MONOCLONAL ANTIBODY FOR THE DIAGNOSIS, TREATMENT AND/OR PREVENTION OF BRAIN TUMORS AND BRAIN LESIONS

Abstract

The invention relates to the use of the monoclonal antibody NILO1 for the diagnosis, treatment and/or prevention of brain tumors and lesions. Particularly, the invention relates to methods for the diagnosis of brain tumors and brain lesions in which cells marked with said antibody, or with immunologically active fragments thereof, are detected. The invention also relates to the use of said monoclonal antibody, or immunologically active fragments thereof, as a medicament for the treatment and/or prevention of brain tumors and brain lesions. In a preferred embodiment of the invention, the monoclonal antibody NILO1, or its immunologically active fragments, are humanized.

Claims

1. Use of a monoclonal antibody, called NILO1, produced by the hybridoma deposited under the DSM access number No. ACC2887, or an immunologically active fragment thereof, for the manufacture of a reactive for the diagnosis of brain tumors and lesions.

2. Use according to claim 1, wherein the monoclonal antibody or the immunologically active fragment thereof is coupled to a support or a particle, preferably nanoparticle, more preferably magnetic nanoparticle, more preferably magnetic glyconanoparticle, even more preferably wherein said nanoparticle further comprises a G protein immobilized in its surface.

3. Use according to any of claim 1 or 2, wherein the diagnosis of brain tumors and lesions is performed by magnetic resonance imaging, flow cytometry, immunohistochemistry, ELISA, and/or Western Blot.

4. Use according to any of claims 1 to 3, wherein the brain lesion is produced by a neurodegenerative process, demyelination, mechanical injury or stroke, and the brain tumor is glioblastoma.

5. A monoclonal antibody, called NILO1, produced by the hybridoma deposited under the DSM access number No. ACC2887, or an immunologically active fragment thereof, for use as a medicament.

6. The monoclonal antibody, called NILO1, produced by the hybridoma deposited under the DSM access number No. ACC2887, or an immunologically active fragment thereof, for use in the treatment and/or prevention of brain tumors or lesions, preferably wherein the brain lesion is produced by a neurodegenerative process, demyelination, mechanical injury or stroke, and the brain tumor is glioblastoma.

7. The monoclonal antibody, or an immunologically active fragment thereof, according to any of claim 5 or 6, wherein said antibody or said fragment is coupled to an active principle and/or a particle, wherein preferably the active principle is a drug, and more preferably the particle is a nanoparticle, more preferably magnetic nanoparticle, more preferably magnetic glyconanoparticle, even more preferably wherein said nanoparticle further comprises a G protein immobilized in its surface.

8. The monoclonal antibody or an immunologically active fragment thereof according to claim 7, wherein the active principle is for the treatment and/or prevention of brain tumors and lesions, preferably wherein the brain lesion is produced by a neurodegenerative process, demyelination, mechanical injury or stroke, and the brain tumor is glioblastoma.

9. The monoclonal antibody, or an immunologically active fragment thereof, according to any of claims 5 to 8, wherein the antibody or the immunologically active fragment thereof is humanized.

10. A method for the in vitro diagnosis of brain tumors, more preferably glioblastoma, which comprises: a. Putting in contact an isolated biological sample with the monoclonal antibody called NILO1, produced by the hybridoma deposited under the DSM access number No. ACC2887, or an immunologically active fragment thereof, b. detecting, preferably by flow cytometry, immunohistochemistry, ELISA, and/or Western Blot, the presence of cells marked with the antibody or an immunologically active fragment thereof, and c. associate the presence of said cells to a brain tumor.

11. A method for obtaining data for the diagnosis of a brain tumor or lesion, preferably brain lesions produced by a neurodegenerative process, demyelination, mechanical injury or stroke, which comprises: a. detecting the presence of cells marked with the monoclonal antibody, called NILO1, produced by the hybridoma deposited under the DSM access number No. ACC2887, or an immunologically active fragment thereof, in the brain of a subject previously injected with said antibody or an immunologically active fragment thereof, wherein the step (a) is preferably performed by magnetic resonance imaging, flow cytometry, immunohistochemistry, ELISA, and/or Western Blot, more preferably by magnetic resonance imaging.

12. The method according to any of claim 10 or 11, wherein said antibody, or an immunologically active fragment thereof, is coupled to a support or a particle, preferably nanoparticle, more preferably wherein the magnetic nanoparticle is a magnetic glyconanoparticle, more preferably wherein said nanoparticle further comprises a G protein immobilized in its surface.

13. A monoclonal antibody, called NILO1, produced by the hybridoma deposited under the DSM access number No. ACC2887, or an immunologically active fragment thereof, coupled to a support, an active principle and/or a particle, wherein preferably the active principle is a drug and more preferably the particle is a nanoparticle, more preferably magnetic nanoparticle, more preferably magnetic glyconanoparticle, even more preferably wherein said nanoparticle further comprises a G protein immobilized in its surface.

14. The monoclonal antibody or an immunologically active fragment thereof according to claim 13, wherein the active principle is for the treatment and/or prevention of brain tumors or lesions, preferably wherein the brain lesion is produced by a neurodegenerative process, demyelination, mechanical injury or stroke, and the brain tumor is glioblastoma.

15. The monoclonal antibody or an immunologically active fragment thereof according to any of claim 13 or 14, wherein the antibody or the immunologically active fragment thereof is humanized.

