BACKGROUND OF THE INVENTION Multiple sclerosis (MS) is an inflammatory demyelinating disease of the central nervous system, also known as disseminated sclerosis or encephalomyelitis disseminata. This damage disrupts the ability of parts of the nervous system to communicate, resulting in a wide range of signs and symptoms, including physical, mental, and sometimes psychiatric problems. MS takes several forms, with new symptoms either occurring in isolated attacks (relapsing forms) or building up over time (progressive forms). Between attacks, symptoms may disappear completely; however, permanent neurological problems often occur, especially as the disease advances.

Relapsing-remitting multiple sclerosis (RRMS) is characterized by an unpredictable relapsing-remitting course, wherein clearly defined attacks of worsening neurologic function. - often called relapses, flare-ups or exacerbations - are followed by partial or complete recovery periods (remissions). During recovery periods symptoms improve partially or completely, and there is no apparent progression of disease. RRMS is the most common disease course at the time of diagnosis. Approximately 85 percent of people are initially diagnosed with RRMS, compared to 10-15 percent with progressive forms of the disease.

Diagnosis of MS involves blood tests, lumbar punctures, evoked potentials and magnetic resonance imaging.

Today MRI diagnosis and therapeutic assessment of multiple sclerosis is mostly based on quantification of contrast-enhancing lesions and change of lesion load (Polman et al., 2011, Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol. 2011 Feb;69(2):292-302).

New tools are required to improve the sensitivity of imaging in MS and to provide valuable information for disease physiopathological and therapeutic assessment. In particular, there is an unmet need for methods for predicting the risk of relapse in patients suffering from RRMS.

Indeed, the ability to predict the occurrence of a subsequent relapse (yes/no) and to estimate the time when that process will occur has important clinical and practical implications. This knowledge can help in decisions related to treatment - e.g. either treat patients with more aggressive disease or avoid over-treatment of patients with a more favorable disease course. Prediction of the time to next relapse can also be useful in the design of clinical trials as an additional criterion for selecting active patients. For patients with clinically isolated syndrome (CIS), who have just experienced the first relapse, this knowledge may allow to estimate the probability to convert to definite MS by predicting the time until the second relapse.

Leukocyte recruitment is mediated by adhesion molecule expressed on endothelial cells (Rossi et al., 2011, Vascular inflammation in central nervous system diseases: adhesion receptors controlling leukocyte-endothelial interactions J. Leukoc. Biol., 89 (2011), pp. 539-556).

Selectins are important cell adhesion molecules, with high affinities for carbohydrate moieties. They play a prominent and critical role in the initial stages of circulating cellular components and vascular wall interactions by mediating leucocytes/platelet and leucocytes/endothelium interactions. Three types of selectins have been discovered so far: P- selectin, E-selectin and L-selectin. L-selectin is constitutively expressed on almost all circulating leukocytes. The expression of E-selectin is inducible on vascular endothelium upon activation by various mediators including cytokines and endotoxin. P-selectin is contained in intracytoplasmic granules and is rapidly translocated to platelet or endothelial surfaces after cell exposure to thrombin or histamine.

The P-, L- and E-selectins are structurally similar transmembrane proteins. They all possess large, highly glycosylated, extracellular domains, a single spanning transmembrane domain, and a small cytoplasmic tail. At their extracellular amino termini, they have a single calcium-dependent (or C-type) lectin domain (L) followed by an epidermal growth factor (EGF)- like domain (E) and several complement regulatory domains (C). Selectin-mediated cell adhesion results from calcium-dependent interactions of the amino-terminal lectin domain with a large variety of carbohydrate-presenting molecules on the surface of target cells. While the affinity of each of the selectins varies depending on the ligand, they all bind a specific tetrasaccharide carbohydrate structure known as sialyl Lewis X (SLe ), which contains sialic acid and fucose residues.

