FIELD OF THE INVENTION:

The present invention relates to methods for predicting the survival time of a patient suffering from gliobastoma. The present invention also relates to methods and pharmaceutical compositions for the treatment of glioblastoma.

BACKGROUND OF THE INVENTION:

Despite combined modality treatment, including surgery and radio-chemotherapy, the prognosis of patients with glioblastoma remains extremely poor (Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoom MJ, Janzer RC, et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol 2009;10:459-66) and it still a need to develop new therapies. Almost all the patients will die of a relapse in radiation fields or away from the radiation fields, in the brain parenchyma (Konishi Y, Muragaki Y, Iseki H, Mitsuhashi N, Okada Y. Patterns of intracranial glioblastoma recurrence after aggressive surgical resection and adjuvant management: retrospective analysis of 43 cases. Neurol Med Chir (Tokyo) 2012;52:577-86). The recurrence pattern seems to be dependent on the tumor location; i.e. if the initial tumor is in direct contact with the periventricular zone (PVZ), tumors are prone to recur at further distances from the initial lesion. In addition, the patients with a tumor contacting the PVZ (PVZ+) progress quicker and have a decreased overall survival compared to those with tumors not contacting the PVZ (PVZ-) (Adeberg S, Konig L, Bostel T, Harrabi S, Welzel T, Debus J, et al. Glioblastoma recurrence patterns after radiation therapy with regard to the subventricular zone. Int J Radiat Oncol Biol Phys 20l4;90:886-93)( Jafri NF, Clarke JL, Weinberg V, Barani IJ, Cha S. Relationship of glioblastoma multiforme to the subventricular zone is associated with survival. Neuro Oncol 2013;15:91-6). Recent high-resolution genome -wide studies have revealed genetic heterogeneity within individual tumors in several cancers, including glioblastoma (Sottoriva A, Spiteri I, Piccirillo SG, Touloumis A, Collins VP, Marioni JC, et al. Intratumor heterogeneity in human glioblastoma reflects cancer evolutionary dynamics. Proc Natl Acad Sci U S A 2013;110:4009-14).

To invade, glioma cells must initiate dynamic changes in the cytoskeleton organization, notably via integrins. Integrins specifically bind to extracellular matrix proteins, connect their cytoplasmic domain to cytoskeleton and signaling proteins. In this way, integrins control RhoGTPases and actin reorganization that leads to cell migration. Among integrins expressed in glioblastoma cell lines, anb3 and anb5 integrins are involved in glioma invasion and progression to high grade glioma (Uhm JH, Gladson CL, Rao JS. The role of integrins in the malignant phenotype of gliomas. Front Biosci l999;4:Dl 88-99). The fibronectin receptor, a5b1 integrin, has also been shown as a promising therapeutic target for high grade glioma (10). Besides, elevated expression of a6 integrin from samples of glioblastoma patients is also correlated with a poor patient prognosis. A subset of glioma cells called glioma stem-like cells (GSCs) -which are responsible for the development and the maintenance of tumors - have been proposed to be responsible for tumor recurrences (Singh SK, Clarke ID, Terasaki M, Bonn VE, Hawkins C, Squire J, et al. Identification of a cancer stem cell in human brain tumors. Cancer Res 2003;63:5821-8). In fact, GSCs are more invasive than their differentiated progeny cells. Moreover, GSCs are resistant to conventional radio-chemotherapy treatment (Bao S, Wu Q, McLendon RE, Hao Y, Shi Q, Hjelmeland AB, et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature 2006;444:756-60). After treating mice with temozolomide, a drug currently used in standard glioblastoma treatment, GSCs can still form new tumors. GSCs preferentially reside in perivascular niches where they interact and communicate with tumor associated endothelial cells, via their basement membrane. When orthotopically xenografted, GSCs form tumors that recapitulate the phenotype of patient tumors, notably the ability of glioblastoma cells to infiltrate diffusely (Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, et al. Identification of human brain tumour initiating cells. Nature 2004;432:396-401). In addition, recent studies demonstrate that a6b1 and a3b1 integrins are overexpressed in GSCs in comparison to non-GSCs suggesting a role of these laminin receptors in GSC invasion.

SUMMARY OF THE INVENTION:

The present invention relates to methods for predicting the survival time of a patient suffering from gliobastoma. The present invention also relates to methods and pharmaceutical compositions for the treatment of glioblastoma. In particular, the invention is defined by the claims.

DETAILED DESCRIPTION OF THE INVENTION:

As clinical data demonstrates that relapses in glioblastoma after radio-chemotherapy are more aggressive in PVZ+ tumors than in PVZ- tumors, the inventors hypothesized that GSC migration properties are dependent on their anatomical origin (PVZ+ or PVZ-). They also hypothesized that it could exist a genomic signature in PVZ+ tumors that could explain their high capacity to infiltrate diffusely the healthy adjacent parenchyma and consequently, the worse progression of PVZ tumors. They hypothesized that this genomic signature first determined in PVZ+ tumors could predict the survival of GBM patients treated by the same Stupp protocol.

To this end, they established GSC paired cultures from CT and PVZ tumor part for two patients diagnosed with glioblastoma. They demonstrated that GSCs obtained from PVZ and CT present gene expression hallmarks and, that PVZ GSCs have higher migration capacities than CT GSCs. By investigating the effect of radiotherapy on GSC migration phenotype, they showed that GSC spreading is enhanced by ionizing radiation via abbΐ integrin. Moreover, the PVZ migration is controlled by the small RhoGTPase, RND1. Low expression of RND1 constitutes a bad prognosis factor for glioblastoma patients.

Interestingly, the inventors then identified a RNDl ^low signature of six genes (ITGA5, COL3A1, COL5A1, MET, COL1A2, LAMC1) that are significantly associated with overall survival of GBM patients treated with standard radio-chemotherapy. This RNDl ^low signature remains a good prognostic factor, independently of clinical parameters (independent prognostic factor). Among these six genes, MET and ITGA5 are promising therapeutic targets for the treatment of glioblastoma.

Prediction methods

A first object of the present invention relates to a method for predicting the survival time of a subject suffering from glioblastoma comprising i) determining the expression level of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 in a tumor sample obtained from the subject, ii) comparing the expression levels determined at step i) with their respective predetermined reference levels and iii) providing a good prognosis when the expression level determined at step i) is lower than the predetermined reference levels or providing a poor prognosis when the expression level determined at step i) is higher that the predetermined expression levels.

In a particular embodiment, the survival time is the overall survival time.

In a particular embodiment, the expression level of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 or variants thereof is determined. In a particular embodiments, variants of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 are obtained by splicing.

In a particular embodiment, the subject is treated with standard radio-chemotherapy.

As used herein, the term glioblastoma (GBM), also called glioblastoma multiforme or "grade 4 astrocytoma", has its general meaning in the art and refers to central nervous system primary tumor derived from glial cells. GBM is one of the deadliest human cancers with an incidence of about 3.5/100,000 per year worldwide (Cloughesy, T.F., W.K. Cavenee, and P.S. Mischel, Glioblastoma: from molecular pathology to targeted treatment. Annu Rev Pathol, 2014. 9: p. 1-25). Despite the aggressive standard of care currently used including surgery, chemo- and radiotherapy, the prognosis remains very poor with -15 months overall survival (Weathers, S.P. and M.R. Gilbert, Advances in treating glioblastoma. FlOOOPrime Rep, 2014. 6: p. 46).

As used herein, the term“tumor” refers to any growth deregulated cell(s) which may be part of a mass of tissue.

As used herein, the term“periventricular zone” refers to the structure found closed to and in the contact of the lateral walls of the lateral ventricles. The periventricular zone is a known site where neurogenesis continues into adulthood and harbours neural stem and progenitors cells.

As used herein, the term“subject” denotes a mammal, such as a rodent, a feline, a canine, and a primate. Preferably, a subject according to the invention is a human.

As used herein, the term“tumor sample” means any tissue tumor sample derived from the subject. Said tissue sample is obtained for the purpose of the in vitro evaluation. In some embodiments, the tumor sample may result from the tumor resected from the subject. In some embodiments, the tumor sample may result from a biopsy performed in the primary tumor of the subject or performed in metastatic sample distant from the primary tumor of the subject.

As used herein, the term“predicting” refers to a method of forming a prognosis, wherein a medically trained person analyzes biomarker information.

As used herein, the term "predetermined reference level" refers to the expression levels of MET, ITGA5, LAMC1, COL1A2, COL3A1 or COL5A1 in samples obtained from the patients diagnosed for GBM. A "predetermined reference level" may be determined, for example, by determining the expression level of MET, ITGA5, LAMC1, COL1A2, COL3A1 or COL5A1 nucleic acids or encoded polypeptides, in a corresponding sample obtained from one or more control subject(s). When such a predetermined reference level is used, a higher or increased levels determined in a sample (i.e. a test sample obtained from the subject) is indicative for example that said patient has a poor prognosis.

The method of the present invention is particularly suitable for predicting the duration of the overall survival (OS). Those of skill in the art will recognize that OS survival time is generally based on and expressed as the percentage of people who survive a certain type of cancer for a specific amount of time. Cancer statistics often use an overall five-year survival rate. As used herein, the expression“short survival time” indicates that the subject will have a survival time that will be lower than the median (or mean) observed in the general population of subjects. When the subject will have a short survival time, it is meant that the subject will have a“poor prognosis”. Inversely, the expression“long survival time” indicates that the subject will have a survival time that will be higher than the median (or mean) observed in the general population of subjects. When the subject will have a long survival time, it is meant that the subject will have a“good prognosis”.

As used herein, the terms“MET”,“c-MET” or“HGFR” relates to hepatocyte growth factor receptor (HGFR), a receptor tyrosine kinase that transduces signals from the extracellular matrix into the cytoplasm by binding to hepatocyte growth factor (HGF) ligand (MET Uniprot reference: P08581 for Homo sapiens and P16056 for Mus musculus). MET is also known as “Tyrosine -protein kinase Met”,“Proto-oncogene c-Met”,“HGF/SF receptor” or“Scatter factor receptor”. MET is encoded by MET gene (NCBI gene ID: 4233 for Homo sapiens and 17295 for Mus musculus). MET regulates many physiological processes including proliferation, scattering, morphogenesis and survival. Ligand binding at the cell surface induces autophosphorylation of MET on its intracellular domain that provides docking sites for downstream signaling molecules. Following activation by ligand, it interacts with the PI3- kinase subunit PIK3R1, PLCG1, SRC, GRB2, STAT3 or the adapter GAB1. The recruitment of these downstream effectors by MET leads to the activation of several signaling cascades including the RAS-ERK, PI3 kinase-AKT, or PLCgamma-PKC.

As used herein, the term“ITGA5” refers to Integrin alpha-5 chain protein that belongs to the integrin alpha chain family (ITGA5 Uniprot reference: P08648 for Homo sapiens and Pl 1688 for Mus musculus). Alpha chain 5 undergoes post-translational cleavage in the extracellular domain to yield disulfide-linked light and heavy chains that join with beta 1 chain to form a fibronectin receptor. ITGA5 is also known as“CD49 antigen-like family member E”, “Fibronectin receptor subunit alpha”,“Integrin alpha-F” or“VLA-5”. IGA5 is encoded by ITGA5 gene (NCBI gene ID: 3678 for Homo sapiens and 16402 for Mus musculus).

As used herein, the term“LAMC 1” refers to laminin subunit gamma- 1 protein (LAMC 1 Uniprot reference: P11047 for Homo sapiens and P02468 for Mus musculus). Laminins are a family of extracellular matrix glycoproteins that represent the major non collagenous constituent of basement membranes. LAMC1 is encoded by the LAMC1 gene (NCBI gene ID: 3915 for Homo sapiens and 226519 for Mus musculus).

As used herein, the term “COL1A2” refers to collagen alpha-2(I) chain protein (Collagen, type I, alpha 2) (COL 1A2 Uniprot reference: P08123 for Homo sapiens and Q01149 for Mus musculus). COL1 A2 is encoded by the COL1A2 gene (NCBI gene ID: 1278 for Homo sapiens and 12843 for Mus musculus).

