WO2005113770A1 - Anti-rhoa and -rhoc sirnas and therapeutic compositions comprising them. - Google Patents

Abstract

The present invention pertains to the use of anti-RhoA and -RhoC siRNAs in order to inhibit specifically RhoA or RhoC synthesis. Isolated double-stranded RNA molecules which are able to mediate interference of the human RhoA and RhoC mRNAs, in particular to inactivate these genes by transcriptional silencing, are disclosed. The invention also relates to therapeutical compositions comprising such anti-RhoA and / or anti-RhoC siRNAs, as well as their use for the manufacture of a therapeutical composition, in particular against aggressive breast cancer.

Description

ANTI-RhoA AND -RhoC siRNAS AND THERAPEUTIC COMPOSITIONS COMPRISING THEM The present invention pertains to the field of anti-cancer therapy. More particularly, the invention is based on the demonstration, by the inventors, that the inhibition of rhoA and/or rhoC in cancer and/or angiogenic endothelial cells could inhibit tumor growth and tumoral angiogenesis . The present invention hence relates to anti- cancercompositions, comprising means to inhibit the expression of RhoA and/or RhoC. Low molecular weight GTP/GDP binding GTPases of Ras superfamily, RhoA and RhoC, have been shown to promote cell proliferation and cell invasion (Aznar and Lacal, 2001; Schmitz et al., 2000). Accumulating studies indicate that Rho protein-dependant cell signalling might be important for malignant transformation (Denoyelle et al., 2003; Frame and Brunton, 2002; Price and Collard, 2001). Once activated, protein RhoA triggers a complex set of signal transduction pathways, which include the activation of ROCK, responsible for actin polymerization required for cell locomotion, and the phosphatidyl inositol 3 phosphokinase/AKT pathway thought to be critical for cell survival and expression of genes involved in cell proliferation (Denoyelle et al., 2003). In several human cancers, RhoA is over-expressed during tumor, and this correlates with poor prognosis (Fritz et al., 2002; Kamai et al., 2003; Kamai et al., 2004; Sahai and Marshall, 2002). RhoC, another low molecular weight GTPase, was also shown to be involved in cancer invasion in melanoma (Fritz et al., 1999), inflammatory breast cancer (Clark et al., 2000; Kleer et al., 2002), and ovarian cancer (Horiuchi et al., 2003). Several reports also demonstrate the implication of RhoA in angiogenesis, which participates to cancer growth and metastasis (Abecassis et al., 2003; Hoang et al., 2004; Liu and Senger, 2004; Vincent et al . , 2001; Vincent et al., 2002). The low molecular weight GTPases require prenylation for their biological activity (Seabra, 1998) . The prenylated derivatives, the farnesyl pyro-phosphate (FPP) and the geranyl-geranyl-pyro-phosphate (GGPP) which is a precursor of cholesterol, are respectively required for Ras- and Rho-regulated cell signalling (Malaney and Daly, 2001; Sahai and Marshall, 2002) . The prenylation of Ras and Rho is controlled by enzymes farnesyl transferase and geranyl-geranyl-transferase . As Ras has been found to be activated in more than 30% of cancers due to mutation and RhoA is over-expressed and spontaneously activated in several aggressive cancers, farnesylation inhibitors (FTI) (Haluska et al., 2002; Li and Sparano, 2003) and geranyl- geranyl transferase inhibitors (GGTI) (Lobell et al., 2001; Sebti and Hamilton, 2000) have been produced as anticancer agents. Interestingly, FTI activity does not necessarily act on Ras mutational status (Dy and Adjei, 2002), indicating that Ras is not the only FTI target and other farnesylated proteins may be involved in oncogenesis. It has been proposed that FTI recruit the antioncogenic RhoB protein via RhoB-GG (Prendergast , 2001) . Inhibition of Ras and Rho signalling could also be achieved by statins (such as cerivastatin) which are inhibitors of hydroxy-methyl-glutaryl coenzyme A reductase (HMGCoA reductase) because this enzyme is required for biosynthesis of cholesterol and its prenylated precursors (FPP and GPP) (Denoyelle et al., 2003; Vincent et al., 2001). In a variety of aggressive breast cancers, such as MDA-MB 231 cells (with a constitutive activation of Ras and overexpression of RhoA and RhoC) , cerivastatin induces in vi tro a strong inhibition of aggressive breast cells proliferation and cell invasion (Denoyelle et al., 2001). The inhibitory effect of statin has also been reported in other types of cancer cells (Kusama et al., 2002). Cerivastatin was also shown to inhibit angiogenesis (Vincent et al., 2001; Vincent et al., 2002). These inhibitory effects on cancer cells and endothelial cells were related to the inhibition of two pathways. The first effect was the inhibition of Rho-associated coiled-coil-containing protein kinase (ROCK) which was involved in cells spreading and cells motility and was activated during tumor invasion (Imamura et al., 2000; Sahai and Marshall, 2002; Sahai and Marshall, 2003; Vincent et al., 2003; Vincent et al., 2002) . The second effect was the inhibition of the cell signal pathway RhoA/Fak/PI3 Kinase/AKT. This inhibition might be crucial for an anti-cancer action as the activation of this pathway induces the activation of 3 different transcriptional factors involved in gene expression implicated in cancer cell proliferation NF- KappaB, API and β-catenin (Denoyelle et al., 2003). In addition, these inhibitory effects on cancer cells and endothelial cells were due to inhibition of Rho-induced cell signalling and not by Ras inhibition, as it was reversed by geranyl-geranyl-pyrophosphate (GGPP) but not by farnesyl pyrophosphate (FPP) (Denoyelle et al., 2003; Vincent et al., 2001; Vincent et al., 2002). In this context, the inventors have investigated the anticancer action of specific RhoA and RhoC inhibitors and compared their effects with that of cerivastatin, FTI and GGTI in MDA-MB 231 cells and endothelial cells. In an attempt to inhibit specifically RhoA and RhoC, both involved in cancer aggressivity, they used chemically synthesized short interfering RNA (siRNA) , i.e., small double-stranded oligonucleotides which can specifically and effectively direct homology-dependent post-transcriptional gene silencing. The siRNA triggers the degradation of the endogenous mRNA to which the siRNA hybridizes (Elbashir et al., 2001; Tuschl, 2002). This mechanism is specific and switches off the expression of the target protein. Importantly, siRNA is able to penetrate into tumors, for example using cytofectin as a delivery system (Bertrand et al., 2002). However, intravenous administration is more convenient for clinical use than intra-tumoral injection, and the IV route is likely to enable the siRNA to reach both primary tumor and metastasis. Therefore, the efficacy and the toxicity of IV administration of encapsulated anti-RhoA siRNA have also been investigated, using chitosan-coated nanoparticles as carrier. Chitosan is deacylated derivative of chitin, which is one of the most abundant mucopolysaccharide in crustaceans and insects. It is used as a non-viral delivery system which presents several advantages including low cost, non-infectivity, absence of immunogenicity, and the possibility of repeated clinical administration (Pouton and Seymour, 2001). Nanoparticles entirely made of chitosan have already been described (Borchard, 2001) . The electrostatic interactions between the negatively charged siRNA and cationic polymer result in the formation of a complex (speckled particle) that is protected by exposed steric polymer strands. Chitosan is not toxic, easily biodegradable and was shown to protect nucleotides from enzymatic digestion both in vivo and in vi tro (Mao et al., 2001; Nsereko and A iji, 2002; Quong and Neufeld, 1998); moreover, the presence of serum does not interfere with the transfection ability (Mao et al., 2001). For these reasons, chitosan nanoparticles containing reporter genes are being used in the literature for the transfection of mammalian cells both in vi tro and in vivo (Mansouri et al., 2004; Mao et al., 2001). Chitosan-coated polyisohexylcyanoacrylate (PIHCA) nanoparticles (Bioalliance, WO 2004/000287) represent a novel device for oligonucleotides transfer. In these nanoparticles, the chitosan used is 5000 Da. The inventors have tested the efficacy of chitosan-coated nanoparticles loaded with anti-RhoA siRNA administered via the IV route, as antitumor therapy of xenografted aggressive breast tumor. The duration of its effect was also investigated. Indeed, chitosan-nucleotides nanoparticles may enable intracellular sustained release of nucleotides (Dastan and Turan, 2004); this might be beneficial for a more prolonged and controlled expression of siRNA. However, the duration of drugs release depends on the drug tested (it can be complete in some hours while it can continue for several months with DNA) . The in vivo efficacy of these nanoparticles as antitumor drug was determined by measuring growth of xenografted aggressive breast cancer (MDA-MB 231) in mice. As there is no data regarding the toxicity of anti-RhoA siRNA, clinical, biochemical and histopathological parameters were investigated to test nephro-, hepato- and muscle- toxicities in anti-RhoA treated mice. As shown below and in (Pille et al . , 2005), the inventors have demonstrated that intra-tumoral injections of anti-RhoA or anti-RhoC siRNA (100 μl, 85 nM) every 3 days for 20 days, using cytofectin for delivery, almost totally inhibited the growth and angiogenesis of xenografted MDA MB-231 tumors. Surprisingly, the effect observed with these siRNAs was greater than that observed with prenylation inhibitors. The efficacy and the absence of toxicity of the anti-RhoA siRNA delivered through chitosan nanoparticles administered by intravenous route has also been demonstrated. The invention hence pertains to an isolated polynucleotide, which is able to inhibit RhoA or RhoC expression by transcriptional interference. By "transcriptional interference" is meant the RNAi mechanism described, for example, by Hutvagner and Zamore (Hutvagner and Zamore, 2002) . In particular, the invention concerns an isolated double-stranded RNA molecule, such as a small interfering RNA (siRNA) , which is able to mediate interference of the human RhoA or RhoC mRNA, in particular to inactivate the RhoA or RhoC gene by transcriptional silencing . In a preferred embodiment, an isolated polynucleotide according to the invention is able to down- regulate RhoA or RhoC mRNA and protein levels by at least 30%, preferably at least 50%, and more preferably at least 80% in cultured MDA-MB 231 cells or in cultured HMEC-1 cells, when said cells are transfected twice at 24 hours interval with 8.5 nM of said double-stranded RNA molecule. Materials and methods appropriate for testing the down- regulation of RhoA and RhoC expression are described in the examples below. Polynucleotides according to the invention also preferably have some functional characteristics that can be tested in vi tro . In particular: - double transfection of cultured MDA-MB 231 cells or cultured HMEC-1 cells, at 24 hours interval, with 8.5 nM of a polynucleotide according to the invention, leads to an inhibition of cell proliferation which is greater than the inhibition observed upon treatment of said cells with GGTI at lOμM, or FTI at lOμM; and/or - double transfection of cultured MDA-MB 231 cells, at 24 hours interval, with 8.5 nM of a polynucleotide according to the invention, leads to an inhibition of cell invasion through matrigel which is greater than the inhibition observed upon treatment of said cells with GGTI at lOμM, or FTI at lOμM; and/or double transfection of cultured HMEC-1 cells, at 24 hours interval, with 8.5 nM of a polynucleotide according to the invention, leads to an inhibition of capillary tube formation in a capillary tube formation assay on matrigel matrix; and/or double transfection of cultured HMEC-1 cells, at 24 hours interval, with 8.5 nM of a polynucleotide according to the invention, leads to an inhibition of β catenin nuclear localization. When the isolated polynucleotide according to the invention is an anti-RhoA (or anti-RhoC) siRNA, it is preferably in the form of a double-stranded RNA molecule, wherein each RNA strand has a length from 19 to 25 nucleotides, more preferably 19 to 23 nucleotides, and even more preferably 20 to 22 nucleotides. In a preferred embodiment of such siRNAs, one or both strand (s) has (have) a 3' overhang from 1 to 5 nucleotides, preferably from 1 to 3 nucleotides, and for example, 2 nucleotides. In a particular embodiment, each 3' overhang is stabilized against degradation. For example, an isolated double- stranded RNA molecule as described above is a chemically synthesized RNA, or an analogue of a naturally occurring RNA. Of course, the isolated polynucleotide according to the invention can also be a DNA that is transcribed into anti-RhoA (or anti-RhoC) siRNA in mammalian cells. For example, it can be a plasmid comprising a palindromic sequence that will be transcribed in the cell into a single-strand RNA molecule designed so that it forms a short hairpin RNA (shRNA) . In the cell, this shRNA will be processed into siRNA (Yu et al., 2002). A number of protocols are now provided to help the skilled artisan design a sequence for a shRNA. For example, mention can be made of the instruction manuals for the pSilencer™ vectors (Ambion® website: http: //www . ambion. com) . Preferred isolated polynucleotides according to the invention are the double-stranded RNA molecules used in the experiments disclosed in the examples below, i . e . :