Selectins are considered as potentially useful markers for the diagnosis of some of these pathologies. Numerous efforts are in progress to image selectins predominantly through Magnetic Resonance Imaging (MRI) (S. Bouty et al, Contrast Media Mol. Imaging, 2006, 1: 15-22), scintigraphy (G. Hairi et al, Ann. Biomed. Eng., 2008, 36: 821-830), and more recently using ultrasons (F.S. Villanueva et al, Nat. Clin. Pract. Cardiovasc. Med., 2008, 5: S26-S32). Most selectin imaging agents developed so far are anti-selectin antibodies (B.A. Kaufman et al, Eur. Heart J., 2007, 28: 2011-2017; G. Hairi et al, Ann. Biomed. Eng., 2008, 36: 821-830; K. Licha et al, J. Biomed. Opt., 2005, 10: 41205; and P. Hauff et al, Radiology, 2004, 231 : 667-673) and sialyl Lewis X analogs or derivatives (S. Bouty et al, Contrast Media Mol. Imaging, 2006, 1: 15- 22; F.S. Villanueva et al, Circulation, 2007, 115: 345-352). These imaging agents have been suggested to allow the in vivo non-invasive detection of selectins in inflammation, neurodegenerative disorders, cancer and thrombosis.

However, to date, there exists no reliable method for predicting the risk of relapse in an asymptomatic patient suffering from RRMS, or for predicting the chances of remission in a symptomatic patient suffering from RRMS. SUMMARY OF THE INVENTION

The inventors have found that imaging targeting-p-selectin allows non-invasive measurement of vascular inflammation and can thus be used as additional information inaccessible to conventionally used imaging techniques for the diagnosis, follow-up and prognosis of MS patients.

The inventors have developed a molecular imaging technique which can reveal the blood- spinal cord barrier (BSCB) dysfunctions together with the immune cell infiltration. This imaging technique can accurately predict the onset of neurological symptoms, notably in asymptomatic patients, but also the appearance of remission.

Thus, the present invention relates to an imaging agent for use in a method for predicting the onset of neurological symptoms or the remission from neurological symptoms in a patient suffering from remitting-relapsing multiple sclerosis, wherein said imaging agent comprises a selectin ligand. The invention also relates to the use of spinal cord P-selectin as a biomarker for multiple sclerosis. The invention also provides a method for predicting the onset of neurological symptoms or the remission from neurological symptoms in a patient suffering from remitting-relapsing multiple sclerosis, said method comprising the steps of:

- administering to said patient a detectable amount of an imaging agent comprising a selectin ligand,

- subjecting said patient to MRI,

- collecting the MRI signal,

- quantifying the percentage of signal void,

- comparing said percentage of signal void to a predetermined value,

wherein an increase in the percentage of signal void compared to the predetermined value is indicative of a risk of relapse in a patient who does not present any neurological symptoms, and/or wherein a decrease in the percentage of signal void compared to the predetermined value is indicative of a risk of remission in a patient who presents neurological symptoms.

DETAILED DESCRIPTION OF THE INVENTION

As described above, the present invention relates to an imaging agent for use in a method for predicting the onset of neurological symptoms or the remission from neurological symptoms in a patient suffering from remitting-relapsing multiple sclerosis (RRMS), wherein said imaging agent comprises a selectin ligand.

As used herein, the term "patient" refers to a human or another mammal (e.g., mouse, rat, rabbit, hamster, dog, cat, cattle, swine, sheep, horse or primate). In many embodiments, the subject is a human being. As used herein, the term "imaging agent" refers to a compound that can be used to detect specific biological elements (e.g., biomolecules) using imaging techniques. Imaging agents of the invention are molecules comprising a selectin ligand associated with at least one detectable moiety. Imaging agents of the present invention can be used to detect selectins in in vitro and ex vivo biological systems as well as in subjects.

The detectable moiety according to the invention is typically detectable through any one of the following imaging techniques Planar Scintigraphy (PS) or Single Photon Emission Computed Tomography (SPECT), Positron Emission Tomography (PET), ultrasound imaging such as contrast-enhanced ultrasonography (CEUS), Magnetic Resonance Imaging (MRI) or fluorescence spectroscopy.

Typically, said method for predicting the onset/relapse of neurological symptoms is carried out by Magnetic Resonance Imaging (MRI), or by ultrasound imaging such as CEUS.

Typically, said method for predicting the onset/relapse of neurological symptoms is carried out by imaging the central nervous system of said patient. In one embodiment, said imaging technique, notably the MRI or ultrasound imaging, is carried out on the brain of said patient.

In another embodiment, said imaging technique is carried out on the spinal cord of said patient.