As used herein, the term“COL3A1” refers to collagen alpha-l(III) chain protein (Collagen, type III, alpha 1) (COL3A1 Uniprot reference: P02461 for Homo sapiens and P08121 for Mus musculus). COL3A1 is encoded by the COL3A1 gene (NCBI gene ID: 1281 for Homo sapiens and 12825 for Mus musculus).

As used herein, the term“COL5A1” refers to collagen alpha- l(V) chain protein (Collagen, type V, alpha 1) (COL5A1 Uniprot reference: P20908 for Homo sapiens and 088207 for Mus musculus). COL5A1 is encoded by the COL5A1 gene (NCBI gene ID: 1289 for Homo sapiens and 12831 for Mus musculus).

As used herein, the term“RND1” refers to Rho-related GTP -binding protein Rho6, also known as RhoS, which is a member of the Rho family of GTPases (RND1 Uniprot reference: Q92730 for Homo sapiens and Q8BLR7 for Mus musculus). Members of this family regulate the organization of the actin cytoskeleton in response to extracellular growth factors. RND1 is encoded by the RND1 gene (NCBI gene ID: 27289 for Homo sapiens and 223881 for Mus musculus). RND1 is also known as RHOS.

Methods for determining the expression level of MET , ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1:

Determination of the expression level of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 genes may be performed by a variety of techniques. Generally, the expression level as determined is a relative expression level. For example, the determination comprises contacting the sample with selective reagents such as probes or ligands, and thereby detecting the presence, or measuring the amount, of nucleic acids or polypeptides of interest originally in said sample. Contacting may be performed in any suitable device, such as a plate, microtiter dish, test tube, well, glass, column, and so forth. In specific embodiments, the contacting is performed on a substrate coated with the reagent, such as a nucleic acid array or a specific ligand array. The substrate may be a solid or semi-solid substrate such as any suitable support comprising glass, plastic, nylon, paper, metal, polymers and the like. The substrate may be of various forms and sizes, such as a slide, a membrane, a bead, a column, a gel, etc. The contacting may be made under any condition suitable for a detectable complex, such as a nucleic acid hybrid or an antibody-antigen complex, to be formed between the reagent and the nucleic acids or polypeptides of the biological sample. In a particular embodiment of the invention, the expression level of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 genes may be determined by determining the quantity of mRNA.

Methods for determining the quantity of mRNA are well known in the art. For example the nucleic acid contained in the samples is first extracted according to standard methods, for example using lytic enzymes or chemical solutions or extracted by nucleic-acid-binding resins following the manufacturer's instructions. The extracted mRNA is then detected by hybridization (e.g., Northern blot analysis) and/or amplification (e.g., RT-PCR). Quantitative or semi-quantitative RT-PCR is preferred. Real-time quantitative or semi-quantitative RT-PCR is particularly advantageous. Another methods that can be used are for instance microarray or RNA sequencing.

Nucleic acids having at least 10 nucleotides and exhibiting sequence complementarity or homology to the mRNA of interest herein find utility as hybridization probes. It is understood that such nucleic acids need not be identical, but are typically at least about 80% identical to the homologous region of comparable size, more preferably 85% identical and even more preferably 90-95% identical. Probes typically comprise single-stranded nucleic acids of between 10 to 1000 nucleotides in length, for instance of between 10 and 800, more preferably of between 15 and 700, typically of between 20 and 500. The probes and primers are "specific" to the nucleic acids they hybridize to, i.e. they preferably hybridize under high stringency hybridization conditions (corresponding to the highest melting temperature Tm, e.g., 50 % formamide, 5x or 6x SCC. SCC is a 0.15 M NaCl, 0.015 M Na-citrate).

In the context of the invention, "hybridization" relates to the fact of obtaining a close interaction of the nucleotide probe and the target region that is expected to be revealed by the detection of the nucleotide probe. Such an interaction can be achieved by the formation of hydrogen bonds between the nucleotide probe and the target sequence, which is typical of the interactions between complementary nucleotide molecules capable of base pairing. Hydrogen bonds can be found, for example, in the annealing of two complementary strands of DNA.

It will be advantageous to use nucleic acids in combination with appropriate means, such as a detectable label, for detecting hybridization. A wide variety of appropriate indicators are known in the art including, fluorescent, radioactive, enzymatic or other ligands.

Conventional methods and reagents for isolating RNA from a sample comprise High Pure miRNA Isolation Kit (Roche), Trizol (Invitrogen), Guanidinium thiocyanate -phenol- chloroform extraction, PureLink™ miRNA isolation kit (Invitrogen), PureLink Micro-to- Midi Total RNA Purification System (invitrogen), RNeasy kit (Qiagen), Oligotex kit (Qiagen), phenol extraction, phenol-chloroform extraction, TCA/acetone precipitation, ethanol precipitation, Column purification, Silica gel membrane purification, PureYield™ RNA Midiprep (Promega), PolyATtract System 1000 (Promega), Maxwell® 16 System (Promega), SV Total RNA Isolation (Promega), geneMAG-RNA / DNA kit (Chemicell), TRI Reagent® (Ambion), RNAqueous Kit (Ambion), ToTALLY RNA™ Kit (Ambion), Poly(A)Purist™ Kit (Ambion) and any other methods, commercially available or not, known to the skilled person.

In one embodiment, the expression level of one or more mRNAs is determined by the quantitative polymerase chain reaction (QPCR) technique. The QPCR may be performed using chemicals and/or machines from a commercially available platform. The QPCR may be performed using QPCR machines from any commercially available platform; such as Prism, geneAmp or StepOne Real Time PCR systems (Applied Biosystems), LightCycler (Roche), RapidCycler (Idaho Technology), MasterCycler (Eppendorf), BioMark™ HD System (Fluidigm), iCycler iQ system, Chromo 4 system, CFX, MiniOpticon and Opticon systems (Bio-Rad), SmartCycler system (Cepheid), RotorGene system (Corbett Fifescience), MX3000 and MX3005 systems (Stratagene), DNA Engine Opticon system (Qiagen), Quantica qPCR systems (Techne), InSyte and Syncrom cycler system (BioGene), DT-322 (DNA Technology), Exicycler Notebook Thermal cycler, TF998 System (lanlong), Fine-Gene-K systems (Bioer Technology), or any other commercially available platform. The QPCR may be performed using chemicals from any commercially available platform, such as NCode EXPRESS qPCR or EXPRESS qPCR (Invitrogen), Taqman or SYBR green qPCR systems (Applied Biosystems), Real-Time PCR reagents (Eurogentec), iTaq mix (Bio-Rad), qPCR mixes and kits (Biosense), and any other chemicals, commercially available or not, known to the skilled person. The QPCR reagents and detection system may be probe-based, or may be based on chelating a fluorescent chemical into double-stranded oligonucleotides.

The QPCR reaction may be performed in a tube; such as a single tube, a tube strip or a plate, or it may be performed in a microfluidic card in which the relevant probes and/or primers are already integrated.

In a particular embodiment, the expression level of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 genes may be determined by determining of the quantity of protein encoded by the MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 genes.

Such methods comprise contacting the sample with a binding partner capable of selectively interacting with the protein present in said sample. The binding partner is generally an antibody that may be polyclonal or monoclonal, preferably monoclonal. As used herein, the term "monoclonal antibody" refers to a population of antibody molecules that contains only one species of antibody combining site capable of immunoreacting with a particular epitope. A monoclonal antibody thus typically displays a single binding affinity for any epitope with which it immunoreacts. A monoclonal antibody may therefore contain an antibody molecule having a plurality of antibody combining sites, each immunospecific for a different epitope, e.g. a bispecific monoclonal antibody. Although historically a monoclonal antibody was produced by immortalization of a clonally pure immunoglobulin secreting cell line, a monoclonally pure population of antibody molecules can also be prepared by the methods of the invention.

Laboratory methods for preparing monoclonal antibodies are well known in the art (see, for example, Harlow et al., 1988). Monoclonal antibodies (mAbs) may be prepared by immunizing purified MET, ITGA5, LAMC1, COL1A2, COL3A1 or COL5A1 into a mammal, e.g. a mouse or a rat. The antibody-producing cells in the immunized mammal are isolated and fused with myeloma or heteromyeloma cells to produce hybrid cells (hybridoma). The hybridoma cells producing the monoclonal antibodies are utilized as a source of the desired monoclonal antibody. This standard method of hybridoma culture is described in Kohler and Milstein (1975).

While mAbs can be produced by hybridoma culture the invention is not to be so limited. Also contemplated is the use of mAbs produced by an expressing nucleic acid cloned from a hybridoma of this invention. That is, the nucleic acid expressing the molecules secreted by a hybridoma of this invention can be transferred into another cell line to produce a transformant. The transformant is genotypically distinct from the original hybridoma but is also capable of producing antibody molecules of this invention, including immunologically active fragments of whole antibody molecules, corresponding to those secreted by the hybridoma. See, for example, U.S. Pat. No. 4,642,334 to Reading; European Patent Publications No. 0239400 to Winter et al. and No. 0125023 to Cabilly et al.

Antibody generation techniques not involving immunisation are also contemplated such as for example using phage display technology to examine naive libraries (from non-immunised animals); see Barbas et al. (1992), and Waterhouse et al. (1993).

Alternatively, binding agents other than antibodies may be used for the purpose of the invention. These may be for instance aptamers, which are a class of molecule that represents an alternative to antibodies in term of molecular recognition. Aptamers are oligonucleotide or oligopeptide sequences with the capacity to recognize virtually any class of target molecules with high affinity and specificity. Such ligands may be isolated through Systematic Evolution of Ligands by Exponential enrichment (SELEX) of a random sequence library, as described in Tuerk C. and Gold L., 1990. The random sequence library is obtainable by combinatorial chemical synthesis of DNA. In this library, each member is a linear oligomer, eventually chemically modified, of a unique sequence. Possible modifications, uses and advantages of this class of molecules have been reviewed in Jayasena S.D., 1999. Peptide aptamers consists of a conformationally constrained antibody variable region displayed by a platform protein, such as E. coli Thioredoxin A that are selected from combinatorial libraries by two hybrid methods (Colas et al., 1996).

The binding partners of the invention such as antibodies or aptamers, may be labelled with a detectable molecule or substance, such as a fluorescent molecule, a radioactive molecule or any others labels known in the art. Labels are known in the art that generally provide (either directly or indirectly) a signal. As used herein, the term "labelled", with regard to the antibody or aptamer, is intended to encompass direct labeling of the antibody or aptamer by coupling (i.e., physically linking) a detectable substance, such as a radioactive agent or a fluorophore (e.g. fluorescein isothiocyanate (FITC) or phycocrythrin (PE) or lndocyanine (Cy5)) to the antibody or aptamer, as well as indirect labelling of the probe or antibody by reactivity with a detectable substance. An antibody or aptamer of the invention may be labelled with a radioactive molecule by any method known in the art.

The aforementioned assays generally involve the coating of the binding partner (ie. antibody or aptamer) in a solid support. Solid supports which can be used in the practice of the invention include substrates such as nitrocellulose (e. g., in membrane or microtiter well form); polyvinylchloride (e. g., sheets or microtiter wells); polystyrene latex (e.g., beads or microtiter plates); polyvinylidine fluoride; diazotized paper; nylon membranes; activated beads, magnetically responsive beads, and the like.

In another embodiment of the invention, the measurement of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 in the sample may be achieved by a cytometric bead array system wherein the antibodies that bind to the biomarkers are coated directly or indirectly on beads. Typically, Luminex® technology which is a new technology based on fluorescent detection using a flow cytometer, microbeads dyed with multiple fluorescent colours and lasers detection may be used.

For example, the level of a biomarker protein such as MET, ITGA5, LAMC 1 , COL1 A2, COL3A1 and COL5A1 may be measured by using standard electrophoretic and immunodiagnostic techniques, including immunoassays such as competition, direct reaction, or sandwich type assays. Such assays include, but are not limited to, Western blots; agglutination tests; enzyme- labeled and mediated immunoassays, such as ELISAs; biotin/avidin type assays; radioimmunoassays; Immunoelectrophoresis; immunoprecipitation.

