The present invention relates to multimodal agents that can be used for the diagnosis of diseases related to protein misfolding.

Background art

There is considerable evidence to suggest that misfolding and/or aggregation of proteins, for example misassembly into amyloid and other pathological forms is associated with a number of diseases such as Alzheimer's disease and prion diseases. Alzheimer's disease (AD), and several other age-related neurodegenerative disorders such as Prion diseases and Parkinson's disease are referred to as protein conformation or misfolding disorders. These neurodegenerative diseases show early effects on cognition and memory and affect reasoning and decision making. Significant evidence suggests that the early accumulation of soluble and/or insoluble protein aggregates is central to their pathogenesis.

It is of considerable interest to be able to carry out diagnosis of such diseases by non-invasive means. In particular, whilst imaging using fluorescence requires perforation of the cranial cavity, magnetic resonance imaging (MRI) can be carried out in a fully non-invasive manner. It would be desirable to develop novel agents and methods for diagnostic imaging that will rely on reporter functionalities that are active as molecular reporting agents for enhanced MRI contrast. Such a molecular probe would need design of ligands that are both superparamagnetic and amyloidotropic. Such imaging of amyloid imaging with MRI (amMRI) is currently not available. Today derivatives of amyloidotropic dyes, like Pittsburgh compound B (PIB) (a derivative of Thioflavins) are at the forefront for molecular imaging of amyloid lesions in the living brain. These small molecular agents are radiolabelled to function in a Positron Emission Tomography or PET camera. The molecular imaging field is hampered due to lack of imaging agents for amMRI which would enable correlated imaging of MRI (physiology) fMRI (function) and amMRI (amyloid pathology) all in one diagnostic session.

There are considerable difficulties in obtaining magnetic nanoparticles (MNPs) capable of both crossing the blood-brain barrier, specifically target mis- folded proteins, and contain strong reiaxivity properties so it can bind in sufficient amount to enable reliable detection based on the change in Tl and/or T2 or T2* relaxation of the surrounding water. All these properties must be present in the MNP in order to enable diagnostic imaging. Summary of the Invention

The present invention provides an agent having the structure:

(WSP) _n (MNP)-(ML)-(HCM)-(LL)-(LCP)

wherein:

-WSP is a water soluble polymer

-MNP is a magnetic nanoparticle, a non-magnetic nanoparticle surrounded by magnetic complexes or a magnetic nanoparticle surrounded by magnetic complexes;

-HCM is a hydrosoluble connecting moiety;

-LCP is a luminescent conjugated oligothiophene moiety having 4 to 7 thiophene rings;

-ML is a group enabling binding of the (ML)-(HCM)-(LL)-(LCP) system to the nanoparticle of the MNP moiety;

-LL is a group providing a covalent linkage between the oligothiophene moiety LCP and the hydrophilic connecting moiety HCM,

wherein the oligothiophene moiety shows at least one carboxylic acid group in a free acid or salt form,

wherein n is at most 100.

Water soluble polymers may not be necessary for carrying out the present invention successfully, particularly if a highly hydrosoluble polymer spacer is used. Therefore, in certain embodiments, n in the above formula may be zero. However, in preferred embodiments, n is greater than zero, and indeed in preferred embodiments, a molar excess of WSP may be used, advantageously as much as 99 mol% of WSP to LCP. Thus, in advantageous embodiments, n is at least 0.5 and at most 99.5. More preferably, n is at least 1, even more preferably at least 10, still more preferably at least 50, and in preferred embodiments n is at most 99.

In the agents of the invention, the oligothiophene moiety LCP preferably contains one or more of the following units: thiophene, 3-thiophene acetic acid, 2- thiophene carboxylic acid, triazole-linked thiophene.

In the agents of the invention, the -(ML)-(HCM)-(LL)-(LCP) part of the agent preferably has a number of chain atoms counted between extremities ML and LL of between 8 and 150 atoms.

In the agents of the invention, the hydrosoluble connecting moiety (HCM) preferably contains a polymer or oligomer chain. In generally advantageous embodiments, the hydrosoluble connecting moiety (HCM) contains a polyoxyalkylene polymer or oligomer, a polysaccharide or oligosaccharide, a poly(meth)acrylate or oligo(meth)acrylate.

Appropriate magnetic complexes of the MNP moiety can cause change in Tl relaxation and/or T2 relaxation and/or T2* relaxation of the surrounding water. In this context, the above-defined agent of structure (WSP)-(MNP)-(ML)-(HCM)-(LL)- (LCP) can be used in magnetic imaging. In particular, the agent according to the invention can be used as an imaging agent for the diagnosis of diseases related to misfolding and/or aggregation of proteins. Such diseases may for example be Alzheimer's disease or a prion disease or other diseases associated with protein aggregation and/or misfolding, such as systemic amyloidoses.