In one embodiment, said imaging technique, notably MRI imaging or ultrasound imaging, is carried out on the upper spinal cord, preferably the cervical spinal cord.

In one embodiment, said imaging technique, notably MRI imaging or ultrasound imaging is carried out on the lower spinal cord, preferably the lumbar spinal cord.

Without wishing to be bound by theory, it is believed that, in certain cases of MS, lesions in the spinal cord appear prior to brain lesions. Indeed, it has been reported that presence of asymptomatic spinal lesions may help confirm a diagnosis of MS when few or no brain lesions are present (Lycklama et al. Spinal-cord MRI in multiple sclerosis. Lancet Neurol. 2003 Sep;2(9):555-62). As used herein, the term "neurological symptoms" refers to the clinical manifestations of MS. They include, but are not limited to: numbness or weakness in one or more limbs,

partial or complete loss of vision, usually in one eye at a time, often with pain during eye movement,

Prolonged double vision,

- Tingling or pain in parts of the patient's body

Electric- shock sensations that occur with certain neck movements, especially bending the neck forward (Lhermitte sign)

Tremor, lack of coordination or unsteady gait

Slurred speech

- Fatigue

Dizziness

Problems with bowel and bladder function.

The patient according to the invention can be asymptomatic. Typically such a patient does not exhibit some of the more specific symptoms as mentioned above. For example, such a patient does not exhibit:

numbness or weakness in one or more limbs,

partial or complete loss of vision, usually in one eye at a time, often with pain during eye movement,

Prolonged double vision,

Tingling or pain in parts of the patient's body

Electric- shock sensations that occur with certain neck movements, especially bending the neck forward (Lhermitte sign)

Tremor, lack of coordination or unsteady gait

Slurred speech

Preferentially, an asymptomatic patient does not exhibit any neurological symptoms as previously described.

The method of the invention can predict the risk of relapse in a patient suffering from RRMS. As used herein, the term "relapse" has its general meaning in the art. It refers to episodes of inflammatory activity. A relapse is defined by the occurrence of new or worsening neurological symptoms, during at least 24 hours, separated from the previous relapse by at least one month. Relapses are also called attacks, flare-ups, or exacerbations.

Advantageously, the method for predicting a relapse according to the invention can be useful for selecting and/or adapting the appropriate treatment regimen. Typically, if a risk of relapse is identified, the physician will prescribe a treatment in order to attenuate or prevent the neurological symptoms in said patient. Prior to the development of the prediction tool according to the present invention, treatment could only be initiated once the relapse had actually occurred, i.e. after the onset of the neurological symptoms.

Without wishing to be bound by theory, it is believed that the method of the invention enables a better management of the neurological symptoms.

Conversely, the method of the invention can predict the chances of remission from neurological symptoms in a patient suffering from RRMS who is experiencing symptoms at the time the imaging is carried out. This knowledge can also be very helpful in decisions related to treatment - e.g. it may avoid over-treatment of patients with a favorable disease course.

As used herein, "remission" has its general meaning in the art. It refers to a period between two relapse episodes, during which the number and/or intensity of the neurological symptoms is reduced.

The invention also provides a method for predicting the onset of neurological symptoms or the remission from neurological symptoms in a patient suffering from remitting-relapsing multiple sclerosis, said method comprising the steps of:

- administering to said patient a detectable amount of an imaging agent comprising a selectin ligand,

- subjecting said patient to an imaging technique,

- collecting the imaging signal from the imaging agent,

- quantifying the signal,

- comparing the level of said signal to a predetermined value, wherein an increase in the level of the signal compared to the predetermined value is indicative of a risk of relapse in an asymptomatic patient, typically who does not present any neurological symptoms,

and/or wherein a decrease in the level of the signal compared to the predetermined value is indicative of a chance of remission in a patient who presents some neurological symptoms.

Typically, the invention provides a method for predicting the onset of neurological symptoms or the remission from neurological symptoms in a patient suffering from remitting-relapsing multiple sclerosis, said method comprising the steps of:

- administering to said patient a detectable amount of an imaging agent comprising a selectin ligand,

- subjecting said patient to MRI,

- collecting the MRI signal,

- quantifying the percentage of signal void,

- comparing said percentage of signal void to a predetermined value,

wherein an increase in the percentage of signal void compared to the predetermined value is indicative of a risk of relapse in a patient who does not present any neurological symptoms, and/or wherein a decrease in the percentage of signal void compared to the predetermined value is indicative of a chance of remission in a patient who presents neurological symptoms.