More particularly, an ELISA method can be used, wherein the wells of a microtiter plate are coated with a set of antibodies against MET, ITGA5, LAMC1, COL1A2, COL3A1 or COL5A1. A sample containing or suspected of containing MET, ITGA5, LAMC1, COL1A2, COL3A1 or COL5A1 is then added to the coated wells. After a period of incubation sufficient to allow the formation of antibody-antigen complexes, the plate(s) can be washed to remove unbound moieties and a detectably labeled secondary binding molecule added. The secondary binding molecule is allowed to react with any captured sample marker protein, the plate washed and the presence of the secondary binding molecule detected using methods well known in the art.

Measuring the level of a biomarker protein such as MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 (with or without immunoassay-based methods) may also include separation of the proteins: centrifugation based on the protein's molecular weight; electrophoresis based on mass and charge; HPLC based on hydrophobicity; size exclusion chromatography based on size; and solid-phase affinity based on the protein's affinity for the particular solid-phase that is use. Once separated, MET, ITGA5, LAMC1, COL1A2, COL3A1 or COL5A1 may be identified based on the known "separation profile" e. g., retention time, for that protein and measured using standard techniques.

Alternatively, the separated proteins may be detected and measured by, for example, a mass spectrometer.

Kit of the invention

A further object of the present invention relates to a kit suitable for predicting the survival time of a subject suffering from glioblastoma comprising:

At least a means for determining the expression levels of MET, ITGA5 , LAMC 1 , COL1 A2, COL3A1 and COL5A1 in a tumor sample obtained from the subject,

Instructions for use.

Typically the kit may include primers, probes, an antibody, or a set of antibodies. In a particular embodiment, the antibody or set of antibodies are labelled. The kit may also contain other suitably packaged reagents and materials needed for the particular detection protocol, including solid-phase matrices, if applicable, and standards.

Therapeutic methods and uses

Among the six genes of the predictive signature described above, MET and ITGA5 represent promising targets for treating glioblastoma. Accordingly, a further object of the present invention relates to a method of treating glioblastoma with periventricular zone tumor localization in a subject in need thereof comprising administering to the subject a therapeutically effective amount of a MET antagonist or ITGA5 antagonist.

A further object of the present invention relates to a method of treating glioblastoma with periventricular zone tumor localization in a subject in need thereof comprising administering to the subject a therapeutically effective amount of a MET antagonist in combination with a therapeutically effective amount of an ITGA5 antagonist.

As used herein,“treatment” or“treating” is an approach for obtaining beneficial or desired results including clinical results. For purposes of this invention, beneficial or desired clinical results include, but are not limited to, one or more of the following: alleviating one or more symptoms resulting from the disease, diminishing the extent of the disease, stabilizing the disease (e.g., preventing or delaying the worsening of the disease), preventing or delaying the spread of the disease, preventing or delaying the recurrence of the disease, delaying or slowing the progression of the disease, ameliorating the disease state, providing a remission (partial or total) of the disease, decreasing the dose of one or more other medications required to treat the disease, delaying the progression of the disease, increasing the quality of life, and/or prolonging survival. The term“treatment” encompasses the prophylactic treatment. As used herein, the term "prevent” refers to the reduction in the risk of acquiring or developing a given condition.

As used herein, the term "in combination" means a process whereby the MET antagonist and the ITGA5 antagonist are administered to the same patient. The use of the term "in combination" does not restrict the order in which said therapeutic agents are administered to the subject. The MET antagonist and the ITGA5 antagonist may be administered simultaneously, at essentially the same time, or sequentially.

As used herein, the term“MET antagonist” has its general meaning in the art and refers to any compound, natural or synthetic, that blocks, suppresses, or reduces the biological activity of MET or to any compound that inhibits MET gene expression.

As used herein, the term“ITGA5 antagonist” has its general meaning in the art and refers to any compound, natural or synthetic, that blocks, suppresses, or reduces the biological activity of ITGA5 or to any compound that inhibits ITGA5 gene expression.

In some embodiments, the MET antagonist or/and the ITGA5 antagonist is a small organic molecule. The term "small organic molecule" refers to a molecule of a size comparable to those organic molecules generally used in pharmaceuticals. The term excludes biological macromolecules (e. g., proteins, nucleic acids, etc.). Preferred small organic molecules range in size up to about 5000 Da, more in particular up to 2000 Da, and most in particular up to about 1000 Da.

In a particular embodiment, the ITGA5 antagonist is SJ749 (compound n°20 in Smallheer JM et al. Synthesis and biological evaluation of nonpeptide integrin antagonists containing spirocyclic scaffolds. Bioorg Med Chem Lett. 2004 Jan 19;14(2):383-7).

In a particular embodiment, the ITGA5 antagonist is K34c (Martinkova E et al. alpha5betal integrin antagonists reduce chemotherapy-induced premature senescence and facilitate apoptosis in human glioblastoma cells. Int J Cancer. 2010 Sep 1 ; 127(5): 1240-8).

In a particular embodiment, the ITGA5 antagonist of the invention may be one of the compounds described in:

- Stragies R et al., Design and synthesis of a new class of selective integrin alpha5betal antagonists (J Med Chem. 2007 Aug 9;50(l6):3786-94);

- Umeda N et al, Suppression and regression of choroidal neovascularization by systemic administration of an alpha5betal integrin antagonist (Mol. Pharmacol. 2006;69: 1820- 1828);

- Muether P.S et al., The role of integrin alpha5betal in the regulation of corneal neovascularization (Exp. Eye Res. 2007;85:356-365);

- Zahn G. et al., Preclinical evaluation of the novel small-molecule integrin alpha5betal inhibitor JSM6427 in monkey and rabbit models of choroidal neovascularization (Arch. Ophthalmol. 2009;127: 1329-1335);

- Zischinsky G., et al., SAR of N-phenyl piperidine based oral integrin alpha5betal antagonists (Bioorg. Med. Chem. Lett. 2010;20:65-68);

- Zischinsky G. et al., Discovery of orally available integrin alpha5betal antagonists (Bioorg. Med. Chem. Lett. 2010;20:380-382);

- Delouvrie B., et al., Structure-activity relationship of a series of non peptidic RGD integrin antagonists targeting alpha5betal : Part 1 (Bioorg. Med. Chem. Lett. 20l2;22:4l 11- 4116);

- Delouvrie B. et al., Structure-activity relationship of a series of non peptidic RGD integrin antagonists targeting alpha5betal : Part 2 (Bioorg. Med. Chem. Lett. 20l2;22:4l 17- 4121).

In a particular embodiment, the ITGA5 antagonist of the invention may be one of the following compounds:

-ATN-161 (Ac-PHSCN-NH(2)), a 5-mer capped peptide derived from the synergy region of fibronectin that binds to a5b 1 and anb3 in vitro (P. Khalili, A. Arakelian, G. Chen, M.L. Plunkett, I. Beck, G.C. Parry, F. Donate, D.E. Shaw, A.P. Mazar, S.A. Rabbani Mol. Cancer Ther., 5 (2006), p. 2271);

-SJ749, a specific nonpeptidic alpha5betal antagonist (Marinelli Ll, Meyer A, Heckmann D, Lavecchia A, Novellino E, Kessler H. J Med Chem. 2005 Jun 30;48(l3):4204- v);

-CLT-28643, a small molecule integrin a5b1 inhibitor (Van Bergen Tl, Zahn G2, Caldirola P2, Fsadni M3, Caram-Lelham N2, Vandewalle E4, Moons L5, Stalmans I Invest Ophthalmol Vis Sci. 2016 Nov l;57(l4):6428-6439).

In a particular embodiment, the MET antagonist is DecoyMET, it equals to the extracellular domain of MET and acts as a decoy for HGF. DecoyMET blocks the dimerization of MET.

In a particular embodiment, the MET antagonist is NK4.

In a particular embodiment, the MET antagonist of the invention may be one of the compounds described in Joanne V. Allen et al., The discovery of benzanilides as c-Met receptor tyrosine kinase inhibitors by a directed screening approach (Bioorganic & Medicinal Chemistry Letters, 21 (18): 5224-5229).

In a particular embodiment, the MET antagonist is ATP mimics.

In a particular embodiment, the MET antagonist is a tyrosine kinase inhibitor.

In a particular embodiment, the MET antagonist is BMS-777607.

In a particular embodiment, the MET antagonist is PF-02341066.

In a particular embodiment, the MET antagonist is crizotinib (Xalkori).

In a particular embodiment, the MET antagonist is AMG-458.

In a particular embodiment, the MET antagonist is MK-2461.

In a particular embodiment, the MET antagonist of the invention may be one of the compounds described in Underiner TL. et al., Discovery of Small Molecule c-Met Inhibitors: Evolution and Profiles of Clinical Candidates (Anti-Cancer Agents in Medicinal Chemistry, 10 (1): 7-27).

In a particular embodiment, the MET antagonist is JNJ-38877605.

In a particular embodiment, the MET antagonist is PF-04217903.

In a particular embodiment, the MET antagonist is GSK 1363089 (XL880, foretinib).

In a particular embodiment, the MET antagonist of the invention may be one of the compounds described in Porter, J., Small molecule c-Met kinase inhibitors: a review of recent patents (Expert opinion on therapeutic patents, 20 (2): 159-177). In a particular embodiment, the MET antagonist is Tivantinib (ARQ197) (Eathiraj S et al., Discovery of a Novel Mode of Protein Kinase Inhibition Characterized by the Mechanism of Inhibition of Human Mesenchymal-epithelial Transition Factor (c-Met) Protein Autophosphorylation by ARQ 197, Journal of Biological Chemistry, 286 (23): 20666-20676).

In a particular embodiment, the MET antagonist of the invention may be one of the compounds described in Liu XD et al., Developing c-MET pathway inhibitors for cancer therapy: progress and challenges (Trends in Molecular Medicine, 16 (1): 37-45).

In a particular embodiment, the MET antagonist is cabozantinib (XL184, BMS-907351) (COMETRIQ).

In a particular embodiment, the MET antagonist is PHA-665752.

In some embodiments, the MET antagonist or/and the ITGA5 antagonist is an antibody or a portion thereof. In some embodiments, the ITGA5 antagonist is selected from the group consisting of chimeric antibodies, humanized antibodies or full human monoclonal antibodies.

As used herein, "antibody" includes both naturally occurring and non-naturally occurring antibodies. Specifically, "antibody" includes polyclonal and monoclonal antibodies, and monovalent and divalent fragments thereof. Furthermore, "antibody" includes chimeric antibodies, fully synthetic antibodies, single chain antibodies, and fragments thereof. The antibody may be a human or nonhuman antibody. A nonhuman antibody may be humanized by recombinant methods to reduce its immunogenicity in man.

In one embodiment of the antibodies or portions thereof described herein, the antibody is a monoclonal antibody. In one embodiment of the antibodies or portions thereof described herein, the antibody is a full human monoclonal antibody. In one embodiment of the antibodies or portions thereof described herein, the antibody is a polyclonal antibody. In one embodiment of the antibodies or portions thereof described herein, the antibody is a humanized antibody. In one embodiment of the antibodies or portions thereof described herein, the antibody is a chimeric antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises a light chain of the antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises a heavy chain of the antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises a Fab portion of the antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises a F(ab')2 portion of the antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises a Fc portion of the antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises a Fv portion of the antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises a variable domain of the antibody. In one embodiment of the antibodies or portions thereof described herein, the portion of the antibody comprises one or more CDR domains of the antibody.