Anti-RhoA siRNA: sense 5' GACAUGCUUGCUCAUAGUCTT 3' (SEQ ID No: 1) antisense 3' TTCUGUACGAACGAGUAUCAG 5' (SEQ ID No: 2), and

Anti-RhoC siRNA: sense 5' GACCUGCCUCCUCAUCGUCTT 3' (SEQ ID No: 3) antisense 3' TTCUGGACGGAGGAGUAGCAG 5' (SEQ ID No: 4). Of course, these sequences are not limitative, and other siRNA sequences can be used to perform the present invention, as well as DNA sequences that can be transcribed into shRNAs which will subsequently be processed into said siRNAs. The isolated polynucleotides according to the invention can be used for the preparation of a pharmaceutical composition. Hence, another object of the present invention is a pharmaceutical composition comprising an isolated polynucleotide as described above, and a pharmaceutically acceptable carrier. A particular pharmaceutical composition according to the invention comprises a first isolated polynucleotide which is able to inhibit RhoA expression, a second isolated polynucleotide which is able to inhibit RhoC expression, and a pharmaceutically acceptable carrier. Alternatively, when both RhoA and RhoC inhibition is sought, the polynucleotides specific for each gene can be formulated in two separate pharmaceutical compositions, which will be administered to a patient in need thereof either simultaneously or sequentially, through the same route or not, depending on the physician's choice. The invention hence also pertains to a kit of parts comprising at least one pharmaceutical composition comprising an isolated polynucleotide as described above, specific for RhoA, and another pharmaceutical composition comprising an isolated polynucleotide as described above, specific for RhoC. In preferred embodiments of the pharmaceutical compositions according to the present invention, the polynucleotide is in the form of an injectable solution. Depending on the therapeutic indication, this solution is preferably appropriate for intravenous and/or for local administration - examples of local administrations are intratumoral administration and administration to the eye, in order to decrease retinal vascularization. According to a particular embodiment of the invention, the RhoA- (or RhoC-) down-regulating polynucleotide is comprised in a vector for introducing it into a mammalian cell. The skilled artisan will choose, amongst the wide variety of vectors described in the scientific literature, an appropriate vector, depending on the target cells, the mode of administration that is contemplated (ex vivo or in vivo, intravenous or intratumoral, ...) , the expected duration of expression of the polynucleotide, etc. Plasmids and viruses, such as lentiviruses, can be cited as non-limiting examples (Abbas-Terki et al., 2002; Hannon and Conklin, 2004). Such vectors can be used advantageously to deliver DNA that will subsequently be transcribed into shRNA, in the cells. Integrative vectors carrying a sequence that can be transcribed into an anti-RhoA or anti-RhoC shRNA, can advantageously be used to obtain a long-term expression, for example by local injection in tumors. Other carriers that can be used in the pharmaceutical composition, in complement or as an alternative to the above-mentioned vectors, comprise nanospheres or nanoparticles. In particular, chitosan nanoparticles such as, for example, the chitosan-coated PIHCA nanoparticles described in WO 2004/00287, can be used to perform this aspect of the invention. The molecular weight of chitosan in the nanoparticles which can be used as vehicles in the compositions according to the invention preferably ranges 2000 to 8000 Da. For example, chitosan of 5000 Da can be used in the pharmaceutical compositions according to the invention. A pharmaceutical composition according to the invention is preferably formulated so as to enable the administration to a patient in need thereof, of between 100 to 2000 μg of the anti-RhoA or anti-RhoC isolated polynucleotide (s) , per kg body weight, in one day. Greater amounts of polynucleotides according to the invention can also be administered to a patient, for example and the physician will be able to adjust the above quantities, depending on the patient, the type of cancer, and the administration route. In certain situations, for example for stability issues, it is preferable to prepare a pharmaceutical composition extratemporaneously . The present invention hence also concerns a kit of parts comprising at least one isolated polynucleotide as described above, and a pharmaceutical vector which enables the transfection of cells by said polynucleotide, when combined with it. The pharmaceutical compositions and kits of parts according to the invention can also further comprise an anticancer chemotherapeutic agent. As far as the therapeutic applications are concerned, an isolated polynucleotide according to the invention, which is specific for RhoA, will of course be advantageously used for the preparation of a pharmaceutical composition against a disease or condition associated with an over-expression of RhoA, and an isolated polynucleotide specific for RhoC will be advantageously used for the preparation of a pharmaceutical composition against a disease or condition associated with an over-expression of RhoC. The polynucleotides and compositions of the invention are particularly useful for the treatment or prophylaxis of an hyperproliferative condition, such as a solid tumor cancer, in particular aggressive breast cancer. The polynucleotides according to the present invention can also be advantageously used for the preparation of a pharmaceutical composition targeting metastasis in cancer, especially pharmaceutical compositions formulated for systemic administration. The present invention will be further illustrated by the additional description which follows, which refers to examples of obtention and use of siRNAs and compositions of the invention. It should be understood however that these examples are given only by way of illustration of the invention and do not constitute in any way a limitation thereof.