Prediction is based on the comparison of the signal collected by MRI in a given patient to a reference signal.

Typically, the MRI signal for a patient can be quantified by quantifying the percentage of signal void in the images collected from said patient, wherein the percentage of signal void is defined as the percent ratio of the selectin positive signal over the total area of the imaged zone (e.g. brain and/or spinal cord). The percentage of signal void is representative of the number of lesions present in the brain and/or spinal cord.

Typically said percentage of signal void is compared to a predetermined value vl or v2. If the percentage of signal void measured for a patient is higher than the predetermined value vl, the patient is deemed to be at risk of presenting a relapse. Conversely, if the percentage of signal void measured for said patient is lower than the predetermined value v2, the patient is deemed to have an increased chance of entering into remission. Typically, the predetermined values can standard values observed in the general population of healthy control individuals. Alternatively, the predetermined values can be values corresponding to patients suffering from MS at predetermined stages of the disease. In one embodiment, the predetermined values are values obtained during previous MRI imaging carried out on the same patient. In this case, the method of the invention serves to monitor the evolution of the RRMS in a given patient. According to this embodiment, an increase in the number of lesions is indicative of a risk of relapse. A decrease in the number of lesions is indicative of a chance of remission.

In another embodiment, the method for predicting the onset of neurological symptoms or the remission from neurological symptoms in a patient suffering from remitting-relapsing multiple sclerosis comprises the steps of:

- administering to said patient a detectable amount of an imaging agent comprising a selectin ligand,

- subjecting said patient to ultrasound imaging such as CELTS,

- collecting the ultrasound signal,

- quantifying the mean acoustic signal from retained imaging agents

- comparing said mean acoustic signal to a predetermined value,

wherein an increase of the mean acoustic signal compared to the predetermined value is indicative of a risk of relapse in a patient who does not present any neurological symptoms,

and/or wherein a decrease in the percentage of signal void compared to the predetermined value is indicative of a chance of remission in a patient who presents neurological symptoms.

Prediction is based on the comparison of the signal collected by ultrasound imaging in a given patient to a reference signal.

According to the present invention, the imaging agent comprises a selectin ligand.

As used herein, the term "selectin" has its art understood meaning and refers to any member of the family of carbohydrate-binding, calcium-dependent cell adhesion molecules that are constitutively or inductively present on the surface of leukocytes, endothelial cells or platelets. The term "E-selectin", as used herein, has its art understood meaning and refers to the cell adhesion molecule also known as SELE; CD62E; ELAM; ELAM1 ; ESEL; or LECAM2 (Genbank Accession Numbers for human E-selectin: NM_000450 (mRNA) and NP_000441 (protein)). As used herein, the term "L-selectin" has its art understood meaning and refers to the cell adhesion molecule also known as SELL; CD62L; LAM-1 ; LAM1 ; LECAM1 ; LNHR; LSEL; LYAM1; Leu-8; Lyam-1; PLNHR; TQ1; or hLHRc (Genbank Accession Numbers for human L- selectin: NM_000655 (mRNA) and NP_000646 (protein)). The term "P-selectin", as used herein, has its art understood meaning and refers to the cell adhesion molecule also known as a SELP; CD62; CD62P; FLJ45155; GMP140; GRMP; PADGEM; or PSEL (Genbank Accession Numbers for human P-selectin: NM_003005 (mRNA) and NP_002996 (protein)).

As used herein, the term "selectin ligand" refers to compound that specifically binds to selectin. Selectin ligands include, but are not limited to, anti-selectin antibodies, sialyl Lewis X analogs and fucoidans.

In one embodiment, the selectin ligand is an antibody that specifically binds to selectin, preferably P-selectin.

Anti P-selectin antibodies have been extensively described in the art. Typically, the anti P-selectin antibody can be the polyclonal goat anti-mouse antibody commercialized by R&D Systems under reference AF737.

In one embodiment, said anti P-selectin antibody is a humanized antibody recognizing human P-selectin. Methods for obtaining such a antibodies are known in the art.

Any MRI imaging agent that can be coupled to a selectin ligand can be used in the method according to the invention.