Antibodies are prepared according to conventional methodology. Monoclonal antibodies may be generated using the method of Kohler and Milstein (Nature, 256:495, 1975). To prepare monoclonal antibodies useful in the invention, a mouse or other appropriate host animal is immunized at suitable intervals (e.g., twice-weekly, weekly, twice-monthly or monthly) with antigenic forms of MET or ITGA5. The animal may be administered a final "boost" of antigen within one week of sacrifice. It is often desirable to use an immunologic adjuvant during immunization. Suitable immunologic adjuvants include Freund's complete adjuvant, Freund's incomplete adjuvant, alum, Ribi adjuvant, Hunter's Titermax, saponin adjuvants such as QS21 or Quil A, or CpG-containing immunostimulatory oligonucleotides. Other suitable adjuvants are well-known in the field. The animals may be immunized by subcutaneous, intraperitoneal, intramuscular, intravenous, intranasal or other routes. A given animal may be immunized with multiple forms of the antigen by multiple routes. Briefly, the recombinant MET or ITGA5 may be provided by expression with recombinant cell lines. In particular, MET or ITGA5 may be provided in the form of human cells expressing MET or ITGA5 at their surface. Following the immunization regimen, lymphocytes are isolated from the spleen, lymph node or other organ of the animal and fused with a suitable myeloma cell line using an agent such as polyethylene glycol to form a hydridoma. Following fusion, cells are placed in media permissive for growth of hybridomas but not the fusion partners using standard methods, as described (Coding, Monoclonal Antibodies: Principles and Practice: Production and Application of Monoclonal Antibodies in Cell Biology, Biochemistry and Immunology, 3rd edition, Academic Press, New York, 1996). Following culture of the hybridomas, cell supernatants are analyzed for the presence of antibodies of the desired specificity, i.e., that selectively bind the antigen. Suitable analytical techniques include EFISA, flow cytometry, immunoprecipitation, and western blotting. Other screening techniques are well-known in the field. Preferred techniques are those that confirm binding of antibodies to conformationally intact, natively folded antigen, such as non-denaturing EFISA, flow cytometry, and immunoprecipitation.

In one embodiment, the antibody of the invention is a chimeric antibody, particularly a chimeric mouse/human antibody. According to the invention, the term "chimeric antibody" refers to an antibody which comprises a VH domain and a VL domain of a non-human antibody, and a CH domain and a CL domain of a human antibody.

This invention provides in certain embodiments compositions and methods that include humanized forms of antibodies. As used herein, "humanized" describes antibodies wherein some, most or all of the amino acids outside the CDR regions are replaced with corresponding amino acids derived from human immunoglobulin molecules. Methods of humanization include, but are not limited to, those described in U.S. Pat. Nos. 4,816,567, 5,225,539, 5,585,089, 5,693,761, 5,693,762 and 5,859,205, which are hereby incorporated by reference. The above U.S. Pat. Nos. 5,585,089 and 5,693,761, and WO 90/07861 also propose four possible criteria which may used in designing the humanized antibodies. The first proposal was that for an acceptor, use a framework from a particular human immunoglobulin that is unusually homologous to the donor immunoglobulin to be humanized, or use a consensus framework from many human antibodies. The second proposal was that if an amino acid in the framework of the human immunoglobulin is unusual and the donor amino acid at that position is typical for human sequences, then the donor amino acid rather than the acceptor may be selected. The third proposal was that in the positions immediately adjacent to the 3 CDRs in the humanized immunoglobulin chain, the donor amino acid rather than the acceptor amino acid may be selected. The fourth proposal was to use the donor amino acid reside at the framework positions at which the amino acid is predicted to have a side chain atom within 3 A of the CDRs in a three dimensional model of the antibody and is predicted to be capable of interacting with the CDRs. The above methods are merely illustrative of some of the methods that one skilled in the art could employ to make humanized antibodies. One of ordinary skill in the art will be familiar with other methods for antibody humanization.

In some embodiments, the antibody is a fully human antibody. Fully human monoclonal antibodies also can be prepared by immunizing mice transgenic for large portions of human immunoglobulin heavy and light chain loci. See, e.g., U.S. Pat. Nos. 5,591,669, 5,598,369, 5,545,806, 5,545,807, 6,150,584, and references cited therein, the contents of which are incorporated herein by reference. These animals have been genetically modified such that there is a functional deletion in the production of endogenous (e.g., murine) antibodies. The animals are further modified to contain all or a portion of the human germ-line immunoglobulin gene locus such that immunization of these animals will result in the production of fully human antibodies to the antigen of interest. Following immunization of these mice (e.g., XenoMouse(Abgenix), HuMAb mice (Medarex/GenPharm)), monoclonal antibodies can be prepared according to standard hybridoma technology. These monoclonal antibodies will have human immunoglobulin amino acid sequences and therefore will not provoke human anti mouse antibody (KAMA) responses when administered to humans. In vitro methods also exist for producing human antibodies. These include phage display technology (U.S. Pat. Nos. 5,565,332 and 5,573,905) and in vitro stimulation of human B cells (U.S. Pat. Nos. 5,229,275 and 5,567,610). The contents of these patents are incorporated herein by reference.

As will be apparent to one of ordinary skill in the art, the present invention also provides for F(ab')2, Fab, Fv and Fd fragments; chimeric antibodies in which the Fc and/or FR and/or CDR1 and/or CDR2 and/or light chain CDR3 regions have been replaced by homologous human or non-human sequences; chimeric F(ab')2 fragment antibodies in which the FR and/or CDR1 and/or CDR2 and/or light chain CDR3 regions have been replaced by homologous human or non-human sequences; chimeric Fab fragment antibodies in which the FR and/or CDR1 and/or CDR2 and/or light chain CDR3 regions have been replaced by homologous human or non-human sequences; and chimeric Fd fragment antibodies in which the FR and/or CDR1 and/or CDR2 regions have been replaced by homologous human or non-human sequences. The present invention also includes so-called single chain antibodies.

The various antibody molecules and fragments may derive from any of the commonly known immunoglobulin classes, including but not limited to IgA, secretory IgA, IgE, IgG and IgM. IgG subclasses are also well known to those in the art and include but are not limited to human IgGl, IgG2, IgG3 and IgG4.

In another embodiment, the antibody according to the invention is a single domain antibody. The term“single domain antibody” (sdAb) or "VHH" refers to the single heavy chain variable domain of antibodies of the type that can be found in Camelid mammals which are naturally devoid of light chains. Such VHH are also called“nanobody®”.

In a particular embodiment, the ITGA5 antagonist is Volociximab (Ramakrishnan V et al; Preclinical evaluation of an anti-alpha5betal integrin antibody as a novel anti-angiogenic agent. J Exp Ther Oncol. 2006;5(4):273-86).

In a particular embodiment, the ITGA5 antagonist is PF-04605412 (Li G et al. Dual functional monoclonal antibody PF-04605412 targets integrin alpha5betal and elicits potent antibody-dependent cellular cytotoxicity. Cancer Res. 2010 Dec 15;70(24): 10243-54).

In a particular embodiment, the MET antagonist is DN30 (Petrelli A et al., Ab-induced ectodomain shedding mediates hepatocyte growth factor receptor down-regulation and hampers biological activity. Proc. Natl. Acad. Sci. U.S.A. 103 (13): 5090-5). In a particular embodiment, the MET antagonist is OA-5D5 (Martens T. et al., A novel one-armed anti-c-Met antibody inhibits glioblastoma growth in vivo. Clin. Cancer Res. 12 (20 Pt 1): 6144-52).

In a particular embodiment, the MET antagonist is onartuzumab, a one-armed monovalent MET antibody.

In a particular embodiment, the MET antagonist is H224G11/ABT700, a bivalent anti- MET antibody.

In a particular embodiment, the MET antagonist is LY2875358, a bivalent anti-MET antibody.

In a particular embodiment, the HGF antagonist is rilotumumab, it blocks the interaction between HGF, the ligand of MET, with MET.

In a particular embodiment, the HGF antagonist is Ficlatuzumab, it blocks the interaction between HGF, the ligand of MET, with MET.

In one embodiment, the MET antagonist or/and the ITGA5 antagonist is an aptamer. Aptamers are a class of molecule that represents an alternative to antibodies in term of molecular recognition. Aptamers are oligonucleotide or oligopeptide sequences with the capacity to recognize virtually any class of target molecules with high affinity and specificity. Such ligands may be isolated through Systematic Evolution of Ligands by Exponential enrichment (SELEX) of a random sequence library. The random sequence library is obtainable by combinatorial chemical synthesis of DNA. In this library, each member is a linear oligomer, eventually chemically modified, of a unique sequence. Peptide aptamers consists of a conformationally constrained antibody variable region displayed by a platform protein, such as E. coli Thioredoxin A that are selected from combinatorial libraries by two hybrid methods.

In some embodiments, the MET antagonist or/and the ITGA5 antagonist is a polypeptide.

In a particular embodiment the polypeptide is a functional equivalent of MET or ITGA5. As used herein, a“functional equivalent” of MET or ITGA5 is a compound which is capable of binding to MET ligand or ITGA5 ligand, thereby preventing its interaction with MET or ITGA5. The term "functional equivalent" includes fragments, mutants, and muteins of MET or ITGA5. The term "functionally equivalent" thus includes any equivalent of MET or ITGA5 obtained by altering the amino acid sequence, for example by one or more amino acid deletions, substitutions or additions such that the protein analogue retains the ability to bind to MET ligand or ITGA5 ligand. Amino acid substitutions may be made, for example, by point mutation of the DNA encoding the amino acid sequence. Functional equivalents include molecules that bind a ligand of MET or ITGA5 and comprise all or a portion of the extracellular domains of MET or ITGA5 so as to form a soluble receptor that is capable to trap the ligand of MET or ITGA5. Thus the functional equivalents include soluble forms of MET or ITGA5. A suitable soluble form of these proteins, or functional equivalents thereof, might comprise, for example, a truncated form of the protein from which the transmembrane domain has been removed by chemical, proteolytic or recombinant methods. Typically, the functional equivalent is at least 80% homologous to the corresponding protein. In a preferred embodiment, the functional equivalent is at least 90% homologous as assessed by any conventional analysis algorithm. The term "a functionally equivalent fragment" as used herein also may mean any fragment or assembly of fragments of MET or ITGA5 that binds to a ligand of MET or ITGA5. Accordingly the present invention provides a polypeptide capable of inhibiting binding of MET or ITGA5 to a ligand of MET or ITGA5, which polypeptide comprises consecutive amino acids having a sequence which corresponds to the sequence of at least a portion of an extracellular domain of MET or ITGA5, which portion binds to a ligand of MET or ITGA5. In some embodiments, the polypeptide corresponds to an extracellular domain of MET or ITGA5.

In some embodiments, the polypeptide of the present invention is fused to a heterologous polypeptide to form a fusion protein. As used herein, a“fusion protein" comprises all or part (typically biologically active) of a polypeptide of the present invention operably linked to a heterologous polypeptide (i.e., a polypeptide other than the same polypeptide). Within the fusion protein, the term "operably linked" is intended to indicate that the polypeptide of the present invention and the heterologous polypeptide are fused in- frame to each other. The heterologous polypeptide can be fused to the N-terminus or C-terminus of the polypeptide of the present invention. In some embodiment, the heterologous polypeptide is fused to the C- terminal end of the polypeptide of the present invention.