It has surprisingly been found in the framework of the present invention that for the magnetic conjugates of invention, although the magnetic moiety MNP is typically considerably larger than the oligothiophene (LCP) moiety, the entire conjugate is able to pass the blood-brain barrier. The hydrosoluble linker used to connect the magnetic moiety MNP and the oligothiophene (LCP) is believed to participate in the maintenance of magnetic properties, by allowing access of water molecules. In a preferred embodiment, complexation of magnetic ions using (poly)phosphonates gives rise to favourable magnetic properties in Tl and/or T2 relaxivity. The LCP oligothiophenes appear to enable binding to amyloid in the large conjugates.

Brief Description of the Figures

Figure 1: TEM (top) and HRTEM (bottom) images of naked GdP0 ₄ nanoparticles of example 1 (length 100 nm, diameter 20 nm).

Figure 2: Schematic structure of biphosphonate-PEG (BP-11-104) on particle of Synthesis Example 1.

Figure 3: FT-IR spectrum (bottom) of GdPGvbisphosphonate-PEG of Synthesis Example 1 showing the typical PEG bands.

Figure 4: TEM images of naked GdF ₃ nanoparticles of Synthesis Example 2

(diameter around 15 nm).

Figure 5 : X-ray diffraction patterns of GdF ₃ -PEG of Synthesis Example 2. Figure 6: Ti0 ₂ - bisphosphonate- PEG-DOTA structure with complexation of gadolinium ions as obtained in Synthesis Example 3.

Figure 7: Ti0 ₂ -bisphosphonate-PEG-DOTA-Gd of Synthesis Example 3 in ex vivo brain imaging, reading left-to-right: A) In a Tl weighted spin echo image ; B) in a Tl-weighted gradient-echo image ; C) in a T2-weighted gradient echo image ; and D) in a T2-weighted spin echo images.

Figure 8: Selectivity of GdF ₃ -bisphosphonate-PEG-LCP-a (short spacer) on Αβ fibrils (Fluorescence imaging).

Figure 9: Selectivity of GdF ₃ -bisphosphonate-PEG-LCP-a (short spacer) on Αβ fibrils (TEM imaging).

Figure 10 : Selectivity of Ti0 ₂ -bispnosphonate-PEG-LCP-a (short spacer) on Αβ fibrils (TEM imaging). Figure 11 : staining of GdF ₃ -bisphosphonate-PEG-LCP (short and longer spacer) on Αβ fibrils.

Figure 12: Luminescent spectroscopic properties of multifunctionalized GdF ₃ -PEG-LCP. QE: quantum efficiency. MNP a-LCP: GdF ₃ -PEG-LCP short spacer; MNP b-LCP: GdF ₃ -PEG-LCP longer spacer; a-LCP: LCP short spacer; b-LCP : LCP longer spacer; pFTAA: a pentathiophene reference.

Figure 13: Fluorescence micrographs to show the distributions of LCP-MNP 5011 following two injections of probe into the tail vein of an APP/PS1 mouse. Identical fluorescence spectra are visible for all amyloid plaques in the brain, and a stronger probe distribution is found near the ventricle.

Figure 14: Fluorescence micrographs and fluorescence intensities show the distributions of LCP-MNP 5011 following two injections of probe into the tail vein of an APP/PS1 mouse. Strong probe distribution is found near the ventricles and near arteries in the brain.

Figure 15: Box-plot showing the Gd concentration in brain tissue (mM) measured by ICP-MS seven days after intracranial injection. WT mice are marked in blue and APPPS1 mice in green. Animals are separated based on the perceived success of intracranial injection. Detailed description of the Invention

The present invention relates to multimodal imaging agents for the in vivo diagnosis of protein misfolding diseases. The agents normally comprise all of the following components; MNP, WSP, ML, HCM, LL, LCP. In some embodiments however, as mentioned above, it may be possible to carry out the invention without WSP units. These molecular entities will be specified in detail in the following sections.

Figure 11 shows non-limiting illustrative embodiments of agents according to the present invention (without the water-soluble polymer moiety WSP), these embodiments having a gadolinium fluoride-based magnetic nanoparticle MNP, a phosphonate group-based ML system allowing binding of the magnetic nanoparticle MNP, a hydrosoluble connecting moiety HCM in the form of a polyethylene oxide or poly (2-hydroxyethyl-methacrylate) oligomer/polymer, and an amide or amide-triazole group LL to ensure covalent linkage with the oligothiophene moiety LCP. Although these specific embodiments help to show the general structure of agents according to the invention, numerous variants within the scope of the present invention can be envisaged and be expected to provide similar results, in particular in imaging applications.