In one embodiment said imaging agent is selected in the group consisting of microparticles of iron oxide, iron-containing colloidal particles and gadolinium chelates.

Suitable imaging agents that specifically bind to selectin have been disclosed in the following international patent applications: WO2007/020450 (Isis Innovation Limited) and WO2010116209 (INSERM), the content of which is hereby incorporated by reference. Any ultrasound imaging agent that can be coupled to a selectin ligand as described above can also be used in the method according to the invention.

In one embodiment said imaging agent is selected from an acoustically active (gas-filled) microbubble or an acoustically active lipid particle (i.e.: a gas-filled liposome).

Example of suitable ultrasound imaging agents that may be used in the practice of the invention have been described in Lindner JR et al., Circulation 2001, 104:2107-2112 and in Villanueva FS & Wagner WR, Nat Clin Pract Cardiovasc Med 2008, 5:S26-S32). The invention also relates to the use of P-selectin as a biomarker for multiple sclerosis.

Preferably, the invention relates to the use of spinal cord P-selectin as a biomarker for multiple sclerosis.

The invention also provides a method for treating a patient suffering from RRMS, said method comprising the steps of:

- administering to said patient a detectable amount of an imaging agent comprising a selectin ligand,

- subjecting said patient to an imaging technique such as MRI or ultrasound imaging,

- collecting the imaging signal (i.e.: MRI signal or ultrasound signal),

- quantifying the level of the signal (typically the percentage of signal void for MRI or the mean acoustic signal for ultrasound imaging),

- comparing said level of the imaging to a predetermined value,

- if an increase in the level of the imaging signal compared to the predetermined value is observed, administering a treatment in order to attenuate or prevent the neurological symptoms in said patient, or

- if a decrease in the level of the imaging signal compared to the predetermined value is observed, reducing the treatment for attenuating or preventing the neurological symptoms .

FIGURES LEGENDS

Figure 1: Schematic representation of P-Sel/MPIOs (microparticles of iron oxide). Figure 2: P-Scl/MIPOs allow diagnosis and spatio-temporal monitoring of inflammation in spinal cord of mice with chronic EAE. (A) Experimental design of the experiments. (B) Typical course of chronic EAE. Evolution of clinical score is given in function of time. (C, E) Representative images of molecular imaging for P-Sel showing P-Sel positive vessels in lower (C) and upper (E) spinal cord in sham mice and EAE-MOG mice at CS1, CS2, CS3 and CS4. (D, F) Corresponding quantification of signal void (D: lower spinal cord, E: upper spinal cord) due to P-Sel MPIOs (n=5 per group, *p< 0.05, NS: non significant). Figure 3: Spatial and temporal dissemination of blood-spinal cord opening in mice with chronic EAE. (A) Images of Evan's blue extravasation in the spinal cord of sham and EAE animals at indicated clinical scores. (B) Ex vivo detection of Evan's Blue fluorescence in the same animals and (C) corresponding quantifications (n=5 per group, *p< 0.05). Figure 4: P-Sel MIPOs allow diagnosis and spatio-temporal monitoring of inflammation in spinal cord of mice with relapsing-remitting EAE. (A) Experimental design of the experiments. (B) Typical course of relapsing-remitting EAE. Evolution of clinical score is given in function of time. (C, E) Representative images of molecular imaging for P-Sel showing P-Sel positive vessels in lower (C) and upper (E) spinal cord in sham mice and EAE-MOG mice during presymptomatic, surge, recovery and relapse phases. (D, F) Corresponding quantification of signal void (D: lower spinal cord, E: upper spinal cord) due to P-Sel/MPIOs (n=5 per group, *p< 0.05, NS: non significant).

Figure 5: Spatial and temporal dissemination of blood-spinal cord opening in mice with relapsing-remitting EAE. (A) Images of Evan's blue extravasation in the spinal cord of sham and EAE animals at indicated clinical phases. (B) Ex vivo detection of Evan's Blue fluorescence in the same animals and (C) corresponding quantifications (n=5 per group, *p< 0.05).