In some embodiments, the functional equivalent of MET or ITGA5 is fused to an immunoglobulin constant domain (Fc region) to form an immunoadhesin. Immunoadhesins can possess many of the valuable chemical and biological properties of human antibodies. Since immunoadhesins can be constructed from a human protein sequence with a desired specificity linked to an appropriate human immunoglobulin hinge and constant domain (Fc) sequence, the binding specificity of interest can be achieved using entirely human components. Such immunoadhesins are minimally immunogenic to the patient, and are safe for chronic or repeated use. In some embodiments, the Fc region is a native sequence Fc region. In some embodiments, the Fc region is a variant Fc region. In still another embodiment, the Fc region is a functional Fc region. As used herein, the term "Fc region" is used to define a C-terminal region of an immunoglobulin heavy chain, including native sequence Fc regions and variant Fc regions. Although the boundaries of the Fc region of an immunoglobulin heavy chain might vary, the human IgG heavy chain Fc region is usually defined to stretch from an amino acid residue at position Cys226, or from Pro230, to the carboxyl-terminus thereof The adhesion portion and the immunoglobulin sequence portion of the immunoadhesin may be linked by a minimal linker. The immunoglobulin sequence typically, but not necessarily, is an immunoglobulin constant domain. The immunoglobulin moiety in the chimeras of the present invention may be obtained from IgGl, IgG2, IgG3 or IgG4 subtypes, IgA, IgE, IgD or IgM, but typically IgGl or IgG4. In some embodiments, the functional equivalent of CLEC-l and the immunoglobulin sequence portion of the immunoadhesin are linked by a minimal linker. As used herein, the term“linker” refers to a sequence of at least one amino acid that links the polypeptide of the invention and the immunoglobulin sequence portion. Such a linker may be useful to prevent steric hindrances. In some embodiments, the linker has 4; 5; 6; 7; 8; 9; 10; 11; 12; 13; 14; 15; 16; 17; 18; 19; 20; 21; 22; 23; 24; 25; 26; 27; 28; 29; 30 amino acid residues. However, the upper limit is not critical but is chosen for reasons of convenience regarding e.g. biopharmaceutical production of such polypeptides. The linker sequence may be a naturally occurring sequence or a non-naturally occurring sequence. If used for therapeutical purposes, the linker is typically non-immunogenic in the subject to which the immunoadhesin is administered. One useful group of linker sequences are linkers derived from the hinge region of heavy chain antibodies as described in WO 96/34103 and WO 94/04678. Other examples are poly-alanine linker sequences.

The polypeptides of the invention may be produced by any suitable means, as will be apparent to those of skill in the art. In order to produce sufficient amounts of polypeptides for use in accordance with the present invention, expression may conveniently be achieved by culturing under appropriate conditions recombinant host cells containing the polypeptide of the invention. In particular, the polypeptide is produced by recombinant means, by expression from an encoding nucleic acid molecule. Systems for cloning and expression of a polypeptide in a variety of different host cells are well known. When expressed in recombinant form, the polypeptide is in particular generated by expression from an encoding nucleic acid in a host cell. Any host cell may be used, depending upon the individual requirements of a particular system. Suitable host cells include bacteria mammalian cells, plant cells, yeast and baculovirus systems. Mammalian cell lines available in the art for expression of a heterologous polypeptide include Chinese hamster ovary cells. HeLa cells, baby hamster kidney cells and many others. Bacteria are also preferred hosts for the production of recombinant protein, due to the ease with which bacteria may be manipulated and grown. A common, preferred bacterial host is E coli.

The polypeptides of the invention, fragments thereof and fusion proteins (e.g. immunoadhesin) according to the invention can exhibit post-translational modifications, including, but not limited to glycosylations, (e.g., N-linked or O-linked glycosylations), myristylations, palmitylations, acetylations and phosphorylations (e.g., serine/threonine or tyrosine).

In some embodiments, it is contemplated that polypeptides used in the therapeutic methods of the present invention may be modified in order to improve their therapeutic efficacy. Such modification of therapeutic compounds may be used to decrease toxicity, increase circulatory time, or modify biodistribution. For example, the toxicity of potentially important therapeutic compounds can be decreased significantly by combination with a variety of drug carrier vehicles that modify biodistribution. In example adding dipeptides can improve the penetration of a circulating agent in the eye through the blood retinal barrier by using endogenous transporters.

In a particular embodiment, the ITGA5 antagonist is PHSCN peptide or acetylated amidated PHSCN peptide (also known as ATN-161) (Livant D.L. et al., Anti-invasive, antitumorigenic, and antimetastatic activities of the PHSCN sequence in prostate carcinoma. Cancer Res. 2000;60:309-320) or PHSCN dendrimers (Yao H. et al., Increased potency of the PHSCN dendrimer as an inhibitor of human prostate cancer cell invasion, extravasation, and lung colony formation. Clin. Exp. Metastasis. 2010;27: 173-184).

In a particular embodiment, the MET antagonist or/and the ITGA5 antagonist is an inhibitor of MET or ITGA5 expression.

An“inhibitor of expression” refers to a natural or synthetic compound that has a biological effect to inhibit or significantly reduce the expression of a gene. Therefore, an “inhibitor of MET or ITGA5 expression” denotes a natural or synthetic compound that has a biological effect to inhibit or significantly reduce the expression of the gene encoding for MET or ITGA5. Typically, the inhibitor of MET or ITGA5 expression has a biological effect on one or more of the following events: (1) production of an RNA template from a DNA sequence (e.g., by transcription); (2) processing of an RNA transcript (e.g., by splicing, editing, 5' cap formation, and/or 3’ end formation); (3) translation of an RNA into a polypeptide or protein; and/or (4) post-translational modification of a polypeptide or protein.

In a preferred embodiment of the invention, said inhibitor of gene expression is a siRNA, an antisense oligonucleotide or a ribozyme. Inhibitors of gene expression for use in the present invention may be based on antisense oligonucleotide constructs. Anti-sense oligonucleotides, including anti-sense RNA molecules and anti-sense DNA molecules, would act to directly block the translation of MET or ITGA5 mRNA by binding thereto and thus preventing protein translation or increasing mRNA degradation, thus decreasing the level of MET or ITGA5, and thus activity, in a cell. For example, antisense oligonucleotides of at least about 15 bases and complementary to unique regions of the mRNA transcript sequence encoding MET or ITGA5 can be synthesized, e.g., by conventional phosphodiester techniques and administered by e.g., intravenous injection or infusion. Methods for using antisense techniques for specifically inhibiting gene expression of genes whose sequence is known are well known in the art (e.g. see U.S. Pat. Nos. 6,566,135; 6,566,131; 6,365,354; 6,410,323; 6,107,091; 6,046,321; and 5,981,732).

Small inhibitory RNAs (siRNAs) can also function as inhibitors of gene expression for use in the present invention. Gene expression can be reduced by contacting the subject with a small double stranded RNA (dsRNA), or a vector or construct causing the production of a small double stranded RNA, such that gene expression is specifically inhibited (i.e. RNA interference or RNAi). Methods for selecting an appropriate dsRNA or dsRNA-encoding vector are well known in the art for genes whose sequence is known (e.g. see Tuschi, T. et al. (1999); Elbashir, S. M. et al. (2001); Hannon, GJ. (2002); McManus, MT. et al. (2002); Brummelkamp, TR. et al. (2002); U.S. Pat. Nos. 6,573,099 and 6,506,559; and International Patent Publication Nos. WO 01/36646, WO 99/32619, and WO 01/68836).

Ribozymes can also function as inhibitors of gene expression for use in the present invention. Ribozymes are enzymatic RNA molecules capable of catalyzing the specific cleavage of RNA. The mechanism of ribozyme action involves sequence specific hybridization of the ribozyme molecule to complementary target RNA, followed by endonucleolytic cleavage. Engineered hairpin or hammerhead motif ribozyme molecules that specifically and efficiently catalyze endonucleolytic cleavage of MET or ITGA5 mRNA sequences are thereby useful within the scope of the present invention. Specific ribozyme cleavage sites within any potential RNA target are initially identified by scanning the target molecule for ribozyme cleavage sites, which typically include the following sequences, GUA, GUU, and GUC. Once identified, short RNA sequences of between about 15 and 20 ribonucleotides corresponding to the region of the target gene containing the cleavage site can be evaluated for predicted structural features, such as secondary structure, that can render the oligonucleotide sequence unsuitable. The suitability of candidate targets can also be evaluated by testing their accessibility to hybridization with complementary oligonucleotides, using, e.g., ribonuclease protection assays.

Both antisense oligonucleotides and ribozymes useful as inhibitors of gene expression can be prepared by known methods. These include techniques for chemical synthesis such as, e.g., by solid phase phosphoramadite chemical synthesis. Alternatively, anti-sense RNA molecules can be generated by in vitro or in vivo transcription of DNA sequences encoding the RNA molecule. Such DNA sequences can be incorporated into a wide variety of vectors that incorporate suitable RNA polymerase promoters such as the T7 or SP6 polymerase promoters. Various modifications to the oligonucleotides of the invention can be introduced as a means of increasing intracellular stability and half-life. Possible modifications include but are not limited to the addition of flanking sequences of ribonucleotides or deoxyribonucleotides to the 5' and/or 3’ ends of the molecule, or the use of phosphorothioate or 2’-0-methyl rather than phosphodiesterase linkages within the oligonucleotide backbone.

Antisense oligonucleotides siRNAs and ribozymes of the invention may be delivered in vivo alone or in association with a vector. In its broadest sense, a "vector" is any vehicle capable of facilitating the transfer of the antisense oligonucleotide siRNA or ribozyme nucleic acid to the cells. Preferably, the vector transports the nucleic acid to cells with reduced degradation relative to the extent of degradation that would result in the absence of the vector. In general, the vectors useful in the invention include, but are not limited to, plasmids, phagemids, viruses, other vehicles derived from viral or bacterial sources that have been manipulated by the insertion or incorporation of the antisense oligonucleotide siRNA or ribozyme nucleic acid sequences. Viral vectors are a preferred type of vector and include, but are not limited to nucleic acid sequences from the following viruses: retrovirus, such as moloney murine leukemia virus, harvey murine sarcoma virus, murine mammary tumor virus, and rouse sarcoma virus; adenovirus, adeno-associated virus; SV40-type viruses; polyoma viruses; Epstein-Barr viruses; papilloma viruses; herpes virus; vaccinia virus; polio virus; and RNA virus such as a retrovirus. One can readily employ other vectors not named but known to the art.

Preferred viral vectors are based on non-cytopathic eukaryotic viruses in which non- essential genes have been replaced with the gene of interest. Non-cytopathic viruses include retroviruses (e.g., lentivirus), the life cycle of which involves reverse transcription of genomic viral RNA into DNA with subsequent proviral integration into host cellular DNA. Retroviruses have been approved for human gene therapy trials. Most useful are those retroviruses that are replication-deficient (i.e., capable of directing synthesis of the desired proteins, but incapable of manufacturing an infectious particle). Such genetically altered retroviral expression vectors have general utility for the high-efficiency transduction of genes in vivo. Standard protocols for producing replication-deficient retroviruses (including the steps of incorporation of exogenous genetic material into a plasmid, transfection of a packaging cell lined with plasmid, production of recombinant retroviruses by the packaging cell line, collection of viral particles from tissue culture media, and infection of the target cells with viral particles) are provided in KRIEGLER (A Laboratory Manual," W.H. Freeman C.O., New York, 1990) and in MURRY ("Methods in Molecular Biology," vol.7, Humana Press, Inc., Cliffton, N.J., 1991).

Preferred viruses for certain applications are the adeno-viruses and adeno-associated viruses, which are double-stranded DNA viruses that have already been approved for human use in gene therapy. The adeno-associated virus can be engineered to be replication deficient and is capable of infecting a wide range of cell types and species. It further has advantages such as, heat and lipid solvent stability; high transduction frequencies in cells of diverse lineages, including hematopoietic cells; and lack of superinfection inhibition thus allowing multiple series of transductions. Reportedly, the adeno-associated virus can integrate into human cellular DNA in a site-specific manner, thereby minimizing the possibility of insertional mutagenesis and variability of inserted gene expression characteristic of retroviral infection. In addition, wild-type adeno-associated virus infections have been followed in tissue culture for greater than 100 passages in the absence of selective pressure, implying that the adeno-associated virus genomic integration is a relatively stable event. The adeno-associated virus can also function in an extrachromosomal fashion.