LEGENDS TO THE FIGURES Figure 1 : Inhibition of RhoA and RhoC expression by anti-RhoA and anti-RhoC siRNA at both transcriptional and protein levels. MDA-MB 231 cells were either untreated (lanes 1) or transfected with 8.5 nM anti-RhoA or anti-RhoC siRNA once (lanes 2) or twice at 24 hours interval (lanes 3). Assays were performed 24 h after the end of treatment. For mRNA analyses, cells were treated with (A) anti-RhoA siRNA or (B) anti-RhoC siRNA, and mRNA expression was followed by RT-PCR; β-actin served as an internal control. Western Blots were performed to assay RhoA protein expression after treatment with anti-

RhoA siRNA (C) , and RhoC protein expression after treatment with anti-RhoC siRNA (D) . Note that a double transfection almost completely inhibited RhoA and RhoC expression . Figure 2 : Antiproliferative and anti-invasive effects of anti-RhoA or anti-RhoC siRNA on MDA-MB 231 breast cancer cells in comparison with action of GGTI and FTI Panel A : a ction on MDA-MB 231 prolifera tion 1.5.10⁵ MDA-MB 231 cells were seeded and cells were counted in a particle counter after 24 hours or 48 hours of culture with a minimal concentration of SVF (2%) to assure viability of the cells. Control was done in the absence of agent. It was checked that cytofectin diluted 1/100 had not modified cell proliferation. FTI: in presence of 10 μM FTI-277 GGTI: in presence of 10 μM GGTI-298 Anti-RhoA siRNA: 8.5 nM of anti-RhoA siRNA (final concentration) was added at To and T 24 hours. Anti-RhoC siRNA: 8.5 nM of anti-RhoC siRNA (final concentration) was added at To and T 24 hours Results are expressed as the number of cells/well at indicated time. They represented the mean±SEM of 3 independent experiments (** p<0.01, *** p<0.001) . Panel B : action on MDA-MB 231 invasion . MDA-MB 231 cells were treated with anti-RhoA or anti-RhoC siRNA (8.5 nM) or with FTI-277 (10 μM) , GGTI- 298 (10 μM) for 24 hours before adding to Transwell. Controls were performed in the absence of agent for FTI and GGTI and in the presence of cytofectin diluted 1/100 for siRNA. Then, the cells were detached, resuspended in the presence of agents to be tested at same concentrations as used during the incubation time and seeded into the transwell coated with matrigel. After an 24 hour incubation at 37 °C , the cells in the upper part of the invasion chamber were gently detached, and cells which had traversed the filter (pore 8 μm) were counted by light microscopy after May Grundwald-Giemsa coloration. Results are expressed as % of invaded cells as a mean±SEM of 3 independent experiments (** p<0.01, *** p<0.001). Figure 3 : Inhibition of capillary tube formation by anti RhoA and anti RhoC siRNA on a matrigel matrix. HMEC-1 were transfected twice either with 8.5 nM anti RhoA or anti RhoC siRNA at 24 hours intervals. 4.10⁴ HMEC-1 control (incubated with the same concentration of cytofectin as used for siRNA transfection) were seeded on matrigel. After 9 hours, the formation of capillary tubes was photographed under an inverted light microscope at X 40 magnification. Note the cell rounding in HMEC-1 transfected with anti-RhoA siRNA, while no shape change was noted in anti-RhoC transfected cells. Figure 4 : Effect of anti-RhoA and anti-RhoC siRNA on nuclear β-catenin expression in MDA-MB 231 cells. MDA-MB 231 cells untreated (control) or incubated once or twice at 24 hours interval with 8.5 nM siRNA anti RhoA or anti RhoC were analysed. After indicated incubation (48 hours for one transfection and 60 hours for the double transfection), nuclear extracts were subjected to SDS-PAGE and Western blot analysis using an anti β-catenin monoclonal antibody, as indicated in method section. Figure 5: Effect of intratumoral injection of anti-RhoA or anti-RhoC siRNA on the growth and vascularisation of MDA-MB 231 tumors xenografted in the mice. 4.10⁶ MDA-MB 231 cells were subcutaneously grafted in the dorsa of athymic mice. After 16 days, mice received an intratumoral injection of anti-RhoA or anti-RhoC siRNA (lOOμl at 85 nM) or cytofectin alone for control. Injection was repeated every 3 days. 20 days post siRNA treatment, animals were sacrificed and tumors were removed for immuno-histological analysis. a: control tumor, b: anti RhoA siRNA-treated tumor and c: anti RhoC siRNA- treated tumors. Panel A : Tumor growth . The tumor sizes were measured every day until day 20. Note that in anti-RhoA siRNA-treated group, a complete inhibition of tumor growth was observed in 3 mice. Results are expressed as the tumor size in mm3 at indicated time. They represented the mean ± SEM of 8 animals for each group (** p<0.01, *** p<0.001 at day 20) . Panel B : In tra tumoral vascula iza tion . It was assessed by PECAM-1-immunostaining on paraffin-embeded MDA-MB 231 tumors sections. Figure is shown at high magnification (x400) . Figure 6: Effects of intravenous injections of anti-RhoA siRNA on the growth of MDA-MB-231 cells xenografted into Nude mice. 4 x 10⁶ MDA-MB-231 cells were injected s . c. into the upper hind limb of athymic nude mice and allowed to grow until the tumor reached 20 mm³. Chitosan-coated poly-isohexyl-cyanoacrylate (PIHCA) nanoparticles loaded with anti-RhoA siRNA were injected intravenously every 3 days at the dose of 150 μg/kg and 1500 μg siRNA/kg body weight in anaesthetized mice, in the retro-orbital vein. Control was done with saline. Tumor growth was assessed daily until day 30 of treatment by measuring two perpendicular diameters and calculating the volume in mm³. Mean ± SEM, 8 mice per group. ** p<0.01, *** p<0.001 (Mann-Whitney-test on day 30). Then, control group was euthanazied and treatment was stopped in 50% of siRNA treated animals to analyse the growth of tumor after the cessation of therapy. Figure 7 : intravenous injections of anti-RhoA siRNA inhibits MDA-MB tumor growth. Tumors representative of control and anti-RhoA siRNA-treated groups ( Intraveinous injection every 3 days at 1500 μg/Kg anti RhoA SiRNA) are shown 10 and 30 days after the beginning of therapy, and at day 50, i.e., 20 days after the end of the therapy. Note the necrotic area in the center of tumour in one tumor in a mouse-treated withl500μg anti RhoA siRNA-coated nanoparticles. Figure 8: Toxicology studies of RhoA siRNA in the mice by examining the body weight gains. Control group was constituted of 6 mice which were anesthezied every 3 days in order to be matched with treated group. Body weight was measured in control mice and in both treated groups (150 and 1500 μg/Kg) in animals with or without xenografted tumour. Figure 9 : Liver examination after 30 days treatment with saline (control) or with siRNA antiRhoA- coated nanoparticles at 150 and 1500 μg/Kg

EXAMPLES

EXAMPLE 1: Anti-RhoA and Anti-RhoC siRNAs inhibit the proliferation and invasiveness of MDA-MB-231 Breast Cancer cells in vitro and in vivo Material and methods

Cells Cultures MDA-MB 231, an aggressive human breast carcinoma cell line, was grown in RPMI-1640 medium (Eurobio, Les Ulis, France) supplemented with 10% heat- inactivated foetal calf serum (FCS, Costar, Brumath, France) , 2 mM L-glutamine (Gibco BRL Life Technologies, Grand Island, NY) , and 100 IU/ml penicillin/streptomycin (Sarbach, Suresnes, France/Diamant , Puteaux, France). Cells were cultured at 37°C in a humidified 5% C02 atmosphere. The human microvascular endothelial cell-1 (HMEC-1) line was kindly provided by Dr Ades (Center for Disease Control and Prevention, Atlanta, GA. ) , who established this cell line by transfecting human dermal endothelial cells with SV40 A gene product and large T- antigen (Ades et al., 1992). These cells have properties similar to those of the original primary micro vascular endothelia cells (Bouis et al., 2001). Moreover, HMEC-1 cells proliferation, migration, and capillary tube formation are greatly increased by angiogenic factors such as bFGF and vascular endothelial growth factor (VEGF) (Vincent et al . , 2001; Vincent et al., 2002). HMEC-1 cells were cultured in a complete MCDB131 medium (Sigma, Paris, France) supplemented with 15% foetal calf serum (FCS), 100 IU/ml penicillin, 100 μg/ml streptomycin, 10 ng/ml epidermal growth factor (EGF) (Euromedex, Souffelweyersheim, France) and 1 μg/ml hydrocortisone (Pharmacia-Upjohn, St Quentin en Yvelines, France). HMEC-1 cells were used before the 15th passage, because after 25 passages, changes in the morphology and functions of endothelial cells would make them unsuitable for angiogenesis assessment. SiRNA treatment Two siRNA against human RhoA or Rho C were designed. The coding sequences of RhoA and RhoC were scanned to identify AA(N19)TT sequences. Candidate sequences were compared with the cDNA sequences of RhoA and RhoC and also searched against the non-redundant human DNA database using a BLAST algorithm (accession by international NCBI) . A control siRNA was also tested. It was selected because it exhibited no cellular toxicity. The sequences selected for the sense and antisense strands are: For the anti-RhoA siRNA sense 5' GACAUGCUUGCUCAUAGUCTT 3' (SEQ ID No: 1) antisense 3' TTCUGUACGAACGAGUAUCAG 5' (SEQ ID No: 2) For the anti-RhoC siRNA sense 5' GACCUGCCUCCUCAUCGUCTT 3' (SEQ ID No: 3) antisense 3' TTCUGGACGGAGGAGUAGCAG 5'. (SEQ ID No: 4) For the control siRNA sense 5' CAGUCAGGAGGAUCCAAAGTG 3' (SEQ ID No: 5) antisense 3' TTGUCAGUCCUCCUAGGUUUC 5' (SEQ ID No: 6) They were synthesized as synthetic oligonucleotides by Eurogentech (Belgium) and annealed to form a short double-stranded RNA with a 3 ' -dithymidine overhang. Hybridization was performed in a buffer containing 2 mM sodium acetate, 100 mM potassium acetate and 30 mM Hepes buffer, pH 7.4. The siRNAs were introduced into MDA-MB 231 or HMEC-1 cells by cytofectin-mediated transfection (Ozyme, Paris, France) using the method and siRNA concentration described by Bertrand et al. (Bertrand et al., 2002). Cells were cultured in 6-well plates in 200 μl serum- enriched medium. When confluence reached 50 %, 20 μl of 85 nM siRNA (final concentration) in 1/100 diluted cytofectin were added dropwise to the cell cultures and incubated with the cells for indicated time.