Figure 6: Prediction of first relapse and remission by P-Sel MPIOs in relapsing-remitting EAE. (A) Representative images of molecular imaging for P-Sel showing P-Sel positive vessels in upper and lower spinal cord in mice with a clinical score of 0 (presymptomatic phase). Left: animals with no or negligible P-Sel staining, Right: animals showing substantial P-Sel staining. Clinical score at the day of P-Sel imaging (dO) and the day after (dl) are indicated. (B) Representative images of molecular imaging for P-Sel showing P-Sel positive vessels in upper and lower spinal cord in mice with a clinical score of 3 (first relapse phase). Right: animals showing substantial P-Sel staining, Left: animals with no or negligible P-Sel staining. Clinical score at the day of P-Sel imaging (dO) and the day after (dl) are indicated.

Figure 7: Prediction of first relapse in asymptomatic relapsing-remitting EAE animal by P- Sel MPIOs. (A) Time course of the experimental session. (B) Representative images of molecular imaging for P-Sel obtained from asymptomatic animals (with a clinical score of 0)10 days after PLP-EAE induction and showing P-Sel positive vessels in upper and lower spinal cord. Left: animals with no or negligible P-Sel staining, Right: animals showing substantial P-Sel staining. (C) Mean percentage of signal void measured in both groups. (D) Mean clinical scores at the day of P-Sel imaging (Day 10) and the day after (Day 11). (E) Four asymptomatic mice exhibited low P-Sel staining at the day of MRI (D10) all of them showed no relapse the day after. Conversely, four other asymptomatic mice exhibited high P-Sel staining at the day of MRI, all of them had relapse the day after.

Figure 8: Prediction of remission in symptomatic relapsing-remitting EAE animals by P- Sel MPIOs. (A) Time course of the experimental session. (B) Representative images of molecular imaging for P-Sel obtained from animals at EAE onset (with a mean clinical score around 2) showing P-Sel positive vessels in upper and lower spinal cord. Left: animals showing substantial P-Sel staining. Right: animals with no or negligible P-Sel staining. (C) Mean percentage of signal void measured in both groups. (D) Mean clinical scores at the day of P-Sel imaging (Day 15) and the day after (Day 16). (E) Four symptomatic mice exhibited high P-Sel staining at the day of MRI, none of them had remission the day after. Conversely, four other symptomatic mice at the day of MRI (D15) exhibited low P-Sel staining all of them had a remission the day after. Figure 9: Clearance of the P-Sel-MRI signal from the tissues (A) Time course of the experimental session. (B) Representative images of molecular imaging for P-Sel obtained from EAE animals and showing P-Sel positive vessels in upper and lower spinal cord. MRI sessions were performed 10 min (upper panel), 5 hours (middle panel) and 24 hours (lower panel) after MPIO-P-Sel injection. (C): Mean percentage of signal void obtained at each MRI session in caudal (C) and rostral (D) spinal cord.

Figure 10: Statistical Correlation between P-SEl MRI signal and lymphocyte infiltration in relation to clinical scores (CS). CD4+ infiltration increases with the EAE clinical score (from Sham to CS4) and is also positively correlated with the percentage of signal void, both in rostral (left panel) and caudal (right panel) spinal cord.

EXAMPLES Material and methods

Experimental Autoimmune Encephalomyelitis (EAE):

Chronic EAE was induced in 9- week-old male C57BL6/J mice. Mice were immunized subcutaneously by two investigators blinded to treatment with 200 ; g of MOG _{35 55} peptide (EAE animals) or phosphate buffered saline (Sham-operated animals) in Complete Freund Adjuvant (CFA) (Sigma) containing 800 ; g of M. tuberculosis (H37Ra, Difco). All animals receive 200 ng of pertussis toxin (Sigma) intraperitoneally (i.p.) at the time of immunization and 48 hours later.

Relapsing remitting EAE was induced in 7-8 week-old SJL/J female mice (Janvier, France). Mice were immunized subcutaneously by two investigators blinded to treatment with 200 μ g of PLP ₁₃ 9_i ₅₁ peptide (EAE animals) or phosphate buffered saline (Sham-operated animals) in Complete Freund Adjuvant (CFA) (Sigma) containing 800 ,u g of M. tuberculosis (H37Ra, Difco). The emulsion was administered at four injection sites above the shoulder and the flanks. All animals were additionally intraperitoneally (i.p.) injected with 200 ng pertussis toxin derived from Bordetella pertussis (Sigma) in 200 μL· saline at the time of, and after 48 hours following immunization.