Other vectors include plasmid vectors. Plasmid vectors have been extensively described in the art and are well known to those of skill in the art. See e.g., SANBROOK et ah, "Molecular Cloning: A Laboratory Manual," Second Edition, Cold Spring Harbor Laboratory Press, 1989. In the last few years, plasmid vectors have been used as DNA vaccines for delivering antigen encoding genes to cells in vivo. They are particularly advantageous for this because they do not have the same safety concerns as with many of the viral vectors. These plasmids, however, having a promoter compatible with the host cell, can express a peptide from a gene operatively encoded within the plasmid. Some commonly used plasmids include pBR322, pUCl8, pUCl9, pRC/CMV, SV40, and pBlueScript. Other plasmids are well known to those of ordinary skill in the art. Additionally, plasmids may be custom designed using restriction enzymes and ligation reactions to remove and add specific fragments of DNA. Plasmids may be delivered by a variety of parenteral, mucosal and topical routes. The plasmids may be given in an aqueous solution, dried onto gold particles or in association with another DNA delivery system including but not limited to liposomes, dendrimers, cochleate and microencapsulation. In some embodiments, the MET antagonist or/and the ITGA5 antagonist of the invention is administered to the subject with a therapeutically effective amount.

The terms "administer" or "administration" refer to the act of injecting or otherwise physically delivering a substance as it exists outside the body (e.g., the MET antagonist or/and the ITGA5 antagonist of the present invention) into the subject, such as by mucosal, intradermal, intraperitoneal, intravenous, subcutaneous, intramuscular, intra-articular delivery and/or any other method of physical delivery described herein or known in the art. When a disease, or a symptom thereof, is being treated, administration of the substance typically occurs after the onset of the disease or symptoms thereof. When a disease or symptoms thereof, are being prevented, administration of the substance typically occurs before the onset of the disease or symptoms thereof.

By a "therapeutically effective amount" is meant a sufficient amount of the MET antagonist or/and the ITGA5 antagonist for use in a method for the treatment of glioblastoma with periventricular zone tumor localization at a reasonable benefit/risk ratio applicable to any medical treatment. It will be understood that the total daily usage of the compounds and compositions of the present invention will be decided by the attending physician within the scope of sound medical judgment. The specific therapeutically effective dose level for any particular subject will depend upon a variety of factors including the severity of the glioblastoma, the age, body weight, general health, sex and diet of the subject; the time of administration, route of administration, and rate of excretion of the specific compound employed; the duration of the treatment; and like factors well known in the medical arts. For example, it is well known within the skill of the art to start doses of the compound at levels lower than those required to achieve the desired therapeutic effect and to gradually increase the dosage until the desired effect is achieved. However, the daily dosage of the products may be varied over a wide range from 0.01 to 1,000 mg per adult per day. Typically, the compositions contain 0.01, 0.05, 0.1, 0.5, 1.0, 2.5, 5.0, 10.0, 15.0, 25.0, 50.0, 100, 250 and 500 mg of the active ingredient for the symptomatic adjustment of the dosage to the subject to be treated. A medicament typically contains from about 0.01 mg to about 500 mg of the active ingredient, typically from 1 mg to about 100 mg of the active ingredient. An effective amount of the drug is ordinarily supplied at a dosage level from 0.0002 mg/kg to about 20 mg/kg of body weight per day, especially from about 0.001 mg/kg to 7 mg/kg of body weight per day.

The compositions according to the invention are formulated for any route of administration such as topical route, enteral route (such as oral, rectal) or parenteral route (such as intravenous, intra-arterial, intra-muscular, intracerebral, intracerebroventricular, intrathecal, subcutaneous).

For instance, the compositions according to the invention are formulated for parenteral (such as intravenous or intracerebral) or oral administration.

Suitable unit administration forms comprise oral-route forms such as tablets, gel capsules, powders, granules and oral suspensions or solutions, sublingual and buccal administration forms, aerosols, implants, subcutaneous, transdermal, topical, intraperitoneal, intramuscular, intravenous, subdermal, transdermal, intrathecal and intranasal administration forms and rectal administration forms.

In one embodiment, the compositions according to the invention are formulated for parenteral administration. The pharmaceutical compositions contain vehicles which are pharmaceutically acceptable for a formulation capable of being injected. These may be in particular isotonic, sterile, saline solutions (monosodium or disodium phosphate, sodium, potassium, calcium or magnesium chloride and the like or mixtures of such salts), or dry, especially freeze-dried compositions which upon addition, depending on the case, of sterilized water or physiological saline, permit the constitution of injectable solutions.

In a preferred embodiment, the compositions according to the invention are formulated for intravenous administration. In another embodiment, the compositions according to the invention are formulated for oral administration.

In a preferred embodiment, the compositions according to the invention are formulated for intracerebral administration.

Typically the active ingredient of the present invention (i.e. the MET antagonist or/and the ITGA5 antagonist) is combined with pharmaceutically acceptable excipients, and optionally sustained-release matrices, such as biodegradable polymers, to form pharmaceutical compositions.

The term "pharmaceutically" or "pharmaceutically acceptable" refers to molecular entities and compositions that do not produce an adverse, allergic or other untoward reaction when administered to a mammal, especially a human, as appropriate.

A pharmaceutically acceptable carrier or excipient refers to a non-toxic solid, semi-solid or liquid filler, diluent, encapsulating material or formulation auxiliary of any type. The carrier can also be a solvent or dispersion medium containing, for example, water, ethanol, polyol (for example, glycerol, propylene glycol, and liquid polyethylene glycol, and the like), suitable mixtures thereof, and vegetables oils. The proper fluidity can be maintained, for example, by the use of a coating, such as lecithin, by the maintenance of the required particle size in the case of dispersion and by the use of surfactants. The prevention of the action of microorganisms can be brought about by various antibacterial and antifungal agents, for example, parabens, chlorobutanol, phenol, sorbic acid, thimerosal, and the like. In many cases, it will be preferable to include isotonic agents, for example, sugars or sodium chloride. Prolonged absorption of the injectable compositions can be brought about by the use in the compositions of agents delaying absorption, for example, aluminium monostearate and gelatin.

In some embodiments, the MET antagonist or/and the ITGA5 antagonist of the present invention is administered to the subject in combination with an active ingredient.

In some embodiments, the MET antagonist or/and the ITGA5 antagonist of the present invention is administered to the subject in combination with a standard treatment. Such administration may be simultaneous, separate or sequential. For simultaneous administration the agents may be administered as one composition or as separate compositions, as appropriate. For instance, standard treatment of glioblastoma is surgery, radiotherapy and chemotherapy including for instance temozolomide, procarbazine, vincristine, carboplatin, cisplatin.

In some embodiments, the MET antagonist or/and the ITGA5 antagonist of the present invention is administered to the subject in combination with radiotherapy and temozolomide.

In some embodiments, the MET antagonist or/and the ITGA5 antagonist of the present invention is administered to the subject in combination with the Stupp protocol.

As used herein, the Stupp protocol is well known in the art and refers to a treatment composed of radiotherapy (total 60 Gy, 2 Gy per daily fraction over 6 weeks) and temozolomide (during radiotherapy: 75 mg per square meter of body-surface area per day, 7 days per week and post-radiotherapy (adjuvant): six cycles consisting of 150-200 mg per square meter for 5 days during each 28-day cycle)(Stupp et al; Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma, N Engl J Med. 2005 Mar 10;352(10):987-96) .

In some embodiments, the MET antagonist or/and the ITGA5 antagonist of the present invention is administered to the subject in combination with radiotherapy.

For instance, the treatment of glioblastoma with radiotherapy may be carried out as following: radiotherapy was administered with a total dose of 60 Gy in 2 Gy daily fractions delivered 5 days per week given over a 6-week course to the contrast-enhancing tumor or to the surgical bed with a 2-cm margin according to the EORTC protocol. All treatment was delivered with at least 6 MV beams, every day, 5 days per week.

A further object of the present invention relates to a method of treating glioblastoma with periventricular zone tumor localization in a subject in need thereof comprising i) predicting the survival time of a subject suffering from glioblastoma by performing the method of claim 1 and ii) administering to the subject a therapeutically effective amount of the MET antagonist or/and the ITGA5 antagonist when it is concluded that the subject has a poor prognosis.

A further object of the present invention relates to a method of treating glioblastoma with periventricular zone tumor localization in a subject in need thereof comprising i) predicting the survival time of a subject suffering from glioblastoma by performing the method of claim 1 and ii) administering to the subject a therapeutically effective amount of the MET antagonist or/and the ITGA5 antagonist in combination with radiotherapy when it is concluded that the subject has a poor prognosis.

A further object of the present invention relates to a method of treating glioblastoma with periventricular zone tumor localization in a subject in need thereof comprising i) predicting the survival time of a subject suffering from glioblastoma by performing the method of claim 1 and ii) administering to the subject a therapeutically effective amount of the MET antagonist or/and the ITGA5 antagonist in combination with standard treatment when it is concluded that the subject has a poor prognosis.

Another object of the present invention relates to a method of preventing relapse in a subject suffering from glioblastoma with periventricular zone tumor localization comprising administering to the subject a therapeutically effective amount of an ITGA5 antagonist and/or MET antagonist.

Another object of the present invention relates to a method for monitoring the efficiency of a glioblastoma therapy, said method comprising:

i) determining the expression level of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 in a tumor sample obtained from the subject undergoing said therapy,

ii) repeating step i) on another tumor sample obtained from the same subject taken at a later point in time,

wherein levels of expression of MET, ITGA5, LAMC1, COL1A2, COL3A1 and COL5A1 in comparison to the reference values are indicative of the efficacy of said therapy.

The invention will be further illustrated by the following figures and examples. However, these examples and figures should not be interpreted in any way as limiting the scope of the present invention.

FIGURES:

Figure 1: RND1 suppresses spreading and migration abilities of GSCs

A. RNAs from GSC neurospheres were extracted and then, the expression of RND1 was analyzed by RT-qPCR as described in Material and Methods. Data is shown as fold induction means (±SEM) from at least 3 experiments. *, <0.05; ****, <0.001 . B. After cell sorting, RNAs from GFP positive and negative PVZ1-RND1 cells were extracted and then, the expression of RND1 was analyzed by RT-qPCR as described in Material and Methods.

C. GFP negative (GFP-) and GFP positive (GFP+) PVZ1-RND1 cells were seeded on laminin pre-coated plates then allowed to spread for 3 h. Phase-contrast photographs were taken under xlO magnification.

D. Twenty thousand GFP negative (GFP-) or positive (GFP+) PVZ1-RND1 cells were seeded in the upper reservoir of Transwells coated on their undersurface with 1.5 pg/cm ² of laminin and then, the cells allowed to migrate into the lower chamber for 24 h at 37°C. Migrated cells were fixed, stained with amido black and counted. Data shown as means (±SEM) from 3 experiments performed in duplicate.

E. RNAs from stably transduced CT1 shC and CT1 sh RND1 cells were extracted and then, the expression of RND1 was quantified by RT-qPCR as described in Material and Methods. Data is shown as fold induction means (±SEM) from 3 experiments.

F. GSCs were plated on laminin pre-coated wells. Migration of individual cells was recorded by time-lapse videomicroscopy over 4 h at 37°C. The mean cell velocity of CT1 sh RND1 is compared to the mean cell velocity of CT1 shC used as a reference. *, p<0.05.

G. Stably transduced CT1 shC and CT1 sh RND1 cells were seeded on laminin pre- coated plates then allowed to spread for 1 h. In each experiment, the cell surface and the percentage of polarized cells of at least 30 individual cells were analyzed. Bars represent means (±SEM) from 3 experiments performed in duplicate; *, <0.05.

Figure 2: Down-regulation of RND1 is correlated with a worse prognosis in glioblastoma patients and activates six genes that establish a prognostic signature for glioblastoma

A. RND1 mRNA expression in normal brain tissues, in glioblastoma (GB) cell lines (LN18, U138, SF763, SF767, U251, U87) and in GSCs (Al, CT1, PVZ1, CT2, PVZ2, G, I, K, SC1, SC3) was determined by RT-qPCR as described in Material and Methods. RND1 mRNA levels in the white matter were used as a reference of normal brain expression. Data is shown as fold induction means from at least three experiments.