Description

DESCRIPTION OF THE DRAWINGS

[0070] FIG. 1. Nilo1 mAb identifies type B astrocytes in the adult brain, as well as embryonic radial glia. Immunohistochemistry of subventricular zone from wild type mice by confocal microscopy showing A, double staining with Nilo1 and Sox2, Nestin, GFAP or a triple labeling Nilo1, GFAP and Nestin, indicating that the Nilo1.sup.+ cells are positive for these markers. B, Nilo1.sup.+ cells are CD24.sup.− and had a sub-ependymal localization. Cells labeled with a short pulse of BrdU (50 mg/kg, 1 h before sacrifice) expressing high levels of Pax6 (see arrows), corresponding to the C population, are different from Nilo1.sup.+ cells (middle panels) and from neuroblasts (Nilo2.sup.+ cells, right panel). C, Double staining with Nilo1 and Nilo2 demonstrating that Nilo1.sup.+ cells are different form Nilo2.sup.+ cells (neuroblasts). D, Immunohistochemistry of fixed brains (telencephalic cortex) from E10 mouse embryos double stained with Nilo1 and either nestin or vimentin. Nuclei were counterstained with DAPI. LV, lateral ventricle. V, ventricle. P, pial suface. Scale bars: 50 μm. E, human primary glioblastoma growing in vitro as neurospheres was stained with Nilo1 mAb and revealed with a secondary FITC mouse IgG anti-hamster Ig (left) and counterstained with DAPI (middle). The merged picture is seen on the right panel.

[0071] FIG. 2. Nilo1-mGNPs revealed MRI hypointense signals in brain tumor sites. A, Cartoon representing Nilo1-mGNPs. The nanoparticles were made of a magnetic Fe.sub.3O.sub.4 core, covered by a gold shell (Au) and subsequently coated with carbohydrates and a carboxyl-ending linker to which Protein G was coupled and Nilo1 mAb bound. B, In vivo identification of SVZ Nilo1.sup.+ cells in mice intracranially injected with Nilo1-mGNPs. Brain sections were directly incubated with a secondary biotinylated anti-hamster Ig antibody and revealed with streptavidin A488. DAPI was used to stain nuclei. Scale bar: 25 μm. C, Fluorescence confocal microscopy of Nilo1-mGNPs (top) and Nilo1 mAb (bottom) labeled neurosphere cells grown in Matrigel™. DAPI was used to counter-stain nuclei. Scale bar: 25 μm. D, Flow cytometry analyses of SVZ-derived neurosphere cells stained with Nilo1-mGNPs (light grey) or Nilo1 alone (dark grey) both revealed with a fluorescent secondary antibody. Cells incubated with the secondary antibody in the absence of Nilo1 were used as negative control (black line). E, Schematic representation of the injection sites for the CT-2A astrocytoma cells (left hemisphere, d0) and Nilo1-mGNPs contralaterally in a more rostral position. F, Experimental schedule where Nilo1-mGNPs injection day is indicated with an arrowhead and MRI acquisitions are shown with asterisks, before (empty) or after nanoparticle injection (filled). G, Axial view of T2 MRI image of mouse injected with CT-2A cells (d0) and Nilo1-mGNPs. Dotted line was drawn delimiting the tumor mass. H, Representative T2* MRI study of mice injected with CT-2A cells (d0) and Nilo1-mGNPs (n=3). Axial, coronal and sagittal views just before (−1 h) or 3 h, 22 h, 48 h after nanoparticle injection. Inset in axial panel at 3 h shows the injection site of Nilo1-mGNPs (a more rostral position than the CT-2A injection site). I, Quantification of MRI signal intensity changes from H. The increase in the accumulation of nanoparticles surrounding the tumor is translated as a drop in the mean signal intensity in the region of interest.

[0072] FIG. 3. Migration of B astrocytes towards the tumor site occurs within hours following the insult. A, Experimental schedule where Nilo1-mGNPs injection day is indicated with an arrowhead and MRI acquisitions are shown with asterisks, before (empty) or after tumor cell injection at d0 (filled). B, Representative axial view from the experiment (n=3) showing T2* MRI of a mouse injected with Nilo1-mGNPs (d-4) and CT-2A cells (d0), analyzed just before (−1 h) or 3.5h, 2 or 7 days after tumor injection (left column). T2 MRI analysis was used to follow tumor growth (right column). Arrowhead indicates hypointense signals accumulated just after tumor cell injection. C, Quantification of MRI signal intensity changes from B. The increase in the accumulation of nanoparticles surrounding the tumor is translated as a drop in the mean signal intensity in the region of interest. D-I, Immunohistochemical analyses of fixed brains from mice analyzed by MRI either twenty-four hours (D-G), or seven days (H, I) after the lesion. Nilo1.sup.+ cells were detected by incubation with a secondary biotinilated anti-hamster Ig antibody and revealed with streptavidin A488. Double labeling with DCX (E), EGFR (G) or Sox2 (I). The MRI-hypointense signals detected 3.5 h after tumor cell injection which are maintained and accumulated with time (up to 7 days) corresponded to Nilo1.sup.+ cells which had arrived at the lesion site as undifferentiated B astrocytes. LV, lateral ventricle. Scale bars: 50 μm.

[0073] FIG. 4. The MRI hypointense signals surrounding the tumor correspond to B astrocytes. Immunohistochemical analyses of fixed brains from mice analyzed in FIG. 2 two days after the Nilo1-mGNP injection (d13). Since Nilo1 mAb was already present in the Nilo1-mGNPs, Nilo1.sup.+ cells were revealed by incubation with a secondary biotinilated anti-hamster Ig antibody and streptavidin-A488. A, B, SVZ ipsilateral to the Nilo1-mGNPs injection site demonstrating the presence of type B astrocytes (Nilo1.sup.+ GFAP.sup.+ or Nilo1.sup.+ Sox2.sup.+) exiting the anterior horn (AH) of the lateral ventricle. C-H, Type B astrocytes were also present at the tumor site vicinity since they were Nilo1.sup.+ GFAP.sup.+ (C-E) or Nilo1.sup.+EGFR.sup.+. F-H, Sites with MRI hypointense signals, indicating that Nilo1.sup.+ cells migrated to the tumor site as B astrocytes. CC, corpus callosum; T, tumor. Scale bars: A-D, 50 μm; E-H and inset in B, 25 μm.