Clinical score (CS) of EAE: Mice were examined daily for clinical signs of EAE and were scored as followed: CS 0, no disease; CS1, limp tail; CS2, hindlimb weakness; CS3, complete hindlimb paralysis; CS4, hindlimb paralysis plus forelimb paralysis; and CS5, moribund or dead. Animal were euthanized if they showed a loss of weight superior to 10%. Clinical score was assessed daily by one person blinded to the treatment.

Targeting-moiety conjugation to MPIOs and molecular imaging : Polyclonal goat anti-mouse antibodies for P-selectin (R&D Systems, AF737), were covalently conjugated to MPIOs (4(^g of antibody per mg of MPIOs). Mice received intravenous injection of 2.0mg Fe kg of conjugated MPIOs (200μί) and were immediately subjected to contrast-enhanced MRI. 3D T2*-GEFC (weighted gradient echo imaging with flow compensation) images presented are minimum intensity projections of 5 consecutive slices (Z resolution: 350μπι). Signal voids quantification on 3D T2*-weighted images using automatic triangle threshold in Image J software (vl.45r). The quality of conjugated MPIOs was systematically checked in a naive mouse, by stereotaxic injection of lipopolysaccharide (LPS; ^g in Ιμί) in the striatum (0.5mm anterior, 2.0mm lateral, -3mm ventral to the bregma).

Ex vivo quantification of Evan's blue extravasation: Evan's Blue (200 μΐ, 2%) was intravenously injected in EAE mice 4h before termination and quantified in perfused spinal cords using the BioSpace™ Photoimager.

Tissue sampling, histology and immunohistochemistry: Mice were perfused with heparinised saline. Brains and spinal cords were harvested. Cryomicrotome-cut sections were incubated in the presence of the corresponding antibodies. All sections were examined with a Leica DM6000 microscope. Images were digitally captured using a coolsnap camera and visualized with Metavue software.

Statistical analyses: Results are the mean + SEM. Statistical analyses were performed using Kruskal-Wallis followed by Mann- Whitney's U-test. Statistical significance was concluded for p<0.05. All allocation of animals to experimental and control groups were randomized, and masking of observers to group assignment were incorporated into the design of the study.

Results:

P-Sel/MPIOs allow monitoring of the spatial and temporal dissemination of inflammation in spinal cord of EAE mice. Our first step was to develop a tool for molecular imaging based on the immunodetection of P- Sel adhesion molecule (Figure 1). We selected an antibody against P-Sel for its specificity and its capacity to stain the neuroendothelium in EAE, but not sham, animals (data not shown). These antibodies were then coupled to MPIOs (Figure 1), which we showed were able to bind the endothelium in inflammatory animals after i.v. injection (data not shown).

Using this tool, we were able to show a spatial and temporal dissemination of endothelial activation in chronic EAE (Figure 2). Animals were injected i.v. with P-Sel/MPIOs (Figure 2a) at different stages of chronic EAE corresponding to clinical scores 1 to 4 (Figure 2b). Molecular MRI in the lower and upper part of the spinal cord showed a gradual increase in signal void in a caudo-rostral direction along with the increase in clinical score. (Figure 2c-f) This spatio- temporal dissemination matched the caudo-rostral gradient over time of blood-spinal cord opening shown by the optical and fluorescent assessment of Evan's Blue extravasation (Figure 3a-c). The downstream infiltration of T4 lymphocytes and neutrophils also occurred following this caudo-rostral progression over time (data not shown).

In a second part, we applied the same methods to assess the relevance of this imaging tool in relapsing-remitting EAE (Figure 4-5). P-Sel/MPIOs revealed endothelial activation in early disease, during surge and during relapse in lower and upper parts of the spinal cord (Figures 4c- f), while little if any hyposignal was observed in remitting animals (recovery, Figures 4c-f). Confirming what shown above in chronic EAE, signal void in relapsing remitting EAE matched blood-spinal cord opening (Figure 5) and leukocyte infiltration (data not shown).

Overall, these data indicate that P-Sel/MPIOs enable a follow-up by molecular imaging of disease progression in the spinal cord of animals subjected to two EAE models mimicking the progressive and relapsing-remitting forms of MS. Moreover, the hyposignal due to P-Sel/MPIOs is representative of endothelial activation, blood-spinal cord barrier opening and leukocyte infiltration.