B. RND1 mRNA expression fold change in glioblastoma samples compared to normal brain tissues from thirteen gene expression datasets (at least p<0.05). This analysis was performed using the online NextbioResearch tools. C. Kaplan-Meier survival curves for glioblastoma patients with RND1 overexpression (black line, n=26) or down-regulation (dotted line, n=l58), determined with TCGA database. Log rank - value (down-regulated versus upregulated) = 0.0258.

EXAMPLE:

Material & Methods

Extracellular matrix proteins and antibodies.

Extracellular matrix proteins and antibodies are respectively depicted in tables 1 and 2 below. As previously described, commercial antibodies did not show sufficient affinity to allow the detection of endogenous RND1.

Table 1 : extracellular matrix proteins

Table 2: primary antibodies

Tumor samples.

Before any therapy, GB samples were obtained after informed consent from patients admitted to the neurosurgery department at Toulouse University Hospital. Tumors were histologically diagnosed as glioblastoma according to WHO criteria. For patients 1 and 2, two tumor samples were removed from the cortical area (CT1, CT2) and from the periventricular zone (PVZ1 and PVZ2). These patients had a large tumor that was in contact with both CT and PVZ. For other patients (Al, G, I, K, SC1, SC3), only one tumor sample was removed from different brain zones.

Cell culture and limiting dilution assays.

Using the neurosphere assay, GB primary cell lines were obtained from GB samples and grown in stem cell medium (GSM). The percentage of GSCs was evaluated by flow cytometry. Limiting dilution assays were performed on GSCs. For cell differentiation, GSCs were grown in DMEM-F12 supplemented with 10% FCS (FCSM) for two weeks.

Human LN18, U87, U138, U251, SF763 and SF767 glioblastoma cells were maintained in DMEM (Lonza) supplemented with 10% FCS.

Cell transfection, cell sorting and cell transduction

Five hundred thousand PVZ1 cells per well were seeded in 6-well plates and then transfected using Fugene HD (Promega) with three pg of p-EGFP-RNDl (Addgene) or of p- EGFP (Clontech). A second transfection was realized seven days after the first transfection to improve gene expression. One week after the second transfection, GFP -positive or GFP- negative GSCs were sorted by FACS. Sorted GSCs were immediately seeded on culture plates to study their ability to migrate or pelleted to quantify RND1 mRNA expression levels. Twenty five hundred thousand CT1 cells were transduced with lentiviral particles (MOI of 10: 1) containing the pFKO.1 -neo-CMVtGFP-shRNA plasmid with a sequence directed against RND 1 mRNA or a control sequence (Sigma-Aldrich). Five days after transduction, transduced cells were selected with G418.

RT-qPCR and differential expression analysis

RNA from normal human cortex and white matter were obtained from Biochain, Origene, Clontech, and Agilent. Beta2 microglobulin or actin was used as endogenous control in the ACt analysis.

After RT on GSC RNA with biotinylated desoxyribonucleotides, cDNA were hybridized on an Affymetrix Human Gene 2.0 ST array. Then, the DNA complexes were revealed by fluorescent streptavidin. Images were analyzed by Command Console and normalized with RMA method. The selection of the genes differentially expressed between CT and PVZ cells was performed using a criteria based on adjusted p-value cut off of 0.001 and log2 fold change>0.65 or <-0.65.

Orthotopic xenograft generation and immunohistochemistry In accordance with ARRIVE guidelines, the French Institution animal ethics committee approval was obtained for the protocols used on animals. Orthotopic human GB xenografts were established in 4-6 weeks-old female nude mice (Janvier) with 2.5x105 cells. Each GSC line was xenografted at least in three mice. Immunohistochemistry analysis was performed on the excised brains on paraffin-embedded sections (5 pm).

Immunofluorescence

For nestin and sox2 stainings, isolated GSCs were seeded on laminin-coated Labtek slides for 24 hours. For 3-tubulin and GFAP stainings, GSC neurospheres were seeded on laminin-coated Fabtek slides and were grown in FCSM for five days.

Cell spreading assays

Fifteen thousand GSCs were seeded on pre-coated wells with extracellular matrix proteins and incubated at 37°C. When mentioned, GSCs were pre-incubated or not with 20 pg/ml of function-blocking antibodies for 30 min at 37°C and were then sham irradiated or irradiated at 6 Gy (Gamma-cell Exactor 40, Nordion). Three random fields per well from duplicate wells were pictured under a 10 X objective. Cells were manually delineated. Cell surface (A), perimeter (P) and circularity (0=4pA/R2)) of at least 30 cells per experiment were calculated using the NIS-Elements Advanced Research 3.0 software (Nikon). Cells were classified into two groups: rounded cells and polarized cells as previously described.

Directional migration assays

Directional migration assays were performed as previously described except for the following: twenty thousand GSCs per well were seeded in Transwells (24 wells, BD Biosciences) pre-coated on their undersurface with indicated extracellular matrix proteins and incubated for 24 h at 37°C.

Non-directional migration assays

GSCs (0.75x104 cells/cm2) were seeded on laminin-coated wells (2 duplicate wells per condition) and were allowed to migrate for 4 h at 37°C, 5% C02. Using an inverted Nikon microscope at lOx magnification, two fields per well were imaged and followed at 3 min intervals with a Coolsnap HQ camera (Photometries, Tucson, AZ). Manual tracking of the nucleus was performed to follow individual cell migration (at least 30 individual cells per condition per experiment) using NIS-Elements AR 3.0 software.

Flow cytometry

Flow cytometry was performed as previously described (Monferran S, et al. Alphavbeta3 and alphavbeta5 integrins control glioma cell response to ionising radiation through IFK and RhoB. Int J Cancer 2008;123:357-64). To specifically determine integrin expression in GSCs, the gated strategy was based on the previously described protocol (Dahan P, et al. Ionizing radiations sustain glioblastoma cell dedifferentiation to a stem-like phenotype through survivin: possible involvement in radioresistance. Cell Death Dis 20l4;5:el54).

RND1 gene expression meta-analysis, survival analysis and functional enrichment

For the meta-analysis of RND1 gene expression, we selected the thirteen gene expression datasets comparing glioblastoma samples to normal tissues available on the NextBio research browser (http://www.nextbio.com/). For survival analysis, using the glioblastoma database of TCGA (http://genome-cancer.ucsc.edu/), we focused on patients treated with standard radio-chemotherapy for primary GB, excluding patients with prior glioma history (n=l84 patients). Minimum p-value approach was used to dichotomize RND1 expression (if RNDl>4.8). Overall survival rates were estimated using Kaplan-Meier method and univariate analysis were performed using logrank test. Two-sided p-values of less than 0.05 were considered statistically significant. Using TCGA, two thousand genes that are the most differentially expressed in patients with low and high RND1 expression were identified by Student’s t-test on all available TCGA patients. The p-values were adjusted using the Benjamini-Hochberg procedure for multiple testing. These genes (p-value cut off of 0.01 and log2 fold change>l . l or <0.909) were then analyzed for functional enrichment using the Cytoscape (version 3.4.0) plugin ClueGO (version 2.2.5) compared to the KEGG term. A lasso penalized cox regression was used to identify correlations between overall survival and RND 1 , genes from (KEGG ECM RECEPTOR) and (KEGG FOCAL ADHESION) pathways. A 10- fold cross validation was realized to select the best penalty parameter lambda. Using a resampling approach, bootstrap selection stability (BSS) was computed for each parameter. From genes selected by the lasso procedure, a risk score prediction was created. It is based on the linear predictor given by the model. This score was then dichotomized by taking the median value of the risk score (threshold=2.28). Thus, two groups were established (poor versus good prognostic) and corresponded to the signature for this dataset. To validate our signature, the training model obtained from TCGA was then applied on glioblastoma patients (n=l78) from REMBRANDT database (http://www.betastasis.com/) and a new risk score was obtained. The risk score was dichotomized like mentioned above, by taking the median value of the new risk score (threshold=3.15).

Statistical analysis

To compare the average from different experiments, Student’s test was used. Differences were considered statistically significant at p<0.05; risk of 5 %. Results

Characterization and molecular heterogeneity of GSCs derived from CT and PVZ

Glioblastoma samples from two patients were removed from enhanced contrast regions on MRI in the CT and in the PVZ (data not shown). We established paired cultures of CT and PVZ GSCs (CT1, PVZ1, CT2, and PVZ2) and analyzed their sternness properties. CT and PVZ cells expressed neural tumor stem cell markers CD133, NESTIN, OLIG1, OLIG2, SOX2 and A2B5 (data not shown). After culture in DMEM-F12 supplemented with 10% FCS (FCSM), CT and PVZ GSCs were able to differentiate into neuronal-like and astrocytic-like cells (data not shown) and to express differentiation markers ( GFAP , TUJ1, MAL and OMG) (data not shown). We established with a classic limiting dilution assay that CT and PVZ neurospheres gave rise to secondary neurospheres (data not shown). Both CT and PVZ GSCs had the ability to form diffusely infiltrated tumors when xenografted in nude mice brain (data not shown). Altogether, our data shows that CT and PVZ cells derived from our patient samples possess GSC characteristics.

To determine whether PVZ+ tumors possess a specific genomic signature, we performed a gene expression microarray. Different patterns of gene expression have been obtained (data not shown) according to the initial tumor location of the GSCs. We identified 109 genes differentially expressed between CT and PVZ cells (p<0.00l), associated with essential biological functions including cell adhesion, apoptosis, transcription and metabolic process. Up-regulated genes in PVZ GSC included RhoGTPase activating protein 18, transcription factor DP-2 and mannosidase alpha whereas down-regulated genes in PVZ cells included collagen type CI-alphal, RhoGTPase, RND1 and protocadherin beta3. Besides these protein coding genes, CT and PVZ cells differ in the expression of gene expression regulators (antisense RNA, miRNA, long intergenic RNA) and of regulators that guide chemical modifications of others RNAs (small nucleolar RNA...) (data not shown). These results demonstrate the molecular heterogeneity of GSCs according to their brain tumor location, notably in the migration processes.

Laminin-induced spreading of GSCs is mediated by aόbΐ integrin

Cell spreading is the first step of cell migration. As GSCs reside preferentially in perivascular niches, we performed spreading assays on laminin, fibronectin and vitronectin, three extracellular matrix proteins of the brain blood vessel basement membrane involved in glioma pathogenesis. We showed that laminin is a critical extracellular matrix protein for CT and PVZ GSCs (data not shown). The main receptors for laminin in glioma cells are aόbΐ, a6b4 and a3b1 integrins. By RT-qPCR, we showed that a6 integral and bΐ integrin mRNAs were strongly expressed in GSCs whereas a3 integrin and b4 integrin mRNAs were weakly expressed (data not shown). FACS analyses determined that CT and PVZ GSCs expressed a6 integrin and bΐ integrin at high levels on their surfaces, and did not express b4 integrin (data not shown). Integrin expression may slightly differ between GSC lines, but independently of the initial tumor location of GSCs. In order to determine which integrin was involved in laminin GSC spreading, we performed cell spreading assays using functional blocking antibodies. Inhibition of a6 integrin significantly decreased CT1 and PVZ1 GSC spreading in a similar manner (data not shown). Quantification of CT1 spreading in the presence of a3, a6, bΐ or b4 function-blocking antibody showed that cell spreading on laminin was dependent on a6b1, but not on a6b4 or on a3b1 integrin (data not shown).

Ionizing radiation increases GSC spreading on laminin via a6b1 integrin

Because radiation therapy is part of the standard treatment of patients with glioblastoma, and because ionizing radiation enhances invasion properties in glioblastoma cell lines, we tested whether irradiation could stimulate CT and PVZ GSC spreading. One hour after a 6 Gy irradiation, a significant increase of both CT1 and PVZ1 irradiated cells spreading was observed in comparison to control cells as shown by the average cell surface and the percentage of polarized cells (data not shown). To determine whether a6 integrin was involved in GSC radiation induced-spreading, we pre-treated cells with a blocking antibody against a6 subunit before irradiation and, performed cell spreading assays. Radiation-induced cell spreading required functional a6b1 integrin for both CT1 and PVZ1 GSCs (data not shown). These results, reproduced in the CT2/PVZ2 cell lines, demonstrated that ionizing radiation enhances CT and PVZ GSC spreading on laminin via a6b1 integrin.