Concerning the magnetic nanoparticle component MNP, typically the MNP contains metals, metal oxides, metal fluorides or metal phosphates. The MNP may be advantageously chosen from the group consisting of: gadolinium fluorides, gadolinium oxides, iron oxides, manganese oxide, manganese sulfide, gadolinium phosphate. The MNP moiety may also be metal clusters (iron, manganese, gadolinium), and another possibility for the MNP moiety is a metal oxide surrounded by magnetic complexes (gadolinium, iron, manganese). General MNP moieties may thus be a magnetic nanoparticle based on GdF ₃ , GdP0 _4/ MnO, Fe ₃ 0 ₄ , Gd ₂ 0 _3/ manganese sulfide or manganite, or a magnetic cluster, or a nanoparticle based on Ti0 ₂ , Si0 ₂ or Au, to which a magnetic complex is bound.

The magnetic nanoparticle of the invention may be a nanorod, a spherical nanoparticle or have a core-shell structure. The largest diameter of the nanoparticle may advantageously range from 2 to 200 nm.

In order to join the organic moiety -(HCM)-(LL)-(LCP), containing the hydrosoluble connecting moiety HCM and the conjugated oligothiophene LCP, to the magnetic nanoparticle MNP, a group ML is used. Preferred ML groups are ones showing chelating ability for the metal atoms of the MNP group. Generally advantageous ML groups include groups containing a moiety selected from the group consisting of phosphonate, bisphosphonate, polyphosphonate, phosphinate, phosphate, thiol, alkoxysilane, silane, silanol, amino, thiophosphonate, thiophosphate and carboxylate functions. The ML group may appropriately contain a chain length, counting from the hydrosoluble connecting moiety HCM and in particular the closest atom of a repeating unit of an oligomeric or polymeric HCM to the ML group, of less than 20 atoms, preferably less than 10 atoms.

In the present invention, the magnetic nanoparticle MNP part of the agent of the invention is, in preferred embodiments, also associated with a water-soluble polymer (WSP) able to bind to the MNP. The water-soluble polymer will typically be a biocompatible material, and will have functional groups with heteroatoms (such as oxygen, nitrogen, or sulphur) able to coordinate to metal atoms in the MNP. Preferred examples of WSP are polyoxyalkylenes such as polyethylene glycol (PEG) chains, oilgo- or polysaccharides such as dextran chains, or hydrophilic (meth)acrylate polymers such as oligomers or polymers of N-acryloylmorpholine (PNAM), p(hydroxyethyl)acrylate (PHEA). The number of repeating polymer units may advantageously lie between 3 and 20 e.g. between 3 and 20 alkylene oxide, monosaccharide or C=C-derived e.g. (meth)acrylate units.

In the organic moiety -(ML)-(HCM)-(LL)-(LCP), the hydrosoluble connecting moiety HCM may appropriately contain the same sorts of polymer chain as the water-soluble polymers WSP that are directly associated with the magnetic nanoparticle MNP. The number of repeating polymer units may advantageously lie between 3 and 100 e.g. between 3 and 100 alkylene oxide, monosaccharide or C=C-derived e.g. (meth)acrylate units.

In what follows, exemplary but non-limiting synthetic routes enabling (ML)-

(HCM)-(LL)-(LCP) units to be prepared will be presented.

The aggregation of proteins is a major hallmark of several neurological diseases, such as Alzheimer's disease (the A-beta and tau-protein). Oligo- and poly-thiophenes of four units or more have previously been shown to specifically target and stain protein aggregates in vitro [Klingstedt et al. ORGANIC & BIOMOLECULAR CHEMISTRY, 9, 24, 8356-8370] and in for some specific molecules in vivo [Aslund et al. ACS CHEMICAL BIOLOGY, 4, 8, 673-684].

In this context the oligothiophene acts as a targeting device that accumulates magnetic resonance imaging contrast agents (MRI-CA) at sites of protein aggregates. The oligothiophene can be between four and seven thiophenes long and will also assist the MRI-CA in its passage across the blood brain barrier.

One exemplary and non-limiting synthesis is shown by following Scheme 1, in which the ML group is phosphonate-based, the HCM moiety is based on a hydrophilic acrylate derivative (2-hydroxyethyl acrylate), polymerized by atom transfer radical polymerization, and the (ML)-(HCM) unit is coupled to the oligothiophene (LCP) via an alkynyl (-C≡C-) bond on the latter reacting with an azide group introduced into the (ML)-(HCM) unit to form a 1,2,3-triazole ring, the final LL unit thus containing the 1,2,3-triazole ring.

