277 prevented membrane localization of GFP-CaaX but were not found to induce apoptosis in U- 2 OS cells, whereas the geranylgeranyltransferase inhibitor GGTI-298 potently induced apoptosis. This difference results most probably from the fact that in U-2 OS cells more essential proteins are geranylgeranylated than farnesylated.

We observed that zoledronic acid and mevastatin first prevented membrane localization of our reporter protein GFP-CaaX, and only thereafter signs of apoptosis became apparent in the cells. Therefore it is possible to separate inhibitors of the mevalonate pathway from merely toxic compounds using the GFP-CaaX localization assay described above. I.e. the real inhibitors of the pathway can be discriminated from the merely toxic compounds by the fact that they would first induce cytosolic and/or nuclear localization of GFP-CaaX in healthy- looking cells, and only thereafter, or with higher compound concentrations, apoptosis would be detectable.

We analyzed the effect of various drugs on GFP-CaaX localization after 48 hour incubation. There are several reasons for the long incubation time. First, after inhibition of FPPS most of the isoprenoid lipids present at that time point in cells must be consumed before an effect on GFP-CaaX localization can be seen. Second the protein turnover time must be taken into account. A notable portion of the GFP-CaaX must get degraded and new non-prenylated protein must be synthesized before the effect on GFP-CaaX localization can be seen. Untagged GFP is very stable, its half-life is about 26 h (Corish and Tyler-Smith, 1999). Whether the prenylation affects the half-life of GFP to one direction or another, is not known.

By using the tagged GFP instead of endogenous small GTPases or GFP fused to a full- length small GTPase we can better monitor the compounds modulating the mevalonate pathway. The localization of the small GTPases is affected not only by the lipids at the C- terminus, but also by their activation status. Thus, hits found by using full-length small GTPases could affect also the GTPases themselves or any upstream signaling molecule instead of the mevalonate pathway.

References

1. Bergstrom,J.D., Bostedor,R.G., Masarachia,P.J., Reszka,A.A., and Rodan,G. (2000). Alendronate is a specific, nanomolar inhibitor of farnesyl diphosphate synthase. Arch.Biochem.Biophys. 373, 231-241.

2. Boukamp,P., Petrussevska,R.T., Breitkreutz,D., Hornung,J., Markham,A., and Fusenig,N.E. (1988). Normal keratinization in a spontaneously immortalized aneuploid human keratinocyte cell line. J. Cell Biol. 106, 761-771.

3. Chalfie,M., Tu,Y., Euskirchen,G., Ward,W.W., and Prasher,D.C. (1994). Green fluorescent protein as a marker for gene expression. Science. 263, 802-805.

4. Choy,E., Chiu,V.K., Silletti, J ., Feoktistov,M., Morimoto,T., Michaelson,D., IvanovJ.E., and Philips, M. R. (1999). Endomembrane trafficking of ras: the CAAX motif targets proteins to the ER and Golgi. Cell 98, 69-80.

5. Choy,E. and Philips, M. (2001 ). Green fluorescent protein-tagged Ras proteins for intracellular localization. Methods Enzymol. 332:50-64., 50-64.

6. Cohen, L. H., Pieterman,E., van Leeuwen,R.E., Overhand, M., Burm,B.E., van der Marel,G.A., and van Boom, J. H. (2000). Inhibitors of prenylation of Ras and other G-proteins and their application as therapeutics. Biochem. Pharmacol. 60, 1061-1068.

7. Conte,P. and Guarneri,V. (2004). Safety of intravenous and oral bisphosphonates and compliance with dosing regimens. Oncologist. 9 Suppl 4:28-37., 28-37.

8. Corish,P. and Tyler-Smith, C. (1999). Attenuation of green fluorescent protein half-life in mammalian cells. Protein Eng 12, 1035-1040.

9. Der,C.J., Krontiris,T.G., and Cooper,G.M. (1982). Transforming genes of human bladder and lung carcinoma cell lines are homologous to the ras genes of Harvey and Kirsten sarcoma viruses. Proc.Natl.Acad.Sci.U.S.A 79, 3637-3640.

10. Dunford,J.E., Thompson, K., Coxon,F.P., Luckman,S.P., Hahn,F.M., Poulter,C.D., Ebetino,F.H., and Rogers, M.J. (2001 ). Structure-activity relationships for inhibition of farnesyl diphosphate synthase in vitro and inhibition of bone resorption in vivo by nitrogen-containing bisphosphonates. J. Pharmacol. Exp.Ther. 296, 235-242.