[0074] FIG. 5. Migration of B astrocytes occurred following several types of injury. Immunohistochemistry of fixed brains analyzed by confocal microscopy. A, Three days after a cryolesion, Nilo1.sup.+ cells were present surrounding the lesion site (dotted line, left panel); magnification showing fluorescent (middle) or bright field (right panel) microscopy where Nilo1.sup.+ cells were associated to blood vessels (n=5). B, Seven days following a demyelination, Nilo1.sup.+ cells were found between the niche (LV) and the lesion site (left panel) where they accumulated (right panel) (n=3). C, This movement of Nilo1.sup.+ cells was followed by the appearance of O4.sup.+ and myelin.sup.+ cells 25 days after demyelination (n=3). D, Three days after a mechanical damage produced by stereotaxic injection of PBS (n=6), Nilo1.sup.+ cells were detected surrounding the lesion site (left panel) revealed by nigrosine (right panel) on a bright field microscopy. E, Three days after a mechanical damage produced by stereotaxic injection of PBS, Nilo2.sup.+ neuroblasts were detected filling the lesion site. Scale bars: A and B left panels, 150 μm; A, central and right panels; B right panel, C-E, 50 μm.

[0075] FIG. 6. GFAP.sup.+ Nilo1.sup.+ processes coalesced and filled the damaged tissue. A, Immunohistochemical analyses of fixed brains from the needle track site on mice intracranially injected with CT-2A cells (top row) (n=3) or PBS (bottom row) (n=3) and analyzed by confocal microscopy 24 hours after the lesion. Staining with Nilo1-Cy5 allows detection, in addition to Nilo1.sup.+ cells from SVZ towards the lesion site, of processes that at these early times fill the broken tissue (arrowheads). B, Double immunohistochemistry staining of fixed brains from mice injected with PBS 24 h before, demonstrating that these Nilo1.sup.+ processes were compatible with adult radial glia since they were Nilo1.sup.+ GFAP.sup.+, where associated with PSA-NCAM.sup.+ structures and red blood cells Ter-119.sup.+, CD24.sup.+. These processes did not represent reactive astrocytes since they were vimentin.sup.−. LV, lateral ventricle. Scale bars: A, left panels 50 μm, right panels 25 μm; B, 25 μm.

[0076] FIG. 7. In vivo identification of B astrocytes surrounding a brain tumor after intraperitoneal injection of the Nilo1 mAb. Immunohistochemical analyses of fixed brain from mice injected with GFP-CT-2A cells (GFP.sup.+) in the left striatum, in which Nilo1 was intraperitoneally injected one week after tumor cell injection, and mice (n=4) sacrificed 24 hours later. A, Low and B, high magnification showing that Nilo1.sup.+ cells surrounding the tumor were host-derived (left panel). An accumulation of Nilo1.sup.+ cells was detected surrounding infiltrated tumor cells (right panel). Nilo1.sup.+ cells were revealed with anti-hamster Cy5.5. Scale bars: A, 100 μm; B, 75 μm.

EXAMPLES

1. Materials and Methods

1.1. Animals

[0077] For these experiments, 6-8 weeks old C57B1/6J 103 mice (males), bred and housed under standard conditions were used.

1.2. Antibodies

[0078] Nilo1 mAb was generated by the fusion of hamster B cells and the mouse myeloma X63Ag8, as described [EP2690112A1]. Purification of Nilo1, biotinylation and Cy5 labeling was from ProteinTools (CNB-CSIC, Madrid, Spain).

1.3. Characterization of the Protein G-Magnetic Glyconanoparticles (mGNPs) and Coupling to Nilo1

[0079] Water-soluble magnetic glyconanoparticles, consisting on a magnetic core (4 nm of diameter) covered with a 1 nm gold shell and coated with carbohydrates and an amphiphilic linker ended in a carboxyl group, were prepared and characterized as previously described [Gallo J, Garcia I, Padro D, et al., J. Mat. Chem. 2010; 20:10010]. Recombinant protein G was covalently immobilized to these particles, which enabled the subsequent capture of IgG antibodies.

[0080] Characterization of Protein G-glyconanoparticles (mGNPs) including size and T2* estimation was made as described [Elvira G, Garcia I, Benito M, et al., PLoS One. 2012; 7:e44466]. mGNPs (100 μg) were incubated 5 h at 4° C. with Nilo1 mAb (135 μg) in 0.1 M glycine buffer pH 9.0 on a final volume of 50 μl.

[0081] Characterization of the amount and functionality of the coupled antibody was made as described [Elvira G, Garcia I, Benito M, et al., PLoS One. 2012; 7:e44466].

1.4. Cell Culture

[0082] CT-2A mouse astrocytoma (a gift from Prof. T. N. Seyfried, Boston, Mass., USA), and GFP-CT-2A (a gift from A. Martinez, I. Cajal, CSIC, Madrid, Spain), were grown in RPMI medium, 10% heat-inactivated fetal bovine serum in 5% CO.sub.2 at 37° C. and 95% humidity.

1.5. Isolation and Culture of Cancer-Initiating Cells from Human Glioblastoma Samples

[0083] Glioblastoma (GBM) Tumor Stem Cells were isolated from 4 different human fresh GBM samples. Tissue samples were obtained from patients operated at the Neurosurgery department, Hospital la Fe, Spain.