Total RNA Extraction and Reverse Transcriptase-Polymerase Chain Reaction Assay (RT-PCR) . Briefly, after siRNA transfection, cells were detached with the non enzymatic cell dissociation solution (CDS, Sigma) and washed twice in PBS. Total RNA extraction was performed using the "SV total RNA isolation system" (Promega, Charbonnieres-les-Bains, France) according to the manufacturer's instructions. Primers were chosen using a biomolecular sequences databases (Genbank) and oligonucleotides used as primers were synthesized by Genset (Paris, France); the sequences were as follows: For RhoA :

Forward primer: 5' -AGAGGTGTATGTGCCCACAGT-3' (from position 93-113 bp) (SEQ ID No: 7)

Reverse primer: 5' -CTTCGGAATGATGAGCACAC-3' (from position of 361-380 bp) (SEQ ID No: 8). For RhoC :

Forward primer: 5' -GGAGGTCTACGTCCCTACTGT-3' (from position of 93-113 bp) (SEQ ID No: 9)

Reverse primer: 5' -TACCCGGACACTGATGTCATC-3' (from position of 220-240 bp) (SEQ ID No: 10) For β-actin :

Forward primer: 5' -ATCTGGCACCACACCTTCTACAATGAGCTGCG-3' (from position of 253-284 bp) (SEQ ID No: 11) Reverse primer: 5' -CGTCAACTCCTGCTTGTTGATCCACATCTGC-3' (from position of 1049-1080 bp) (SEQ ID No: 12). The predicted sizes for RhoA, RhoC and actin

PCR products were respectively 287 bp for RhoA, 147 bp for RhoC and 838 bp for actin Reverse transcription was performed at 48 °C for 45 min. The reaction products were then subjected to 30 cycles of PCR for RhoA and RhoC and 35 for actin. An amplification cycle consisted of 30 sec at 94 °C for denaturation, 30 sec at 60°C and 30 sec at 68°C. Finally, an extension step at 68 °C (7 min) improved the quality of the final product by extending truncated product to full length. RT-PCR was done on control cells and after one or two treatment with siRNA.

Evaluation of RhoA and RhoC protein expression by Western blotting Western blotting was performed 24 h after two or three transfections with siRNA. Extracts from breast cancer cell lines were prepared by hypotonic buffer disruption of cells (5 mM HEPE,S pH 7.4, 1.5 mM MgCl₂, 10 mM KC1, 0.5 mM dithiothreitol (DTT), 0.1 mM phenylmethylsulfonyl fluoride (PMSF) and 0.1 mM aprotinin) . Protein concentration was measured by the technique of Bradford (Bradford, 1976) using the Bio-Rad protein assay (Hercules, CA, USA) . Equal amounts of protein extracts (20 μg) were subjected to 10% polyacrylamide gel electrophoresis . Proteins were electrotransferred onto polyvinylidene difluoride (PVDF) membranes (Amersham, Saclay, France) using a semi-dry system (Schleicher & Schuell, Dassel, Germany) . Membranes were immunoblotted overnight with mouse anti-RhoA monoclonal antibody (Santa CruzBiotechnology, Santa Cruz, CA, USA) , or goat anti-RhoC polyclonal antibody (Santa CruzBiotechnology, Santa Cruz, CA, USA) diluted 1/500. Binding of primary antibody was detected by enhanced chemiluminescence visualization system (ECL, Amersham) using the horseradish peroxidase (HRP) -coupled anti-mouse or anti-goat antibody (1:10,000; Dako, Glostrup, Denmark) for 40 min at room temperature. The autoradiography was carried-out for 1 min to 10 min. Western blot was performed 24 hours after one or two treatments with siRNA.

Cell treatment for Rho cell signalling inhibition Adherent cells were incubated for 18 hours with geranyl-geranyl transferase inhibitor (10 μM GGTI- 298) or a farnesyl transferase inhibitor (lOμM FTI-277) or cerivastatin at 25 ng/ml (concentration already shown to totally translocate RhoA from membrane to cytosol) (Denoyelle et al., 2003). β-catenin expression in the nuclear fraction Cytosolic and nuclear extracts of MDA-MB 231 cells were prepared at different times after two or three siRNA transfections, by the modified method of Dignam et al . (Dignam et al., 1983) as previously described (Denoyelle et al., 2003). Briefly, mda-mb-231 cells were washed in cold PBS, lysed in ice-cold lysis buffer containing phosphatase and protease inhibitors, and centrifuged at 100,000g for 30 min. at 4°C; the supernatant collected at this step is referred to as the cytosolic fraction. Pellets were rehomogenized in the same lysis buffer containing 2% Triton X-114 (Sigma) and centrifuged at 800g for 10 min. at 4°C; the supernatant collected at this step is the nuclear fraction. Protein concentrations were determined as above (Bradford, 1976) . To detect β-catenin, 20μg of protein extract were subjected to Western blotting using mouse β-catenin monoclonal antibody (Santa Cruz) diluted 1/500. Cell proliferation assay To ensure cell viability, minimal concentration of FCS were added to the medium: 2% for MDA- MB 231, and 7.5% for HMEC-1 cells. Briefly, after trypsination, the cells were seeded at a concentration of 1.5.10⁵ cells per well in a 24-well-plate (Costar, Cambridge, MA) . The agent to be tested was added at the appropriate concentration. For siRNA treatment, two transfections were carried out, at T₀ and after 24 hours, as described above. The cell number was measured after 48 hours by a particle counter (Coulter Zl, Coultronics, France) after detachment with a non-enzymatic cell dissociation solution (CDS, Sigma) . Cell invasiveness through matrigel This was performed as previously described (Denoyelle et al . , 2001). An 8 μm-diameter pore Transwell (Dutscher, Brumath, France) was coated with 500 μl of Matrigel (Becton Dickinson Europe, Meylan, France) diluted at 100 μg/ml. MDA-MB 231 cells were used untreated (incubated in medium alone), or treated for 24 hours FTI- 277 (10 μM), GGTI-298 (10 μM) or transfected by the siRNA for indicated time, or treated with cytofectin alone. Afetre 24h of incubation, the cells were detached by the non enzymatic cell dissociation solution, washed twice with PBS, and resuspended in RPMI 1640 with 0.2 mg/ml bovine serum albumin (BSA, Sigma) in the presence or absence of agents to be tested (at the same concentration as that used for the previous incubation) ; in the case of the siRNAs, this corresponds to a second transfection 24h after the first. In each case, 2. 10⁵ cells were seeded in the upper, Matrigel-coated chamber of the Transwell. The lower chamber was filled with 1 ml of RPMI-1640 together with 2 mg/ml BSA and basic fibroblast growth factor (bFGF, 20 ng/ml) (R&D Systems, Minneapolis, MN, USA) to induce chemotaxis. After 24 h of incubation or indicated time for siRNA, at 37 °C, the non-migrated cells in the upper chamber were gently scraped, and the adherent cells present on the lower surface of the insert were stained by May-Grϋnewald-Giemsa and counted by light microscopy. 10 fields (magnification X200) were counted for each insert. Results were calculated with reference to control values observed after incubation in medium alone (untreated control, for FTI-277 and GGTI-298 treatments), or cytofectin-containing medium (control for all siRNA treatments), arbitrarily set at 100%. Similar experiment was done with HMEC-1 cells, except that detached cells re-suspended in MCDB-131 medium with 0.2 mg/ml bovine serum albumin (BSA, Sigma). Capillary tube formation on matrigel matrix Endothelial cells controls or transfected at indicated conditions using siRNA anti RhoA or anti RhoC were used. Matrigel matrix (Becton Dickinson, France) was kept on ice during 24 hours. Then, 200 μl per well of matrigel was added to a 24-well culture plate. After gelation, at 37°C for 30 minutes, the gels were overlaid with 500 μl of medium containing 4.10⁴ endothelial cells. Then, endothelial cells were incubated with 25 ng/ml of bFGF. Capillary tube formation was inspected at different times during a period of 24 hours under an inverted light microscope at 40X magnification.