11. Green,J.R. (2004). Bisphosphonates: preclinical review. Oncologist. 9 Suppl 4:3-13., 3-13. 12. Hancock,J.F., Magee,A.I., Childs,J.E., and Marshall, C. J. (1989). All ras proteins are polyisoprenylated but only some are palmitoylated. Cell 57, 1167-1177.

13. Heim,R., Prasher,D.C, and Tsien,R.Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc.Natl.Acad.Sci. U.S.A. 91, 12501-12504.

14. Keller,R.K. and Fliesler,S.J. (1999). Mechanism of aminobisphosphonate action: characterization of alendronate inhibition of the isoprenoid pathway. Biochem.Biophys.Res.Commun. 266, 560-563.

15. Keppler,A., Arrivoli,C, Sironi,L, and Ellenberg,J. (2006). Fluorophores for live cell imaging of AGT fusion proteins across the visible spectrum. Biotechniques. 41, 167-5.

16. Keppler,A., Pick, H., Arrivoli,C, Vogel,H., and Johnsson,K. (2004). Labeling of fusion proteins with synthetic fluorophores in live cells. Proc.Natl.Acad.Sci. U.S.A. 101, 9955-9959.

17. Los,G.V., Darzins,A., Zimprich,C, Karassina,N., Learish,R., McDougall,M.G., Encell,L.P., Friedman-Ohana,R., Wood, M., Vidugiris,G., Zimmerman, K., Otto, P., Klaubert,D.H., and Wood, K. (2005) One fusion protein: Multiple functions. Promega Notes Magazine 89, 2 — 5.

18. Luckman,S.P., Coxon,F.P., Ebetino,F.H., Russell, R.G., and Rogers, M.J. (1998). Heterocycle- containing bisphosphonates cause apoptosis and inhibit bone resorption by preventing protein prenylation: evidence from structure-activity relationships in J774 macrophages. J. Bone Miner.Res. 13, 1668-1678.

19. Lundholt,B.K., Linde,V., Loechel,F., Pedersen,H.C, Moller,S., Praestegaard,M., Mikkelsen,!., Scudder,K., Bjorn,S.P., Heide,M., Arkhammar,P.O., Terry,R., and Nielsen,S.J. (2005). Identification of Akt pathway inhibitors using redistribution screening on the FLIPR and the IN Cell 3000 analyzer. J. Biomol. Screen. 10, 20-29.

20 Matz,M.V., Fradkov,A.F., Labas,Y.A., Savitsky,A.P., Zaraisky,A.G., Markelov,M.L., and Lukyanov,S.A. (1999). Fluorescent proteins from nonbioluminescent Anthozoa species. Nat.Biotechnol. 17, 969-973.

21. Needleman,S.B. and Wunsch,C.D. (1970). A general method applicable to the search for similarities in the amino acid sequence of two proteins. J. MoI. Biol. 48, 443-453.

22. Pearson, W.R. and Lipman,D.J. (1988). Improved tools for biological sequence comparison. Proc.Natl.Acad.Sci.U.S.A. 85, 2444-2448. 23. Perrin,D. and Hill, BT. (2000). Tomorrow's anticancer agents: inhibitors of Ras farnesylation. EXS 89:153- 79., 153-179.

24. Roskoski,R., Jr. (2003). Protein prenylation: a pivotal posttranslational process. Biochem.Biophys.Res.Commun. 303, 1-7.

25. Scharovsky,O.G., Rozados,V.R., Gervasoni,S.I., and Matar,P. (2000). Inhibition of ras oncogene: a novel approach to antineoplastic therapy. J.Biomed.Sci. 7, 292-298.

26. Schindler,J.F., Naranjo,P.A., Honaberger,D.A., Chang, C. H., Brainard,J.R., Vanderberg,L.A., and Unkefer,C.J. (1999). Haloalkane dehalogenases: steady-state kinetics and halide inhibition. Biochemistry. 38, 5772-5778.

27. Smith,T.F. and Waterman, M. S. (1981 ). Overlapping genes and information theory. J.Theor.Biol. 91, 379-380.

28. Swanson,K.M. and Hohl,R.J. (2006). Anti-cancer therapy: targeting the mevalonate pathway. Curr.Cancer Drug Targets. 6, 15-37.

29. Tamanoi,F., Gau,C.L, Jiang, C, Edamatsu,H., and Kato-Stankiewicz,J. (2001 ). Protein farnesylation in mammalian cells: effects of farnesyltransferase inhibitors on cancer cells. Cell MoI. Life Sci. 58, 1636-1649.