[0084] GBM Tumor Stem Cells were cultured in media containing: DMEM/F-12 (Gibco, 11039021) with Non Essential Amino Acids (10 mM; Gibco, 11140), Hepes (1M; Gibco, 15630), D-Glucose 45% (Sigma, G8769), BSA-F5 7.5% (Gibco, 15260), Sodium Pyruvate (100 mM; Gibco, 11360), L-Glutamine (200 mM; Gibco), Antibiotic-Antimycotic (100×; Gibco, 15240), N-2 Supplement (100×; Gibco, 17502), Hydrocortisone (1 μg/μl; Sigma, H0135), Tri-iodothyronine (100 μg/ml; Sigma, T5516), EGF (50 ng/μl; Sigma, E9644), bFGF (25 ng/μl; Sigma, F0291) and Heparin (1 ng/μl).

1.6. Surgical Procedures

[0085] Mice were anaesthetized intraperitoneally with 100 mg/Kg of ketamine and 10 mg/Kg of xylacine, their heads were immobilized on a stereotaxic frame and intracranially injected with 1 μl of Nilo1-mGNPs in the right striatum at coordinates +0.9 mm anterior, +0.75 mm lateral, −2.75 mm ventral from bregma point. In control animals, PBS buffer was injected.

[0086] Brain fixations were performed on anesthetized mice by transcardiac perfusion with 4% paraformaldehyde (PF) in 0.1 M phosphate buffer (fixation buffer). Brains were extracted and post-fixed overnight at 4° C. in fixation buffer and cryoprotected in fixation buffer with 30% sucrose for two days at 4° C. before freezing at −80° C. Fixed brains were cut with a cryostat (25 μm thick) and slices were maintained at −20° C. in 30% (v/v) glicerol, 30% (v/v) etilenglicol, PB 0.1M pH 7.4 until analyzed.

[0087] Tumors were generated by grafting 10.sup.2-2×10.sup.5 CT-2A (or GFP-CT-2A) cells intracranially at stereotaxic coordinates +0.1 mm anterior, −2.25 mm lateral, −2.70 mm ventral into the right caudate putamen, in 1 μl of PBS (n=6 for MRI analyses).

[0088] Demyelination was induced by injecting 1 μl of 2% LPC in PBS near the corpus callosum, at stereotaxic positions +1 mm anterior, −1 mm lateral, −2 mm ventral from bregma point, on anaesthetized mice. Mice (n=6) were sacrificed 7 or 25 days later.

[0089] Mechanical injuries were made on anaesthesized mice by inserting a Hamilton needle by stereotaxic surgery (coordinates +0.9 mm anterior, +0.9 mm lateral, −2.75 mm ventral refereed to bregma point) (n=6). In some mice, the lesion site was labeled by nigrosine (0.5 μl i.c. containing 0.5 ng/μl in sterile PBS). Mice were sacrificed one to three days later.

[0090] The cryolesion was generated by applying for 10 s a dry ice pellet onto the left frontoparietal bone of anaesthetized mice. Mice (n=5) were sacrificed three days later.

1.7. Magnetic Resonance Imaging

[0091] MRI studies were performed on a Bruker Biospec 70/20 scanner using a combination of a linear coil (for transmission) with a 4-element mouse head phased array coil (for reception). Animals were anesthetized with sevofluorane (5% for induction and 2% for maintenance) and placed in an MRI-adapted stereotaxic holder. Respiration and body temperature were continuously monitored. MRI acquisition protocol included an initial flash sequence (repetition time: 100 ms, echo time: 6 ms, field of view: 4 cm, matrix: 128×128) to center the Field of View (FOV), followed by a shimming procedure applied to a region of interest covering the head (FOV=3×2×2 cm, matrix=64×64×64) based on a Field Map sequence (TR=20 ms, TE=1.43 and 5.42 ms).

[0092] As an anatomical reference, a T2-weighted axial study (TR=2500 ms; TE, 33 ms; α=180°; FOV=2×2 cm; matrix=256×256; slice thickness=0.5 mm, 15 slices to cover the whole brain) was used and nanoparticles were detected and tracked with a T2*-weighted 3D multi gradient echo (MGE) sequence (TR=200 ms; 8 echoes, TE=10 to 45 ms; echo spacing=5 ms; α=15°; FOV=1.6×1.6×1.5 cm; matrix=192×96×96).

[0093] To increase the signal-to-noise ratio (SNR) and improve image contrast, the different echo images from the MGE sequence were added (in magnitude). To display the results, the tumor area was manually segmented on the T2 axial scans and transferred to the MGE image.

1.8. Immunological Analyses and Staining Procedures

[0094] Single cells suspensions from cultured neurospheres were attached onto Matrigel Basement Membrane Matrix Growth Factor Reduced (BD Pharmingen) pre-coated coverslips with diluted in culture media (1:20 v/v) as described [Elvira G, Garcia I, Benito M, et al., PLoS One. 2012; 7:e44466]. Cells were fixed with 4% PF in PBS buffer for 15 min at RT. Quenching was performed by adding 0.1 M glycine pH 7.4 for 15 min at RT. After three PBS washes, blocking was performed by incubating the coverslips with 10% goat serum in PBS during 1 h at RT. Fixed cells were incubated overnight with Nilo1-mGNPs or purified Nilo1 mAb at 4° C. After three PBS washes, cells were incubated with anti-Ha-FITC secondary antibody 1:100 for 1 h.

[0095] To assess whether MRI signals corresponded to labeled type B astrocytes, brains from mice used in MRI analyses were fixed and cryopreserved for cryostat sectioning in a plane parallel to that of axial MR imaging. Serial 20-25 μm thick frozen sections were collected through the entire mouse brain. Anatomical landmarks such as corpus callosum, lateral ventricles opening, shape and anterior commisure of the brain were used for the spatial alignment of MRI and immunohistochemical sections.