In vivo tumourigenicity assay and Immunohistochemical analysis Female athymic Nude ( nu/nu) mice (Iffa Credo,

L'arbresle, France), 6 week old, were kept in a temperature-controlled room where humidity and light were carefully monitored. MDA-MB 231 cells (4.10⁶ cells in a volume of 200 μl) were injected subcutaneously (s.c.) on the right side of the dorsal area behind the last limb. When tumors reached a volume of 0.2 cm³ (approximately within 2 weeks), the mice were arbitrarily divided into different groups. Mice in group A (n=8) were injected with 100 μl of buffered saline in the tumor whereas mice in group B and C (n=8 for each) were injected with 100 μl siRNA anti-RhoA or anti-RhoC at 85 nM. Intratumoral injections were repeated every 3 days for a total period of 20 days. Tumor sizes were routinely measured every day (perpendiculars diameters) and tumor volumes were systematically calculated. On day 20, animals were euthanized, tumors were removed and then subjected to immunohistochemistry analysis to assess the extent of intratumoral vascularization within the different experimental groups. Immunohistochemical staining was performed as previously described (Li et al., 2001). The tumors were fixed overnight in absolute ethyl alcohol and embedded in paraffin. Five-micrometer sections were then prepared. For vessel staining, tumor sections were deparaffinated in toluene, rehydrated, permeabilized by microwave oven treatment in citrate buffer (pH 6.0), quenched by 3% H₂0₂ for 5 min to remove endogenous peroxidase activity, washed in PBS and then .incubated for 1 hour with a rat antibody against mouse PECAM-1 (platelet-endothelial cell adhesion molecule-1; Pharmingen, France) and then with a biotinylated goat anti-rat IgG antibody for 15 min. After washing, sections were incubated with streptavidin-peroxydase and the vessels were revealed by peroxydase substrate diaminobenzidin . Meyer's hematoxylin was used for counterstaining. The vascularization level was evaluated as described previously by Weidner et al (Weidner et al., 1991) and presented as mean number of micro vessels per microscopic field (200X magnification) .

Results

Down-regulation of RhoA and RhoC mRNA and protein levels by anti RhoA and anti RhoC siRNA transfection to MDA-MB 231 cells and HMEC-1 Efficacy of the inhibition of RhoA and RhoC synthesis by siRNA in MDA-MB 231 and HMEC-1 was analysed by both RT-PCR and Western blot for expression of RhoA and RhoC. As shown in figure 1, both mRNA and protein expressions of RhoA and RhoC were down regulated when MDA- MB 231 cells were incubated with siRNA. 90 % inhibition was obtained when treatment with siRNA was repeated two times at 24 hours interval. However, when the incubation was prolonged for another period of 24 hours, RhoA and RhoC proteins expression increased, because siRNA in mammalian cells was transient (not shown) . Therefore, a double transfection was selected for testing the effect of RhoA or RhoC gene expression inhibition. Similar inhibition was observed with HMEC-1 cells (not shown) . This down regulation is specific as siRNA anti

RhoA did not modify RhoC mRNA expression expression and siRNA anti RhoC did not modify the expression of RhoA mRNA (fig 1) . In addition, β-actin mRNA was not modified by the treatment with siRNA.

Inhibitory action induced by anti-RhoA and anti-RhoC siRNA on MDA-MB 231 and HMEC-1 cell proliferation. Comparison with the inhibitory action of GGTI and FTI or cerivastatin . As shown in figure 2- panel A, a double transfection of MDA-MB 231 with siRNA anti RhoA or RhoC inhibit cell growth by more than 90 %. The effect of siRNA was compared on MDA-MB 231 cells with that of 10 μM GGTI- 298, a potent and selective inhibitor of geranyl-geranyl- transferase. The inhibition was lower than that induced by siRNA. As already shown the selective inhibitor of farnesyl-transferase, FTI-277 (used at 10 μM) has not significant effect on MDA-MB 231 cell proliferation. Thus, the specific inhibition of synthesis of RhoA or RhoC seems to induce a more potent inhibitory action than drugs which inhibit geranyl-geranylation such as GGTI-298. Cerivastatin at 25 ng/ml induced a similar inhibitory effect on MDA-MB 231 cell proliferation than that induced by 10 μM GGTI-298 (not shown) . The double transfection of HMEC-1 with siRNA anti RhoA or anti RhoC also inhibited HMEC-1 proliferation by more than 80% (not shown) .

Inhibitory action induced by anti RhoA and anti RhoC siRNA on MDA-MB 231 cell invasion through matrigel. As observed in figure2-Panel B, the double transfection of MDA-MB 231 cells inhibits cell invasion through matrigel by more than 90 %. The inhibition of MDA- MB 231 cell invasion by 10 μM GGTI-298 was lower and lOμM FTI-277 was ineffective. Inhibitory action induced by anti RhoA and anti RhoC siRNA on capillary tube formation from HMEC-1 Data presented in figure 3, are representative of 3 individual experiments. A 9-hour bFGF stimulation of endothelial cells triggers capillary tube formation, and sprouting of new capillaries is also visible: closed polygons are easily visible and complex mesh like structures with several shunts and branches are markedly formed. When siRNA were transfected in HMEC-1, the formation of capillary tubes was inhibited and endothelial cells remain well separated. A cell rounding was observed when cells were transfected with siRNA anti RhoA but not with siRNA anti RhoC. Inhibitory action by anti RhoA or anti RhoC siRNA on β- catenin nuclear localization As observed in figure 4, the transfection of MDA-MB 231 cells with anti RhoA or anti RhoC siRNA inhibited β-catenin nuclear fraction. Inhibition was maximal with 2 transfections . Inhibitory action induced by intratumoral injection of anti RhoA and anti RhoC siRNA on tumor growth and angiogenesis To investigate the effect of siRNA anti RhoA and anti RhoC transfection on tumor growth, 100 μl siRNA at 85 nM was injected every 3 days into 20 mm3- preestablished human MDA-MB 231 breast carcinoma tumors grown in nude mice. As shown in figure 5-A, the results demonstrate a sustained and significant arrest of tumor growth within the siRNA anti RhoA-injected group. However, only a reduced tumor growth was observed in siRNA anti RhoC-treated group. After 20 days treatment, mice were sacrificed and the tumors were removed for vascularization analysis (Figure 5-B) . PECAM-1 immunostaining was dramatically reduced in tumors excised from anti RhoA siRNA-injected mice. The angiogenesis index were 30.5 ± 4.12 in the control group and 8.75 ± 3.30 in the anti RhoA siRNA- treated group and 22.5± 3.32 in the anti RhoC siRNA- treated group.

Discussion Accumulating studies indicate that RhoA is overexpressed in cancer and Rho-protein- dependant cell signalling might be important for malignant transformation (Sahai and Marshall, 2002) . Therefore the inhibition of Rho proteins might be a good approach to inhibit cancer cell aggressivity. This strategy was already tested by using inhibitors of Rho prenylation, a step required for Rho-cell signaling. In vi tro, cerivastatin, a potent inhibitor of HMG-CoA reductase, efficiently reduced cancer cell proliferation and invasion (Denoyelle et al., 2003) and also angiogenesis (Vincent et al., 2001; Vincent et al . , 2002). However, despite the potency of cerivastatin on cancer cell aggressivity and on angiogenesis in vi tro, it cannot be considered as a potential anticancer agent due to its rapid clearance by the liver. The action of cerivastatin on MDA-MB 231 can also be mimicked by GGTI (Denoyelle et al., 2003). However, several other proteins could also be geranyl-geranylated and their role in Rho- dependant cell signalling remained to be confirmed. In an attempt to inhibit specifically RhoA and RhoC, both involved in cancer aggressivity, chemically synthesized short interfering RNA (siRNA) was used. siRNA can specifically and effectively direct homology-dependent post-transcriptional gene silencing. In the present study, it was shown that anti-RhoA and anti-RhoC siRNAs can suppress the proliferation and the migration of aggressive breast human cancer cells (MDA-MB231 cells) , characterized by an oncogenic mutation of Ras and an over-expression of RhoA and RhoC. The effect of siRNAs was compared with that of FTI, GGTI or cerivastatin used at optimal concentration (25 ng/ml) which totally inhibit prenylation of Rho to cell membrane (Denoyelle et al., 2003). Anti-RhoA or anti- RhoC siRNA was used in conditions which have been determined to suppress more than 90% of mRNA and protein synthesis (2 successive transfections at 8.5 nM in cytofectin at 24 hours interval) . Indeed, an unique transfection of RhoA or RhoC siRNA did not achieved a total inhibition of mRNA and protein expression, and consequently induced a weak inhibition in MDA-MB231 cells proliferation. The specificity of the siRNA was confirmed by the absence of action on actin mRNA expression. It was also observed that anti-RhoA siRNA did not modify RhoC mRNA expression and vice-versa (siRNA anti RhoC inhibits only RhoC expression) . The data obtained in vi tro demonstrate that the use of anti-RhoA or anti-RhoC siRNA was more effective than Rho cell signalling inhibitors, GGTI and cerivastatin, for inhibiting the proliferation and invasion of MDA-MB 231 cells and cytokine-stimulated proliferation from endothelial cells. These siRNA also inhibited the formation of capillary-like tubes by endothelial cells. Anti-RhoA and -RhoC siRNA also induced a potent inhibition of Rho-cell signalling, as demonstrated by the inhibition of β-catenin translocation into the nucleus of MDA-MB 231 cells. The difference between anti-RhoA or -RhoC siRNA and GGTI is probably due to the lack of specificity of GGTI, which at the same time inhibit prenylation of several other proteins including RhoB, which in contrast to RhoA or RhoC, antagonizes malignant transformation. Indeed RhoB-GG was shown to inhibit anchorage-dependent and -independent growth, induce apoptosis, inhibit constitutive activation of Erk and insulin-like growth factor-1 stimulation of Akt and to suppress tumor growth in nude mice (Chen et al., 2000). Taken together, the above data demonstrate that the use of anti-RhoA and -RhoC siRNA is more efficient than the classical Rho-dependant signalling inhibitors for inhibiting cancer cell aggressivity. FTI was ineffective on MDA-MB 231 cells as previously demonstrated (Denoyelle et al., 2003). The mechanism of resistance of MDA-MB 231 cells to FTI is not yet clear. It can be explained by the results of Lobell et al. (Lobell et al., 2001) showing that Ki-Ras (a member of Ras protein family) remained prenylated in FTI-treated cells while it is modified by GGTI. In order to evaluate the in vivo effect of anti-RhoA and anti-RhoC siRNAs, MDA-MB 231 tumor cells were xenografted subcutaneously in nude mice. As indicated by in vi tro results, the efficiency of siRNA is transient and to be efficient the tranfection of siRNA into tumor or endothelial cells has to be repeated. Therefore, the siRNA was introduced (lOOμl per injection with 85 nMol siRNA) directly in the tumor every three days during 20 days. The tumor growth in anti-RhoA siRNA-treated mice was inhibited by 85 % whereas in anti-RhoC siRNA group, the inhibition was less important (53 % inhibition as compared to control group) , although both siRNA anti-RhoA and -RhoC inhibited by 90% cell proliferation and cell invasion in vi tro . This difference between irj vitro and in vivo activities may be due to the fact that the sequences chosen for anti-RhoA siRNA was identical in mouse and human while the sequence of anti-RhoC siRNA is species specific and is not identical in mouse and human. A better designed murine targeted siRNA needs to be synthesized to increase the observed effect Hence, it appears that anti-RhoA siRNA could inhibit both the proliferation of tumor cells from human origin and angiogenesis derived from murine endothelial cells, while anti-RhoC siRNA could inhibit only the proliferation of tumor cells. Indeed, the angiogenic index was greatly inhibited in anti-RhoA siRNA -treated group (8.75 ± 3.30 versus 30.5 ± 4.12 for control group) and less in anti-RhoC siRNA-treated group (22.5± 3.32). The decrease in angiogenic index in the group treated with anti-RhoC, in comparison to control group could be attributed to the decreased secretion of angiogenic factors due to the decrease proliferation of tumor cells. The action in vivo of statins at high concentrations had not reproduced this inhibition and only necrotic area was observed in the tumor injection sites. The remarkable inhibition of tumor growth by anti-RhoA siRNA was explained by the excellent penetration of siRNA in tumor xenografted in mice when using cytofectin to deliver siRNA (Ades et al., 1992). In summary, the above results indicate that anti-RhoA and anti-RhoC siRNA represent a powerful tool to reduce aggressive breast cancer growth by inhibiting both cancer cells invasion and proliferation and angiogenesis under both in vi tro and in vivo conditions. Therefore, siRNA anti-RhoA and -RhoC may have potential therapeutic utility in the treatment of aggressive breast cancers. EXAMPLE 2: Intravenous delivery in mice of anti-RhoA siRNA loaded in nanoparticles of chitosan: safety and efficacy in xenografted aggressive breast cancer