Laminin-induced migration is increased in PVZ GSCs compared to CT GSCs

To compare CT and PVZ migration properties, we performed directional migration assay in Transwells coated on their undersurface with fibronectin or laminin. None of the GSCs was able to migrate toward fibronectin whereas they all successfully migrated toward laminin (data not shown). No significant difference in GSC haptotaxis toward laminin was observed regardless of their initial location in the brain. To sharpen the characterization of GSC migration, we performed time-lapse videomicroscopy of single GSCs seeded on laminin. Quantification of single cell migration revealed that PVZ GSCs migrated significantly faster than CT GSCs -as shown by the mean velocity determination- demonstrating differential migration capacities according to the original tumor location (data not shown). All the GSCs migrated in different directions over the entire surface and no difference in directional persistence was observed according to the tumor location (data not shown). These results showed that laminin is a permissive substrate for CT and PVZ GSC migration and that some GSC migration properties were dependent on their location in the brain.

RND1 suppresses GSC spreading and migration towards laminin

By exploring the differences between CT and PVZ GSC migration at a molecular level, we identified 14 genes known to be involved in migration that are differentially expressed between CT and PVZ cells (data not shown). As RhoGTPases are known regulators of cell migration and have been recently demonstrated as key elements of glioma pathogenesis, we focused our study on the role of RND1, a RhoGTPase, in GSC migration. RND1 regulates cell adhesion via the inhibition of the formation of actin stress fibers. By RT-qPCR, we confirmed the significant down-regulation of RND1 mRNA in PVZ GSCs in comparison to CT GSCs (figure 1A). This result could not be confirmed at protein levels, as there is no commercial antibody that recognizes the endogenous form of RND1 protein. To test a potential correlation between RND1 low-level expression and PVZ GSCs high migration, we investigated whether high levels of RND1 protein in PVZ cells could decrease their migration ability. PVZ1 cells were transfected with a plasmid encoding a fusion protein of EGFP and RND1 (PVZ1-RND1) or with a plasmid encoding EGFP (PVZ1-EGFP). GFP positive or GFP negative GSCs were selected by FACS. We confirmed RND1 overexpression in GFP positive PVZ1-RND1 GSCs by RT-qPCR (figure 1B). We then showed that, three hours after seeding on laminin, GFP positive PVZ1-RND1 cells remained round and stayed in suspension whereas GFP negative PVZ1-RND1 cells spread on laminin (figure 1C) like the control cells (data not shown). Furthermore, overexpression of RND1 in PVZ1 cells dramatically decreased their ability to migrate toward laminin after 24 h in laminin-coated Transwells (figure 1D). We next investigate the consequence of RND1 loss on migration in CT1 cells. CT1 cells were transduced with lentiviral particles expressing a shRNA directed RND1 (CT1 sh RND1) or a control shRNA (CT1 shC). RND1 expression was down regulated in stably transduced CT1 sh RND1 cells in comparison to CT1 shC (figure 1E). RND1 loss induces a slight but significant increase in mean velocity (figure 1F) and in cell spreading (figure 1G). These results demonstrated that RND1 suppresses spreading and migration abilities of GSC. Together our data indicates that the differential RND1 expression level between PVZ and CT GSCs may explain, at least in part, their different migration profiles.

Low expression of RND1 is correlated with a poor prognosis in patients with glioblastoma As loss of RND1 is involved in GSC migration of PVZ+ tumors, known to be more aggressive than PVZ- tumors, we tested whether RND1 gene expression could be correlated with glioblastoma prognosis. First, we analyzed RND1 expression by RT-qPCR in normal brain tissues, in several glioblastoma cell lines and in GSCs established in our laboratory. RND1 expression is significantly down regulated in glioblastoma cells compared to normal tissues (figure 2A). To pursue our analysis, using gene expression databases in open access, a meta analysis of RND1 expression revealed a significant down-regulation of RND1 in glioblastoma samples versus normal tissues (p<0.05, figure 2B). Using TCGA database, we next examined whether this down-regulation of RND1 was related to the prognosis of glioblastoma patients. Patients with an underexpression of RND1 ( i.e <=4.8) showed a worse survival than those with an overexpression of RND1 (HR=0.59, 95% CI:0.37-0.94, p=0.028) (figure 2C). These results clearly demonstrated that expression of RND1 is a prognosis factor of survival for patients with glioblastoma.

RNDl ^iow signature is an independent prognostic factor in glioblastoma

To determine which signaling pathways controlled by RND1 might be involved in the prognosis of glioblastoma recurrence and thus survival, we performed a functional enrichment of genes in patients with low expression of RND1 ( RNDl ^low ). In these patients, we identified thirteen signaling pathways that are activated (data not shown). The most significant being are the“extracellular matrix-receptor interaction” pathway (p=l .32 e-09), the“focal adhesion” pathway (p= 1.05 e- 07) and the“lysosome” pathway (p=l .72 e-04). This raised the hypothesis that genes from these pathways could be involved in the worse survival prognosis of patients with low RND1 expression. To assess the relationship with overall survival of RND1 and genes from“extracellular matrix-receptor interaction” and“focal adhesion” pathways, a penalized cox regression model with lasso selection was used. We identified an RNDl ^low signature of six genes ( ITGA5 , COI3A 1, COL5AJ MET, COL1A2, LAMC1 ) that are significantly associated with overall survival (data not shown). We then calculated the signature risk score for each patient in the TCGA database and divided them into a high-risk group and a low-risk group by taking the mean value of risk score. The median overall survival in the low-risk group is 17.8 months versus 13.8 months for the high-risk group (data not shown). To validate our prognostic signature, we applied the risk score formulas from TCGA to glioblastoma patients from REMBRANDT database. Consistent with TCGA results, our RNDl ^low signature predicts survival of glioblastoma patients (data not shown). Using TCGA, a multivariate cox regression analysis with clinical parameters was carried out to test the strength of RNDl ^low signature in its ability to predict survival. This analysis showed that our RNDl ^low signature remains a strong prognostic factor, independently of clinical parameters (table 3, p=0.0044). To conclude, we identified an RND 1 ^low signature that is an independent prognostic factor in glioblastoma.

Table 3: Multivariate cox regression analysis for RND 1 low signature and other prognostic markers

Discussion:

The aim of this work was to establish whether GSC migration heterogeneity exists according to the initial location of these cells within the tumor (PVZ+ or PVZ-). By using an original model of GSCs isolated from CT and PVZ, we demonstrated that PVZ GSCs migrated faster than CT GSCs and, that their migration was controlled by RND1. Moreover, we demonstrated that low-expression of RND1 in glioblastoma patient samples was correlated with a worse prognosis for patients. Finally, we identified an RND 1 ^low signature that predicts outcome for glioblastoma patients.

We demonstrated here the increased migration ability of PVZ GSCs compared to CT GSCs, which could explain the worse clinical outcome of PVZ+ patients. Only a few studies have previously described different GSC migration abilities according to their location. In fact, a GSC line derived from the PVZ, injected into a mouse brain was shown to invade the corpus callosum and the contralateral hemisphere whereas a GSC line derived from CT was not able to invade these sites. More recently, it has been shown that the GSCs from peritumoral parenchyma are much more invasive than the GSCs from the tumor mass. As CT and PVZ samples come from the same patient, our study illustrates the importance of the intratumor heterogeneity on tumor behavior. In this study, we investigated and identified genes that may distinguish CT and PVZ cells. These genes are involved in cell migration, metabolism, transcription, regulation of apoptosis and cell survival. All these functions are hallmarks of cancer and open new lines of research to explain the higher resistance to treatment and the poor overall survival of PVZ+ patients.

In this study, we investigated the impact of radiation therapy, part of standard treatment for glioblastoma, on GSC migration capacities. We demonstrated that CT and PVZ GSC spreading are equally enhanced after ionizing radiation via abbΐ integrin. Even if it is still controversial, a large number of studies established that ionizing radiation stimulates pro- invasive activities, including in non-stem glioblastoma cells. One study also showed an increase of wound healing ability in GSCs after irradiation. It was previously shown that ionizing radiation enhances: (i) bΐ integrin cell surface in glioma and stabilization of bΐ integrin protein by sialylation in human colon cancer cells. These radiation-induced modifications of b 1 integrin have been correlated with an increase in cancer cell invasion induced by irradiation. However, we did not observe any modifications in a6 and bΐ integrin expression after irradiation in CT1 or PVZ1 cells (data not shown). This result indicates that ionizing radiation induces aόbΐ integrin GSC spreading by a mechanism independent of the overexpression of aόbΐ integrin.

Besides this, we demonstrated in our GSC model that the higher migration ability of PVZ GSCs is due to lower levels of RND1. Only four recent studies explored the role of RND 1 in cell migration. Consistent with our present data, inactivation of RND 1 induces the invasion of immortalized breast cells in 3D matrigel and, overexpression of RND 1 diminishes lung colonization in mice xenografted with breast cancer cells. On the contrary, overexpression of RND1 in esophageal carcinoma cells promotes their migration. This discrepancy concerning the role of RND 1 in invasion is correlated with the difference of RND1 misregulation in these cancers. In fact, RND1 expression is down-regulated in glioblastoma patients and in the most aggressive subtypes of breast cancers but it is up-regulated in esophageal squamous cell carcinoma. Overexpression of RND1 suppresses focal adhesion sites whose formation and turnover are crucial for cell migration. It is known that there is a bi-phasic migration response to cell adhesion since both too weak and too strong adhesion can reduce cell migration. Our functional enrichment of genes in patients revealed that low expression of RND1 in glioblastoma induces an overexpression of focal adhesion proteins like extracellular matrix proteins ( COL1A1 and LAMB1) integrins ( ITGA5 and ITGB1 ); actin-binding proteins ( FLNA ; ACTN1 ) and vinculin (data not shown). Moreover, it was previously shown that overexpression of RND1 in fibroblasts decreases the expression of vinculin. We could hypothesize that misregulation of RND1 expression in glioblastoma cells leads to an optimal formation and turnover of focal adhesion sites and therefore, increased migration.

Recurrence of glioblastoma is caused by the combination of local invasion and therapy resistance. Using a data-driven approach, it has been recently demonstrated that expression of RhoGTPases is a key marker of glioma progression. In fact, the overexpression of RND3, another member of RND subfamily, enhances the invasion of glioblastoma and is correlated with a poor prognosis. In this study, we found that low levels of RND1, involved in GSC migration, are also related to a decreased overall survival in patients. Based on functional enrichment of genes in patients with low expression of RND1, we identified six genes whose expression is inversely correlated to RND1 (data not shown) and that predict the survival of glioblastoma patients. The RNDl ^low signature gathered three qualities: it was discovered from a homogeneous population of glioblastoma patients treated with standard radio-chemotherapy; it involves a short list of genes; and it remains a good prognostic factor, independently of clinical parameters. Moreover, the predictive power of the RNDl ^low signature remains significant for both the training (TCGA) and validation sets (REMBRANDT). Thanks to these qualities, the RNDl ^low signature could be useful in clinical practice to predict the survival of glioblastoma patients. The RNDl ^low signature could also lead to clinical application to improve glioblastoma treatment through the targeting of genes involved in this signature. Indeed, ITGA5 and MET were found to be key contributors to the RNDl ^low signature with their high BSS (data not shown). Integrin a5b1 has recently been described as a fine regulator of glioblastoma cell migration. MET and its ligand HGF create an autocrine signaling loop that promotes GSC invasion. In consequence, targeting ITGA5 or MET genes could inhibit the invasive capacity of glioblastoma cells induced by low RND1 expression and especially the one of PVZ+ cells. Besides, our results also showed that targeting aόbΐ integrin during radiotherapy could be an interesting way to decrease relapse and resistance to this treatment in glioblastoma since radiation-induced GSC spreading appeared to be selectively linked to this integrin.

REFERENCES:

Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.