P-Sel/MPIOs allow prediction of first relapse in asymptomatic animals and remission in symptomatic animals.

We further assessed whether molecular imaging using P-Sel MPIOs could predict changes in symptomatology in relapsing-remitting EAE.

First, we assessed the ability of this tool to predict the appearance of a first relapse in asymptomatic animals (Figure 6a and Figure 7). Asymptomatic animals (clinical score 0 + 0) were injected i.v. with P-Sel/MPIOs at day 10 before EAE onset and subjected to molecular imaging (Figure 7A). Animals with little or no hyposignal (figure 7C) showed no relapse the day after MRI (Clinical score 0 + 0; Figures 6a and 7D-E). Conversely, animals showing substantial hyposignal (Figure 7C) declared a first relapse the day after (Clinical score 2.5 ± 0.5; Figures 6a and 7D-E).

Furthermore, histological analysis demonstrated that the presence of P-Sel MRI signal in asymptomatic animals (at Day 10), which predicts the appearance of a first relapse is correlated with early demyelination in both rostral and caudal spinal cord (data not shown)

Second, we assessed the ability of this tool to predict a phase of remission in symptomatic animals (Figures 6b and 8). Symptomatic animals (clinical score 1.875 + 1.057) were injected i.v. with P-Sel/MPIOs at disease peak (day 15) and subjected to molecular imaging (figure 8A). Animals with substantial hyposignal (Figure 8C) showed no remission the day after MRI (Clinical score 2.5 + 0.58; Figures 6b and 8D-E). Conversely, animals showing little or no hyposignal (Figure 8C) experienced a remission the day after (Clinical score 0.5 ± 0.58; Figure 6b and 8D-E).

P-Sel MRO signal is cleared from the tissues within the first 24 hours after injection.

EAE Animals (Clinical score 2) were injected i.v. with P-Sel/MPIOs and subjected to successive molecular imaging sessions of the lower and upper parts of the spinal cord, performed 10 min, 8h and 24h after the injection (Figure 9A). The results showed a progressive decrease of the signal during the first 24 hours after the injection illustrating its clearance from the tissues.

P-Sel MRI signal appears specifically within the white matter of the spinal cord.

Our results showed (data not shown) that the signal obtained from EAE animals is indeed absent in grey matter, which matches the physiopathology of the disease.

Lastly we have shown that P-SE1 is preferentially expressed in small vessels (> 10 μπι of diameter) of mice subjected to MS model (EAE mice) (data not shown) ant a correlation can be established between P-Sel MRI signal and lymphocyte infiltration, in relation to clinical score (see Figure 9).

Conclusion:

The molecular imaging tool developed in this work, targeting the endothelial adhesion molecule P-Sel, allows diagnostic follow-up of animals subjected to chronic or relapsing-remitting EAE models of MS. The use of this tool revealed the spatio-temporal pattern of dissemination of neuroendothelial activation along the spinal cord in these two models. While the chronic model showed a slow progression of neuroendothelial activation in a caudo-rostral direction, the relapsing remitting model showed a much more rapid progression, with neuroendothelial activation appearing in the whole spinal cord from the early stages. We also showed that, in both models, P-Sel imaging, primarily targeting neuroendothelial activation, was representative of associated pathological processes, BSCB alteration and leukocyte infiltration. Again, differential patterns of BSCB alteration and leukocyte infiltration, matching those of endothelial activation, were noticed in the two models. Finally, we validated the use of P-Sel/MPIO as a predictive tool for relapse and remission in MS models. Indeed, the presence of P-Sel/MPIO-derived hyposignal in asymptomatic animals specifically and accurately predicted the appearance of relapse. In addition, the disappearance of P-Sel/MPIO-derived hyposignal in symptomatic animals predicted their remission. Furthermore, we have established that the presence of the P-Sel MRI signal which predicrs the appereance of a first relapse corresponds at the histological level to early demyelination.

This study is the first to show the validity of molecular imaging targeting P-Sel for exploration of MS-like disease in the spinal cord of mice subjected to chronic or relapsing-remitting EAE. It is also the first study to report the predictive value of this tool in the prognosis of relapse and remission in these models of multiple sclerosis.