Material and methods

Cells Cultures MDA-MB-231 cells were cultured as described in

Example 1. Anti-RhoA siRNA The double-stranded anti-RhoA siRNA described in Example 1 was used. Chitosan nanoparticles Type 2 Chitosan-PIHCA nanoparticles from

Bioalliance (as described in WO 2004/000287), were used.

They were suspended at a concentration of 1 mg/ml in 10 mM Hepes buffer, pH 7.3 containing 100 mM NaCI. Incorporation of siRNA into nanoparticles: siRNA nanoparticles 1 ml of anti RhoA siRNA at 30 μM was added to 300μl of the suspension of chitosan nanoparticles. After an incubation of 15 minutes at room temperature, the mixture was vortexed for 5 minutes. A diluted preparation was also tested: the previous suspension was diluted 1/10 in 10 mM hepes and 100 mM NaCI buffer .

In vivo tumorigenicity assay and siRNA administration Female athymic nude { nu/nu) mice (Iffa Credo,

L'Arbresle, France), 6 weeks old, were housed in a temperature-controlled sterile room where humidity and light were carefully monitored. MDA-MB-231 cells (5 x 10⁶ cells in a volume of 250 μl) were injected subcutaneously (s.c.) into the upper portion of the right hind limb. When thumor reached approximately 20 mm³, 100 μl siRNA were injected every 3 days for a total of 30 days in the retro- orbital vein. Mice were arbitrarily assigned to different groups (n = 12 each) to receive intravenous injection of undiluted or diluted anti-RhoA siRNA nanoparticles in order to obtain a concentration of 1500 and 150 μg of siRNA/Kg body weight respectively . Before intravenous injections the mice were anesthesthetized. Anesthesy was done by inhalation using isoflurane (aerrane, BAXTER, Paris France). In control groups, the same anesthesia was done every 3 days because it influences the body weight . Tumors were measured (perpendicular diameters) every day and their volumes calculated. After siRNA treatment was stopped (day 30) in 6 mice of each group, blood was collected for biochemical analysis. Mice were then euthanazied for examination of different organs after laparatomy. In the 6 other mice of treated groups, the mice were kept for another 30 days period for examination of the tumor growth as previously described.

Toxicological assays : For toxicological assays, a similar procedure of siRNA administration was done in mice with or without xenografted tumor. Toxicology studies of RhoA siRNA were conducted in the mice by examining the following parameters: Body weight gains and final mean body weights of mice Biochemical analysis: blood samples were collected in Eppendorf tubes and then centrifuged at 1000 g for 10 minutes at room temperature and the serum separated and stored at - 20 °C until examination. Serum of mice from each group was mixed to obtain a sufficient amount of serum for examination. Biochemical analysis were done in autoanalyser (Advia 1650 , Bayer, France) . Several enzymes released by organ injury were measured : alkaline phosphatase, alanine aminotransferase, g-glutamyltransferase for examination of liver, lipase for pancreatic examination. Proteins determination was done to determine nutritional status of mice. Urea and creatinin were also measured to examine the renal function. Finally calcium, phosphore and alakaline phosphatase were evaluated for the study of bone metabolism. organs were also examined after mice necropsy at day 30 and day 60.

Results

Intravenous injection of anti-RhoA siRNA inhibits tumor growth The effects of intravenous administration of anti-RhoA siRNA on tumor growth were investigated. To that effect, intravenous injection was done after anaesthesia in order to reach 150 and 1500 μg anti-RhoA siRNA per kg body weight. As shown in Figure 6, the 150 μg anti-RhoA siRNA injected group achieved an important decrease in tumor growth (> 90% decrease in mean tumor volume on day 30) . In the 1500 μg anti-RhoA siRNA injected group, a complete inhibition of tumor growth was obtained in all mice of this group; moreover, in 5 mice out of 12 that had 6

been treated with this high siRNA concentration, a necrotic area was observed in the center of the tumor (figures 6 and 7) . 50 % of mice were also examined after therapy stopping at day 30 for another 30 days period. Tumor growth was observed with a delay varying according to the dosage of siRNA administered (figures 6 and 7) . However, due to the importance of tumor size in control groups all mice were euthanazied at day 30. Evaluation of the toxicity of this treatment. There was no significant difference in body weight progression during the therapy between the control group in which anaesthesia was only done and the treated groups with or without xenografted tumours (figure 8). Indeed, all the mice grew at similar rates throughout the experiment . In addition, 30 days after the beginning of treatment, biochemical parameters were tested in treated and control group. Data were similar in control group and siRNA treated groups, as shown in Table 1 below.