[0096] Brain sections of wild-type mice or mice intracranially injected with Nilo1-mGNPs were blocked with 10% mouse serum in PBS during 1 h at RT and stained with the appropriate antibodies (Table 1).

TABLE-US-00001 TABLE 1 Commercial antibodies used in immunohistochemistry. Ab Antibody Host Source Clone or Cat. # Dilution PRIMARY ANTIBODIES SOX2 Rabbit Chemicon AB5603 1:400 DCX Goat Santa Cruz Biotech. Sc-8066 1:200 Nestin Mouse Chemicon MAB353 1:100 Biotin CD24 Mouse BD Pharmingen 553260 1:200 GFAP Rabbit Dako Z0334 1:3000 Pax6 Rabbit Covance PRB278P 1:300 Anti-BrdU-FITC Mouse Becton Dickinson 347583 1:20 Biotin Ter-119 Mouse BD Pharmingen 553672 1:200 PSA-NCAM Mouse (IgM) Abcys online AbC0019 1:1000 Vimentin Rabbit AbCam Ab 7783 1:200 Oligodendrocyte 2Q92 Mouse AbCam Ab 64547 1:1000 Myelin Mouse Chemicon MAB 328 1:1000 EGFR Rabbit Santa Cruz Biotech. Sc-12357-R 1:100 SECONDARY ANTIBODIES AND REAGENTS BrdU — Sigma B-5002 50 mg/kg (i.p.) Anti-hamster-FITC Mouse BD Pharmingen 554011 1:100 Anti-mouse Alexa Fluor 647 Goat Molecular Probes A-21235 1:300 Anti-mouse IgG-Texas Red Goat Molecular Probes T-862 1:400 Anti-mouse IgM A488 Goat Molecular Probes A-21042 1:400 Anti-rabbit Alexa Fluor 568 Goat Molecular Probes A-11011 1:400 Anti-rabbit IgG-Cy3 Goat Jackson 111-165-003 1:400 Anti-rabbit Alexa Fluor 647 Goat Molecular Probes A-21244 1:400 Anti-goat 594 Chicken Invitrogen A21468 1:400 Anti-hamster biotin Mouse BD Pharmingen 550335 1:100 Streptavidin Texas Red — Molecular Probes S-872 1:400 Streptavidin Alexa Fluor — Molecular Probes S-32354 1:400 DAPI — Invitrogen D1306 0.3 μM Hoechst 33342 — Invitrogen H3570 5 μg/mL

[0097] For the identification of Nilo1.sup.+ cells on NILO1-mGNPs injected mice, anti-Ha-FITC (1:100) or, alternatively, anti-hamster biotin (1:100) followed by streptavidin-A488 (1:400) or streptavidin-Texas Red (1:400) were used. Coverslips were mounted with an anti-fade (Mowioll488), counterstained with either DAPI or Hoechst 33342. Images were collected with a Leica TCS-SP5-AOBS confocal microscope (Mannheim, Germany). Detection ranges were set to eliminate crosstalk between fluorophores.

[0098] For in vivo identification of Nilo1.sup.+ cells surrounding the brain tumor, mice were intraperitoneally injected with Nilo1 ascites at 10 μg/g of body weight one week after stereotaxic injection of 100 GFP-CT-2A cells. Mice were sacrificed 24 h later and 25 μm sections of fixed brains were analyzed using a secondary Cy5.5-labeled anti214 hamster antibody.

[0099] Short BrdU labeling in vivo. Mice were intraperitoneally injected with a single dose of BrdU (50 mg/kg). Mice were sacrificed 1 h later and 25 μm cryostate sections were collected. Fixed brain sections were denatured with 2N HCl in PBS, 0.3% Triton X-100 (PBST) during 30 min at 37° C. After PBST washes, sections were blocked with 10% goat serum in PBST and incubated with a secondary FITC-labeled anti-BrdU antibody during 24 h at 4° C. The Pax6.sup.high BrdU.sup.+ population labeled with this protocol represents the type C transit amplifying progenitors. These samples were additionally stained with Nilo1 mAb.

[0100] C57131/6 mouse embryos were obtained in E10 development stage, fixed by immersion in fixation buffer overnight at 4° C. and cryoprotected in fixation buffer with 30% sucrose for two days at 4° C. before freezing in OCT blocks. Cryostat sections of the embryos were mounted in poly-lysine slides and maintained at −20° C. until analyzed (histology service in CNB-CSIC, Madrid, Spain). Radial glia was identified by using vimentin and nestin antibodies in a Leica TCS-SP5-AOBS confocal microscope (Mannheim, Germany). Images of E10 mouse embryo used to compose overview were collected in a Leica AF6000 LX Live Cell Imaging microscope (Mannheim, Germany).

1.9. Flow Cytometry

[0101] Single cell suspensions from neurospheres were obtained after incubation with accutase (5 min, 37° C.). Unspecific antibody binding was blocked with PBS, 10% goat serum, 3% BSA, 0.0025% NaN.sub.3 for 30 min at 4° C. An excess of Nilo1-mGNPs was added to the cell suspension and incubated for 1 h at 4° C. Cells with purified Nilo1 were incubated as a positive control. After PBS washes, cell suspensions were stained with anti-Ha-FITC (1:100 diluted in PBS, 5% BSA, 0.025% NaN.sub.3). Following additional PBS washes, cells were resuspended in 300 μl cold PBS until FACS measurements (Epics XL, Coulter). Propidium iodine was added (25 μg/ml) to each sample to gate on living cells.