30. Thompson, K., Rogers, M.J. , Coxon,F.P., and Crockett,J.C. (2006). Cytosolic entry of bisphosphonate drugs requires acidification of vesicles after flu id-phase endocytosis. Mol.Pharmacol. 69, 1624-1632.

31. van Beek,E., Pieterman,E., Cohen, L., Lowik,C, and Papapoulos,S. (1999). Farnesyl pyrophosphate synthase is the molecular target of nitrogen-containing bisphosphonates. Biochem.Biophys.Res.Commun. 264, 108-1 11.

32. van Beek,E.R., Lowik,C.W., Ebetino,F.H., and Papapoulos,S.E. (1998). Binding and antiresorptive properties of heterocycle-containing bisphosphonate analogs: structure-activity relationships. Bone 23, 437-442.

33. Willumsen,B.M., Cox,A.D., Solski,P.A., Der,C.J., and Buss,J.E. (1996). Novel determinants of H-

Ras plasma membrane localization and transformation. Oncogene 13, 1901-1909.

Claims

1. A method of screening for modulators of the mevalonate pathway, said method comprising: a. culturing eukaryotic cells expressing a protein comprising a prenylation signal in the presence or in the absence of a test compound; and, b. comparing the subcellular localization of said protein with a prenylation signal in the presence of a test compound with said subcellular localization in the absence of the test compound; wherein a change in the subcellular localization of said protein with a prenylation signal is indicative that said test compound is a modulator of the mevalonate pathway.

2. The method of Claim 1 , wherein said prenylation signal is a peptide comprising (i) the CaaX motif, wherein "C" stands for cystein, "a" for an aliphatic amino acid and the X is selected among the group consisting of a methionine, serine, alanine, glutamine and leucine and, optionally, (ii) a palmitoylation motif.

3. The method of Claim 1 , wherein said prenylation signal is a 4-50 amino acids peptide.

4. The method of Claim 1 , wherein said prenylation signal is a peptide fragment of the C-terminus of human H-Ras consisting of SEQ ID NO:1 or any suitable variant.

5. The method of Claim 1 , wherein a change in the subcellular localization of said protein with a prenylation signal from the plasma membrane to the cytosol in the presence of a test compound is indicative that said test compound is capable of inhibiting the mevalonate pathway.

6. The method of Claim 1 , wherein the subcellular localization of said protein with prenylation signal is detected by monitoring non-tagged protein via immunodetection.

7. The method of Claim 1 , wherein said prenylation signal is fused to a luminophore or a protein tag such as Halo- or SNAP-tag.

8. The method of Claim 7, wherein said luminophore is a fluorophore, for example, a green fluorescent protein.

9. The method of Claim 8, wherein said protein with a prenylation signal essentially consists of the green fluorescent protein fused at its C-terminal to SEQ ID NO:1.

10. The method of Claim 8 or 9, wherein the subcellular localization of the prenylation signal fused to the fluorophore is detected quantitatively either directly by microscopy, or indirectly by using an extracellular quencher and a suitable fluorescence reader.

11. The method of Claim 1 , wherein said eukaryotic cells are mammalian cell lines, such as HaCaT cells, HeLa, MDCK, U-2 OS, CHO-K1 , or primary cells.

12. The method of Claim 1 , wherein said test compound is first selected among those which bind to or modulate a target selected among the group consisting of FPPS enzyme, HMG CoA reductase or farnesyltransferase in a non cellular assay.

13. The method of Claim 1 , wherein the subcellular localization is detected at least in live or fixed cells at a time point when the cells still have normal morphology and would have not started apoptosis if the compound is toxic in order to discriminate among an apparent cytosol localization due to apoptosis and a plasma membrane localization due to inhibition of the mevalonate pathway.

14. The method of Claim 1 , wherein the subcellular localization is detected at least in live or fixed cells, at a compound concentration with which the cells still have normal morphology and do not start apoptosis despite toxicity of the compounds at a higher concentration in order to discriminate among an apparent cytosol localization due to apoptosis and prevention of plasma membrane localization due to inhibition of the mevalonate pathway

15. A eukaryotic cell line, stably transfected with a nucleic acid construct encoding a C- terminal fusion of a prenylation signal with a green fluorescent protein.

16. The eukaryotic cell line of Claim 15, which is selected among the group consisting of HaCaT and U-2 OS cell lines.

17. Use of the cell lines as defined in any of Claims 15-16 in a cellular screening assay for identifying modulators of the mevalonate pathway.

18. Use of the cell lines as defined in any of Claims 15-16 in the method of any of Claims 1-14.