Table 1: Toxicity of IV administration of anti-RhoA siRNA by the evaluation of biological markers in Nude mice without xenografted tumors. Blood collection was done after 30 days treatment (n.d. = not done). In each group, a pool of 10 samples was tested to have enough volume of serum for each test. Discussion In the experiments disclosed in Example 1, a remarkable inhibition of tumor growth and tumoral angiogenesis was obtained with intratumoral injection of anti-RhoA siRNAs. However, a major obstacle to the use of siRNAs as therapeutics is the difficulty involved in effective in vivo systemic delivery. Indeed, to be applicable for clinical purpose, a systemic administration is needed. The efficacy of PIHCA chitosan-coated nanoparticles-mediated intravenous delivery of siRNA in the growth of tumors was hence tested. The effect of this therapy was tested on xenografted tumor (MDA-MB 231, characterised by an overexpression of RhoA and a constitutive activation of RhoA) growth in mice. The amount of siRNA used was that deducted from the results disclosed in Example 1, showing that the antitumoral action of RhoA siRNA was obtained by injecting 0,12 μg in tumour mass every 3 days. Taking into account the dilution in body mass, 150 μg/kg body weight was administered intravenously every 3 days for 30 days and thereafter treatment was stopped. A dose 10 times higher was also administered by the same way to assess both efficacy and toxicity (1500μg/kg) . The 150 μg/kg treatment regimen resulted in an important decrease in mean tumour volume (decrease more than 90 % in comparison to untreated mice) in anti RhoA- siRNA-treated mice in comparison to untreated mice. Tumour growth was totally stopped using 1500 μg/kg, and an important zone of central necrosis in tumor was observed in 5/12 animals. This is probably in relation with the anti-RhoA-induced inhibition of angiogenesis. However, the efficacy of this therapy disappeared when the treatment was stopped and the delay for the relapse depended on the dose of siRNA administered (2 days for the lower dose and 8 days for the higher one) . Whatever this difference, this indicates that release of siRNA was completed in hours and was not comparable to that described administered DNA transfected molecules and therefore it suggests that the treatment cannot be disrupted when using this formulation. Other formulations, for example comprising integrative vectors comprising a DNA sequence that can be transcribed into anti-RhoA shRNA, could be used to obtain a longer effect. Intra-tumoral administration of such vectors could be advantageously performed according to the present invention. The study was also performed to investigate the safety and tolerability of the intravenous anti-RhoA siRNA administration. The uptake of weight did not differ between the control mice group and the groups treated with both doses of anti-RhoA siRNA. The final body weights were similar in the control and the treated groups. This absence of denutrition is in good agreement with the absence of a toxic action of siRNA, as measured by the serum proteins level (Table 1) . A similar procedure of siRNA administration was done in mice without xenografted tumor, to further investigate the safety and tolerability of this strategy. The uptake of weight did not differ between the control mice group and the groups treated with both doses of anti RhoA siRNA. Again, the final body weights were similar in the control and the two treated groups. The cytotoxicity of siRNA was also evaluated using the serum activities of enzymes markers after 30 days treatment. It was shown that siRNA treatment did not induce any cytolytic action on the liver as aminotransferase level was similar in control groups and the two treated groups. In addition, the observation that the levels of alkaline phosphatase and gamma GT were also similar in control and treated groups indicate that the treatment did not induce cholestasis. The similarity in lipase activity between control and treated groups indicate that anti-RhoA siRNA has not induced pancreatitis. In addition, renal function was not modified by the siRNA therapy in mice, as shown by the comparison of urea and creatinin levels in control and treated groups . The absence of organ toxicity of anti-RhoA siRNA could be explained by the observation of Fritz et al . showing that all breast tumors analyzed contained large amounts of RhoA protein, whereas RhoA was hardly or not at all detectable in adjacent normal tissue (Fritz et al., 2002). In addition, it was also shown that in non- stimulated endothelial cells, inhibition of RhoA had no effect whereas it inhibits both proliferation and migration of angiogenic endothelial cells after stimulation by angiogenic factors. As in vasculature endothelial cells are not activated, anti-RhoA siRNA treatment is not toxic for the vessel wall. Furthermore, examination of liver, intestine, muscles etc. after necropsy did not exhibit therapy- induced injury. Therefore, it appears that the treatment with anti-RhoA siRNA is not toxic. These results indicate that anti-RhoA siRNAs represent a powerful tool for inhibiting cancer aggressivity in vivo, with potential therapeutic utility for the treatment of aggressive breast cancers, or cancers of diverse origins in which RhoA is overexpressed .

REFERENCES Abbas-Terki, T., Blanco-Bose, W. , Deglon, N., Pralong, W. and Aebischer, P. (2002) Lentiviral-mediated RNA interference. Hum Gene Ther, 13, 2197-2201. Abecassis, I., Olofsson, B., Schmid, M.,

Zalcman, G. and Karniguian, A. (2003) RhoA induces MMP-9 expression at CD44 lamellipodial focal complexes and promotes HMEC-1 cell invasion. Exp Cell Res, 291, 363-376. Ades, E.W., Candal, F.J., Swerlick, R.A., George, V.G., Summers, S., Bosse, D.C. and Lawley, T.J. (1992) HMEC-1: establishment of an immortalized human microvascular endothelial cell line. J Invest Derma tol , 99, 683-690. Aznar, S. and Lacal, J.C. (2001) Rho signals to cell growth and apoptosis. Cancer Lett, 165, 1-10. Bertrand, J.R., Pottier, M. , Vekris, A., Opolon, P., Maksimenko, A. and Malvy, C. (2002) Comparison of antisense oligonucleotides and siRNAs in cell culture and in vivo. Biochem Biophys Res Commun, 296, 1000-1004. Borchard, G. (2001) Chitosans for gene delivery. Adv Drug Deliv Rev, 52, 145-150. Bouis, D. , Hospers, G.A., Meijer, C, Molema, G. and Mulder, N.H. (2001) Endothelium in vitro: a review of human vascular endothelial cell lines for blood vessel- related research. Angiogenesis, 4, 91-102. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem, 72, 248-254. Chen, Z., Sun, J. , Pradines, A., Favre, G.,

Adnane, J. and Sebti, S.M. (2000) Both farnesylated and geranylgeranylated RhoB inhibit malignant transformation and suppress human tumor growth in nude mice. J Biol Chem, 275, 17974-17978. Clark, E.A., Golub, T.R., Lander, E.S. and Hynes, R.O. (2000) Genomic analysis of metastasis reveals an essential role for RhoC. Na ture, 406, 532-535. Dastan, T. and Turan, K. (2004) In vitro characterization and delivery of chitosan-DNA microparticles into mammalian cells. J Phar Pharm Sci , 1 , 205-214. Denoyelle, C, Albanese, P., Uzan, G., Hong, L., Vannier, J.P., Soria, J. and Soria, C. (2003) Molecular mechanism of the anti-cancer activity of cerivastatin, an inhibitor of HMG-CoA reductase, on aggressive human breast cancer cells. Cell Si gna l, 15, 327-338. Denoyelle, C, Vasse, M., Korner, M., Mishal, Z., Ganne, F. , Vannier, J.P., Soria, J. and Soria, C. (2001) Cerivastatin, an inhibitor of HMG-CoA reductase, inhibits the signaling pathways involved in the invasiveness and metastatic properties of highly invasive breast cancer cell lines: an in vitro study. Carcinogenesis, 22, 1139-1148. Dignam, J.D., Lebovitz, R.M. and Roeder, R.G. (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nuclei c Acids Res, 11, 1475-1489. Dy, G.K. and Adjei, A. A. (2002) The role of farnesyltransferase inhibitors in lung cancer therapy. Clin Lung Cancer, 4, 57-62. Elbashir, S.M., Martinez, J., Patkaniowska, A., Lendeckel, W. and Tuschl, T. (2001) Functional anatomy of siRNAs for mediating efficient RNAi in Drosophila melanogaster embryo lysate. Embo J, 20, 6877-6888. Frame, M.C. and Brunton, V.G. (2002) Advances in Rho-dependent actin regulation and oncogenic transformation. Curr Opin Genet Dev, 12, 36-43. Fritz, G., Brachetti, C, Bahlmann, F. ,

Schmidt, M. and Kaina, B. (2002) Rho GTPases in human breast tumours: expression and mutation analyses and correlation with clinical parameters. Br J Cancer, 87, 635-644. Fritz, G., Just, I. and Kaina, B. (1999) Rho GTPases are over-expressed in human tumors. Int J Cancer, 81, 682-687. Haluska, P., Dy, G.K. and Adjei, A. A. (2002) Farnesyl transferase inhibitors as anticancer agents. Eur J Cancer, 38, 1685-1700. Hannon, G.J. and Conklin, D.S. (2004) RNA interference by short hairpin RNAs expressed in vertebrate cells. Methods Mol Biol , 257, 255-266. Hoang, M.V., Whelan, M.C. and Senger, D.R. (2004) Rho activity critically and selectively regulates endothelial cell organization during angiogenesis. Proc Na tl Acad Sci U S A, 101, 1874-1879. Horiuchi, A., Imai, T., Wang, C, Ohira, S., Feng, Y., Nikaido, T. and Konishi, I. (2003) Up-regulation of small GTPases, RhoA and RhoC, is associated with tumor progression in ovarian carcinoma. Lab Invest, 83, 861-870. Hutvagner, G. and Zamore, P.D. (2002) RNAi: nature abhors a double-strand. Curr Opin Genet Dev, 12, 225-232. Imamura, F. , Mukai, M., Ayaki, M. and Akedo, H. (2000) Y-27632, an inhibitor of rho-associated protein kinase, suppresses tumor cell invasion via regulation of focal adhesion and focal adhesion kinase. Jpn J Cancer Res, 91, 811-816. Kamai, T., Tsujii, T., Arai, K. , Takagi, K. ,

Asami, H., Ito, Y. and Oshima, H. (2003) Significant association of Rho/ROCK pathway with invasion and metastasis of bladder cancer. Clin Cancer Res, 9, 2632-

2641. Kamai, T., Yamanishi, T., Shirataki, H.,

Takagi, K. , Asami, H., Ito, Y. and Yoshida, K. (2004) Overexpression of RhoA, Racl, and Cdc42 GTPases is associated with progression in testicular cancer. Clin

Cancer Res, 10, 4799-4805. Kleer, C.G., van Golen, K.L., Zhang, Y., Wu, Z.F., Rubin, M.A. and Merajver, S.D. (2002) Characterization of RhoC expression in benign and malignant breast disease: a potential new marker for small breast carcinomas with metastatic ability. Am J Pa thol , 160, 579-584. Kusama, T., Mukai, M., Iwasaki, T., Tatsuta,