1.10. T2* MRI Signal Quantification

[0102] The accumulation of nanoparticles over time is reflected as an increase in the T2* hypointensity, equivalent to a drop in the overall intensity of the MRI signal. To evaluate the accumulation of nanoparticles, signal intensity levels in the T2*-weighted images were measured over time. For each animal, a region of interest (ROI) including the tumor area and its surroundings was manually delimited on the anatomical images (T2-weighted), at the final time point of the experiment. This ROI was translated to the T2*-weighed images (all images were previously co-registered) and the overall intensity value in the volume of interest was computed for each day of experiment on all the MRI sections. This processing was performed with the software MMWKS (Multimodality WorkStation). Signal intensity levels over time are expressed as percentage of the baseline T2* intensity for each animal and represented as mean±S.D.

2. Results

2.1. Nilo1 mAb Identifies Type B Astrocytes in the SVZ Niche

[0103] Nilo1 antibody was developed and characterized. Nilo1 was described as identifying Sox2.sup.+, GFAP.sup.+, vimentin.sup.+, EGFR.sup.+, DCX.sup.−, PSA-NCAM.sup.−, Tuj1.sup.− cells, suggesting that it identifies a highly undifferentiated neural precursor. It was corroborated that Nilo1 recognized Nestin.sup.+, GFAP.sup.+ and Sox2.sup.+ at the SVZ niche (FIG. 1A). The antigen recognized by Nilo1 mAb was not expressed in ependymal cells at the wall of the lateral ventricles (CD24.sup.+) (FIG. 1B). In addition, Nilo1.sup.+ cells did not correspond to type C cells (identified by a short pulse of BrdU and co-expressing high levels of Pax6), since the BrdU.sup.+Pax6.sup.high and the Nilo1.sup.+ cells represented two distinct populations (FIG. 1B). As control, it is shown that a short pulse of BrdU did not label neuroblasts (FIG. 1B). Furthermore, Nilo1 did not recognize neuroblasts. Moreover, the antigenic phenotype of Nilo1.sup.+ cells allowed us to exclude that they could represent either intermediate progenitors such as NG2-glia, which are GFAP.sup.−, or differentiated astrocytes, since there is no Nilo1 signal in the brain cortex.

[0104] Taken together, these data indicated that Nilo1 mAb identified surface antigens in SVZ-derived type B astrocytes, defined in adult mice as neural stem cells, since Nilo1.sup.+ cells i) had a subependymal localization and did not recognize ependymal CD24.sup.+ cells, ii) did not represent type C cells, iii) identified a population distinct from neuroblasts, iv) did not represent differentiated astrocytes nor intermediate progenitors NG2-glia like, and v) identified SOX2.sup.+, nestin.sup.+ GFAP.sup.+ cells. Further support to this notion came from the observation that on E10 mouse embryos, the radial glia markers vimentin and nestin identified the same cells as Nilo1 (FIG. 1 D).

[0105] In addition, Nilo1 mAb recognized surface antigens in primary human glioblastoma cell lines growing as neurospheres (FIG. 1 E). Indeed, 5 primary human glioblastoma cell lines, derived from 4 different patients, were analyzed and in all of them there were cells positively stained with Nilo1 mAb. These data indicate that this mAb is able to recognize, in addition to the mouse antigens against which they were raised, the homologous antigen in humans.

2.2. In Vivo Tracking of Nilo1.SUP.+ Cells Mobilized Towards an Induced Glioblastoma

[0106] In order to assess whether more undifferentiated precursors (type B cells), despite neuroblasts, were also able to migrate to an induced tumor, Nilo1 mAb was coupled to mGNPs (FIG. 2A), which had been fully characterized elsewhere in terms of size core, shell composition, relaxivity and functionality. Functionality of the Nilo1-mGNPs was evaluated by flow cytometry, immunocytochemistry and immunohistochemistry of the SVZ following in vivo injection of the Nilo1-mGNP particles (FIG. 2B-D).

[0107] A stereotaxic graft of CT-2A astrocytoma cells (10.sup.2-10.sup.4 cells) into the left hemisphere (FIG. 2E) generated a highly reproducible glioblastoma at the injection site, which was followed by T2 magnetic resonance imaging (MRI) over time (FIG. 2F), demonstrating that it was grown on day 11 (FIG. 2G). At this time, Nilo1-mGNP complexes were intracranially injected by stereotaxic surgery into the right hemisphere (FIG. 2E, F) of anaesthetized mice, which were subsequently imaged by MRI using T2* sequences (FIG. 2H). T2*-weighted MRI studies obtained one hour before Nilo1-mGNP injection (d11, −1 h) were used as baseline for the experiment. To establish basal hypointensities on the MRI experiments several controls were made, including mice injected with PBS, to evaluate effects due to intracranial surgery; mice injected with CT-2A cells either alone or in combination with an isotopic control antibody (CD3c, developed in hamster) coupled to mGNPs, to evaluate signals due to either tumor growth or unspecific movement of mGNPs. None of this controls showed defined hypointense signals on T2*-weighted MRI studies.

[0108] The first T2*MRI study after Nilo1-mGNPs injection (d11, 3h) allowed to discard that the complexes were directly deposited in the cerebrospinal fluid (inset on FIG. 2H).

[0109] At this time, in addition to black spots in the lateral ventricles, which corresponded to Nilo1.sup.+ cells in their niches, we detected an increase in the signal hypointensities around the tumor, as compared with the baseline images (FIG. 2H d11, 3 h). These changes could not be explained by neo-vascularization during the four-hour lapse between MRIs. The hypointense T2* signals surrounding the tumor and additional hypointense signals between the niches and the lesion site accumulated over time (FIG. 2H, d12, d13), as quantified in FIG. 2I.

[0110] To discard that at longer experimental times, part of the hypointense signals could be due to neo-vascularization associated with tumor growth, we analyzed the Nilo1.sup.+ cell migration when the tumor mass was not yet formed (FIG. 3A). For this purpose, CT-2A cells were intracranially injected (d 0) four days after Nilo1-mGNP injection (d-4).

[0111] Baseline T2*-weighted MRI images were taken one hour prior to CT-2A injection, failing to detect any hypointense signal at the astrocytoma injection site (FIG. 3B, d0, −1 h). However, as soon as 3.5 hours after injection of the CT-2A cells, an accumulation of hypointense spots at the cell injection site was detected (FIG. 3B, arrowhead d0, 3.5 h) in all analyzed mice. T2-weighted MRI studies (FIG. 3B) enabled the detection of the lesion produced by the stereotaxic injection of the needle to graft the tumor cells (d0, 3.5 h; d2) and the subsequent tumor formation (d7). An increase with time of hypointense signals (T2*-weighted MRI) was detected at this position (FIG. 2B, C) including the needle track (where CT-2A cells could have also been deposited) (d2), accumulating around the tumor when it was formed (d7). During these analyses, tumor induced angiogenesis was undetectable, since in this glioblastoma model increases on days 12 to 14 (FIG. 2H).

[0112] To demonstrate that the hypointense signals detected by T2* MRI corresponded to Nilo1.sup.+ cells, fixed tissues of the mice used in these MRI experiments were analyzed by immunohistochemistry. Since these tissue sections already contained Nilo1-mGNPs, the presence of Nilo1.sup.+ cells was directly revealed with a fluorescently labeled specific secondary antibody.

[0113] In the experiments where Nilo1-mGNPs were injected before the CT-2A tumor cells, as soon as 24 hours after the damage in the left striatum, we observed a SVZ thickening, concomitant with the presence of neuroblasts (DCX.sup.+) and a high number of Nilo1.sup.+ cells dispersed outside their usual subependymal location, infiltrating the adjacent striatum (FIG. 3 D, E). This unusual location for Nilo1.sup.+ cells could represent cells migrating from the lateral ventricle walls towards the adjacent parenchyma. In addition, in vivo labeled Nilo1.sup.+ cells surrounding the lesion site were detected, confirming the MRI data (FIG. 3F). Moreover, Nilo1.sup.+ cells were also EGFR.sup.+, supporting the notion that these cells arrived undifferentiated at the lesion site (FIG. 3G). Even seven days after the damage, Nilo1.sup.+ Sox2.sup.+ cells between the niche and the damage site and surrounding the tumor were still detected (FIG. 3 H, I).

[0114] These data were confirmed by immunohistochemistry analyses on the converse experiment where the CT-2A tumor was formed before the injection of the Nilo1-mGNPs, Nilo1.sup.+ cells crossing the corpus callosum from the SVZ towards the lesion site were revealed 3 days after Nilo1-mGNP injection (13 days after tumor injection) (FIG. 4A, B). Conversely, on animals devoid of lesion, Nilo1.sup.+ cells were circumscribed to the anterior horn and walls of the lateral ventricles (FIG. 1C), showing the migration specificity of the labeled cells. In addition, Nilo1.sup.+ cells expressed additional stem cell markers while migrating (GFAP.sup.+SOX2.sup.+) or even at the final destination surrounding the tumor (GFAP.sup.+SOX2.sup.+EGFR.sup.+) (FIG. 4) suggesting that these cells migrated undifferentiated.

2.3. Migration of Nilo1.SUP.+ Cells is a General Trait for Brain Injury

[0115] To ascertain whether the migration of B astrocytes was restricted to the presence of tumor cells or rather to a more generalized response mechanism to brain damage, the presence of type B astrocytes in the neighborhood of three different kinds of brain injuries was analyzed.

[0116] Firstly a cryolesion model was used, which consists on applying dry ice over the left frontal bone for a short time. This model of brain injury generates an initial tissue damage which leads to secondary processes such as cell death, inflammation, vascular edema and blood-brain barrier disruption. In this model Nilo1.sup.+ cells were detected on the vicinity of the lesion three days after the injury by immunohistochemistry (FIG. 5A).

[0117] In a second lesion model, it was induced a focal demyelinated lesion by lysolecithin (lysophosphatidilcholine, LPC) injection into the left corpus callosum. The initial LPC effect is a localized demyelination during the first week and maintenance of inflammatory signals during the second week. In these mice, seven days after LPC injection, the optimal time for the localized demyelinization, Nilo1.sup.+ cell recruitment around the lesion site was detected (FIG. 5B) in sites containing GFAP.sup.+ or Sox2.sup.+ cells (data not shown). Furthermore, twenty-five days after LPC injection, when remyelination could be confirmed by the presence of O4.sup.+ or myelin.sup.+ cells, neuroblasts adjacent to remyelinated cells were also detected (FIG. 5C).

[0118] The third injury model consisted on a cortical mechanical lesion induced by the puncture with a stereotaxic needle, where the damage site was revealed by injection of nigrosine. Nilo1.sup.+ cells migrating to the damage site one to three days after the mechanical injury were detected (FIG. 5D) in a location similar to that where neuroblasts were identified (FIG. 5E).

2.4. Migration of Nilo1.sup.+ Cells is Associated with Adult Radial Glia

[0119] Few hours after injuries produced by needle injection, the wound surrounding the needle track was filled by structures expressing Nilo1 (FIG. 6A). These processes might represent adult radial glia, since they express not only Nilo1 but also the glial marker GFAP, are associated to PSA-NCAM and to the erythrocyte markers Ter119.sup.+, CD24.sup.+, whereas did not represent reactive astrocytes since they were negative for vimentin and CD11b (FIG. 6B and data not shown).