M., Matsumoto, Y., Akedo, H., Inoue, M. and Nakamura, H. (2002) 3-hydroxy-3-methylglutaryl-coenzyme a reductase inhibitors reduce human pancreatic cancer cell invasion and metastasis. Gastroenterology, 122, 308-317. Li, H., Lindenmeyer, F. , Grenet, C, Opolon, P., Menashi, S., Soria, C, Yeh, P., Perricaudet, M. and Lu, H. (2001) AdTIMP-2 inhibits tumor growth, angiogenesis, and metastasis, and prolongs survival in mice. Hum Gene Ther, 12, 515-526. Li, T. and Sparano, J.A. (2003) Inhibiting Ras signaling in the therapy of breast cancer. Clin Breast Cancer, 3, 405-416; discussion 417-420. Liu, Y. and Senger, D.R. (2004) Matrix- specific activation of Src and Rho initiates capillary morphogenesis of endothelial cells. Faseb J, 18, 457-468. Lobell, R.B., Omer, C.A., Abrams, M.T., Bhimnathwala, H.G., Brucker, M.J., Buser, C.A., Davide, J.P., deSolms, S.J., Dinsmore, C.J., Ellis-Hutchings, M.S., Krai, A.M., Liu, D., Lumma, W.C., Machotka, S.V., Rands, E., Williams, T.M., Graham, S.L., Hartman, G.D., Oliff, A.I., Heimbrook, D.C. and Kohl, N.E. (2001) Evaluation of farnesyl : protein transferase and geranylgeranyl :protein transferase inhibitor combinations in preclinical models. Cancer Res, 61, 8758-8768. Malaney, S. and Daly, R.J. (2001) The ras signaling pathway in mammary tumorigenesis and metastasis. J Mammary Gland Biol Neoplasia , 6, 101-113. Mansouri, S., Lavigne, P., Corsi, K. ,

Benderdour, M., Beaumont, E. and Fernandes, J.C. (2004) Chitosan-DNA nanoparticles as non-viral vectors in gene therapy: strategies to improve transfection efficacy. Eur J Pharm Biopharm, 57, 1-8. Mao, H.Q., Roy, K., Troung-Le, V.L., Janes, K.A., Lin, K.Y., Wang, Y., August, J.T. and Leong, K.W. (2001) Chitosan-DNA nanoparticles as gene carriers: synthesis, characterization and transfection efficiency. J Con trol Release, 70, 399-421. Nsereko, S. and A iji, M. (2002) Localized delivery of paclitaxel in solid tumors from biodegradable chitin microparticle formulations. Bioma terials , 23, 2723- 2731. Pille, J.Y., Denoyelle, C, Varet, J. , Bertrand, J.R., Soria, J. , Opolon, P., Lu, H., Pritchard, L.L., Vannier, J.P., Malvy, C, Soria, C. and Li, H. (2005) Anti-RhoA and anti-RhoC siRNAs inhibit the proliferation and invasiveness of MDA-MB-231 breast cancer cells in vitro and in vivo. Mol Ther, 11, 267-274. Pouton, C.W. and Seymour, L.W. (2001) Key issues in non-viral gene delivery. Adv Drug Deliv Rev, 46, 187-203. Prendergast, G.C. (2001) Farnesyltransferase inhibitors define a role for RhoB in controlling neoplastic pathophysiology . Histol Histopa thol , 16, 269- 275. Price, L.S. and Collard, J.G. (2001)

Regulation of the cytoskeleton by Rho-family GTPases: implications for tumour cell invasion. Semin Cancer Biol , 11, 167-173. Quong, D. and Neufeld, R.J. (1998) DNA protection from extracapsular nucleases, within chitosan- or poly-L-lysine-coated alginate beads. Biotechnol Bioeng, 60, 124-134. Sahai, E. and Marshall, C.J. (2002) RHO- GTPases and cancer. Na t Rev Cancer, 2 , 133-142. Sahai, E. and Marshall, C.J. (2003) Differing modes of tumour cell invasion have distinct requirements for Rho/ROCK signalling and extracellular proteolysis. Na t Cell Biol , 5, 711-719. Schmitz, A. A., Govek, E.E., Bottner, B. and Van Aelst, L. (2000) Rho GTPases: signaling, migration, and invasion. Exp Cell Res, 261, 1-12. Seabra, M.C. (1998) Membrane association and targeting of prenylated Ras-like GTPases. Cel l Signal , 10, 167-172. Sebti, S.M. and Hamilton, A.D. (2000) Farnesyltransferase and geranylgeranyltransferase I inhibitors and cancer therapy: lessons from mechanism and bench-to-bedside translational studies. Oncogene, 19,

6584-6593. Tuschl, T. (2002) Expanding small RNA interference. Wat Biotechnol , 20, 446-448. Vincent, L., Albanese, P., Bompais, H., Uzan, G., Vannier, J.P., Steg, P.G., Soria, J. and Soria, C. (2003) Insights in the molecular mechanisms of the anti- angiogenic effect of an inhibitor of 3-hydroxy-3- methylglutaryl coenzyme A reductase. Thromb Haemost, 89, 530-537. Vincent, L., Chen, W., Hong, L., Mirshahi, F. , Mishal, Z., Mirshahi-Khorassani, T., Vannier, J.P., Soria, J. and Soria, C. (2001) Inhibition of endothelial cell migration by cerivastatin, an HMG-CoA reductase inhibitor: contribution to its anti-angiogenic effect. FEBS Lett, 495, 159-166. Vincent, L., Soria, C, Mirshahi, F. , Opolon, P., Mishal, Z., Vannier, J.P., Soria, J. and Hong, L. (2002) Cerivastatin, an inhibitor of 3-hydroxy-3- methylglutaryl coenzyme a reductase, inhibits endothelial cell proliferation induced by angiogenic factors in vitro and angiogenesis in in vivo models. Arterioscler Thromb Vase Biol , 22, 623-629. Weidner, N., Semple, J.P., Welch, W.R. and

Folk an, J. (1991) Tumor angiogenesis and metastasis-- correlation in invasive breast carcinoma. N Engl J Med, 324, 1-8. Yu, J.Y., DeRuiter, S.L. and Turner, D.L. (2002) RNA interference by expression of short-interfering RNAs and hairpin RNAs in mammalian cells. Proc Na tl Acad Sci U S A, 99, 6047-6052.

Claims

Claims 1. An isolated polynucleotide, which is able to inhibit RhoA expression by transcriptional interference .

2. The isolated polynucleotide according to claim 1, which is able to down-regulate RhoA mRNA and protein levels by at least 30%, preferably at least 50%, and more preferably at least 80% in cultured MDA-MB 231 cells or in cultured HMEC-1 cells, when said cells are transfected twice at 24 hours interval with 8.5 nM of said double-stranded RNA molecule.

3. The isolated polynucleotide according to claim 1 or claim 2, which is an anti-RhoA siRNA or a DNA that can be transcribed into anti-RhoA siRNA in mammalian cells.

4. The isolated polynucleotide according to any of claims 1 to 3, which is a double-stranded RNA molecule of sequence: sens 5' GACAUGCUUGCUCAUAGUCTT 3' (SEQ ID No: 1) antisens 3' TTCUGUACGAACGAGUAUCAG 5' (SEQ ID No: 2).

5. An isolated polynucleotide, which is able to inhibit RhoC expression by transcriptional interference .

6. The isolated polynucleotide according to claim 5, which is able to down-regulate RhoC mRNA and protein levels by at least 30%, preferably at least 50%, and more preferably at least 80% in cultured MDA-MB 231 cells or in cultured HMEC-1 cells, when said cells are transfected twice at 24 hours interval with 8.5 nM of said double-stranded RNA molecule.

7. The isolated polynucleotide according to claim 5 or claim 6, which is an anti-RhoC siRNA or a DNA that can be transcribed into anti-RhoC siRNA in mammalian cells .

8. The isolated polynucleotide according to any of claims 5 to 7, which is a double-stranded RNA molecule of sequence: sens 5' GACCUGCCUCCUCAUCGUCTT 3' (SEQ ID No: 3) antisens 3' TTCUGGACGGAGGAGUAGCAG 5' (SEQ ID No: 4).

9. A pharmaceutical composition comprising an isolated polynucleotide according to any of claims 1 to 8 and a pharmaceutically acceptable carrier.

10. A pharmaceutical composition comprising a first isolated polynucleotide according to any of claims 1 to 4, a second isolated polynucleotide according to any of claims 5 to 8, and a pharmaceutically acceptable carrier.

11. The pharmaceutical composition according to claim 9 or claim 10, wherein said polynucleotide is in the form of an injectable solution.

12. The pharmaceutical composition according to claim 11, wherein said solution is appropriate for intravenous administration.

13. The pharmaceutical composition according to any of claims 9 to 12, wherein said carrier comprises nanospheres or nanoparticles.

14. The pharmaceutical composition according to claim 13, wherein said carrier comprises chitosan nanoparticles.

15. The pharmaceutical composition according to any of claims 9 to 14, which is formulated so as to enable the administration to a patient in need thereof, of between 100 to 2000 μg of said isolated polynucleotide (s) , per kg body weight, in one day.

16. A kit of parts comprising at least one isolated polynucleotide according to any of claims 1 to 8, and a pharmaceutical vector which enables the transfection of cells by said polynucleotide, when combined with it.

17. A pharmaceutical composition according to any of claims 9 to 15, further comprising an anticancer chemotherapeutic agent.

18. Use of an isolated polynucleotide according to any of claims 1 to 8, for the preparation of a pharmaceutical composition.

19. Use of an isolated polynucleotide according to any of claims 1 to 4, for the preparation of a pharmaceutical composition against a disease or condition associated with an over-expression of RhoA.

20. Use of an isolated polynucleotide according to any of claims 4 to 8, for the preparation of a pharmaceutical composition against a disease or condition associated with an over-expression of RhoC.

21. Use according to claim 19 or claim 20, wherein said disease or condition is a hyperproliferative condition .

22. Use according to claim 21, wherein said hyperproliferative condition is a solid tumor cancer, in particular aggressive breast cancer.

23. Use according to any of claims 18 to

22, for the preparation of a pharmaceutical composition targeting metastasis in cancer.