2.5. Intraperitoneally Injected Nilo1 mAb Enables In Vivo Identification of Brain Glioblastomas

[0120] Since high grade brain tumors induce the blood brain barrier breakdown, one week after intracranially injecting CT-2A-GFP cells, a single dose of Nilo1 mAb (10 μg/g of body weight) was administered intraperitoneally and mice were sacrificed the next day. In the fixed brains, the presence of Nilo1 was revealed by immunohistochemistry with a fluorescently labeled secondary antibody, demonstrating the presence of type B cells surrounding the GFP.sup.+ glioblastoma cells (FIG. 7A). The use of GFP tumor cells allowed the identification of host cells (Nilo1.sup.+ GFP.sup.−, arrowheads in FIG. 7B) in areas where isolated tumor cells infiltrated the parenchyma. This experiment corroborated the notion that type B astrocytes migrate towards the tumor site and opens up the possibility of using intraperitoneally injected antibodies to follow brain damage in situations where the blood brain barrier is disrupted.

3. Conclusions

[0121] Although there are evidences that immature neurons present in brain lesion sites, such as peri-infarcted tissue, come from GFAP.sup.+ SVZ-derived neural stem cells, a direct mobilization of type B astrocytes towards a lesion site had not yet been reported.

[0122] To study the migration of early neural progenitors towards a brain injury, Nilo1 mAb was used in the present invention. Here it has been demonstrated that Nilo1 identified type B astrocytes (or adult neural stem cells) since i) it recognized GFAP.sup.+, Sox2.sup.+ and EGFR.sup.+ cells, being negative for the neuroblast marker DCX; ii) Nilo1 did not stain ependymal CD24.sup.+ cells, having instead subependymal localization on the SVZ; iii) transient amplifying cells (or type C cells, which are Pax6.sup.highBrdU.sup.+ on tissue sections from in vivo labeled animals with a short BrdU pulse) and Nilo1.sup.+ cells represented two distinct populations; iv) Nilo1.sup.+ cells were GFAP.sup.+, excluding the possibility that this mAb identified intermediate progenitors in the differentiation process, like NG2-precursors which are GFAP.sup.−; v) double staining of E10 embryo brains with Nilo1 and either vimentin or nestin suggested that Nilo1 recognizes embryonic radial glia.

[0123] Nilo1 mAb, coupled to functionalized-magnetic nanoparticles in combination with magnetic resonance imaging allowed to identify the Nilo1.sup.+ cells at their niche in the SVZ and, in animals carrying a lesion, at intermediate positions between the niche and the lesion site, where they accumulated over time. The migratory response was very fast since MRI hypointense signals at the lesion site were detected by MRI 3 h after injection of the Nilo1-mGNP complexes in animals bearing tumors, or 3.5 h after injection of the CT-2A cells on animals that had previously been injected with Nilo1-mGNP complexes. The presence of Nilo1.sup.+ cells at the lesion site was corroborated by immunohistochemistry analysis on fixed brain sections from these mice, incubating with a fluorescent secondary antibody, since the Nilo1.sup.+ cells were already labeled in vivo with the Nilo1-mGNPs. Immunohistochemistry confirmed not only that the hypointense signals corresponded to Nilo1.sup.+ cells surrounding the tumor, but also that these cells retained their type B astrocyte phenotype (GFAP.sup.+, EGFR.sup.+, Sox2.sup.+) following their migration.

[0124] The migration of adult neural stem cells (type B astrocytes) towards a tumor site, rather than a tumor-specific response, turned out to be a more generalized response to brain insults since Nilo1.sup.+ cells at the sites where other types of lesions were produced were also detected (i.e. cryolesion, demyelination, mechanical injury). These lesions were chosen because they are very different from each other and from the tumor implantation model described above. In all of them, Nilo1.sup.+ cells were detected surrounding the lesions.

[0125] In addition to Nilo1.sup.+ cells migrating to the lesion sites, both in a mechanical lesion model and after any stereotactic injection, it was detected, one day after the lesion, coalescing structures filling the wound-lesion that expressed high levels of Nilo1 antigen. These Nilo1.sup.+ processes were GFAP.sup.+, vimentin.sup.− and CD11b.sup.−, forming structures associated to PSA-NCAM. On the one hand, the observation that these structures did not express vimentin or CD11b allowed to exclude that they represented proximal reactive astrocytes. On the other hand, their GFAP.sup.+ staining was compatible with the presence of adult radial glia fibers or glial tubes. Radial glia cells have been defined as neuronal precursors and it has been suggested that they could represent a specific subpopulation of astrocytes in adult mammals. Furthermore, the presence of radial glia in adult hippocampus where neurogenesis occurs throughout life, or in non-mammalian vertebrates where neurogenesis persists in a rather wide-spread fashion in the adult brain, raise the possibility that the neurogenic potential of radial glia may extend into adulthood in some brain regions or even as an acute response to brain lesions, were neurogenesis is necessary and commonly associated to brain tissue repairing processes.

[0126] The migration of neural stem cells reported here is in fully accordance with data showing that stimulation with exogenous growth factors increase cell proliferation at the SVZ and promote the migration of SVZ-derived cells into the adjacent parenchyma or even at the vicinity of a brain lesion site.

[0127] Finally, the data presented here represents a proof that MRI using Nilo1-GMPs can be used in humans for detection of a brain primary tumor or a recidive very early on, before contrast substances such as gadolinium would give a detectable signal. In particular, since it has been shown herein that Nilo1 mAb is able to also identify the corresponding antigens present in cells derived from human glioblastomas.