Another exemplary and non-limiting synthesis is shown by following

Scheme 2, wherein a polyethylene glycol derivative with a terminal free amine group is linked directly to a thiophene ring carboxylic acid substituent to form an amide (such that HCM is PEG-based and the linking group LL contains an amide), and then the other end of the PEG is converted into a phosphonate system.

In the first of these two exemplary coupling chemistries (Scheme 1), the LL linking group is constituted (starting at the 2-position of a terminal thiophene of the LCP moiety) by -CO-NH-CH ₂ -(l,2,3-triazole-4-yl)-, the triazole system being directly bound to the repeating unit, here a 2-hydroxyethyl acrylate-derived repeating unit of the HCM. In the second exemplary coupling chemistry (Scheme 2), the LL linking group is a -CO-NH- group. The type of linkage of the LL group is not particularly limited. The LL group may appropriately contain a chain length, counting from the hydrosoluble connecting moiety HCM and in particular the closest atom of a repeating unit of an oligomeric or polymeric HCM to the LCP group, of less than 20 atoms, preferably less than 10 atoms. In view of the availability of carboxylic acid group-terminated thiophenes, in a preferred embodiment, the LL group contains at least an amide (-CO-NH-) functional group.

Scheme 1 : Example synthetic route for an ML-HCM-LL-LCP species

Scheme 2 : Example synthetic route for an ML-HCM-LL-LCP species The oligothiophenes with a a free -C0 ₂ H group on a terminal thiophene can be obtained as previously described and used e.g. in the synthesis route outlined in Scheme 2 above. Either end of the oligothiophene can be provided with an alkynyl group for coupling as in Scheme 1 above by the following processes:

Scheme 3: Preparation of alkynyl group-terminated oligothiophenes

Based on known syntheses of oligothiophenes and the above chemistry for producing terminal alkynyi groups, a number of LCP units can be prepared, such as:

In a further aspect, the present invention provides compounds having the structure (ML)-(HCM)-(LL)-(LCP), wherein ML, HCM, LL and LCP are as defined in claim 1. These compounds may serve as intermediates before their linkage to magnetic nanoparticles or like species, but it may be envisaged that the chemistry developed in the present application could enable these species to serve as useful means to attach luminescent oligothiophene moieties to other chemical units, for example for other types of imaging application.

Preferred compounds of the present invention under this aspect include the following:

In a preferred embodiment of the present invention, the nanoparticle in the MNP moiety is surrounded by water-soluble polymers (WSPs), some of which, or indeed the majority of which, may not have a luminescent conjugated oligothiophene (LCP) moiety. In an illustrative and non-limiting example, the nanoparticle may be coordinated by a mixture of (poly)phosphonate-terminated WSPs (such as (poly)phosphonate-terminated polyethylene glycol(s) (PEG)) and (poly)phosphonate-terminated agents according to the invention containing an LCP moiety. The saturation of the surface of the nanoparticle avoids inter alia coordination of carboxylic acid groups of the oligothiophene (LCP) moiety with the nanoparticle (which could inhibit specific binding to misfolded proteins). In agents according to the invention, the value n, representing the average number of moles of WSP with respect to -(ML)-(HCM)-(LL)-(LCP) units associated with magnetic nanoparticles MNP is preferably between 1 and 99. Thus the molar ratio is typically between 1 and 99 molar percent of LCP-containing -(ML)-(HCM)-(LL)- (LCP) units to 99 to 1 percent of WSP units, such as (poly)phosphonate- terminated polyethylene glycol(s) (PEG)).

In order to prepare agents of the invention of formula (WSP) _n (MNP)-(ML)-

(HCM)-(LL)-(LCP), different approach routes are possible.

In one generally applicable process, ML-HCM-LL-LCP species are mixed with the water-soluble polymers WSP in the desired ratio in a solvent. In a generally preferable case, the solvent is water. Alternatively, any other solvent can be used which is identical or compatible with the solvent in which particles to be further functionalized are dispersed.

The ML-HCM-LL-LCP/WSP solution should be as concentrated possible to optimize the coupling, preferably about 0.1 mol/L.

A nanoparticle colloidal solution (2 to 0.2 mol/L), preferably in water, is added under stirring. The mixture is left stirred for 1 to 24 hours between room temperature and 150 ° C, preferably 70°C.

Purification (to remove free ML-HCM-LL-LCPs and free WSPs) is carried out by dialysis or precipitation and further washing.

In a different process approach, ML-HCM-LL-LCPs and WSP can be introduced separately through sequential addition.

The reactive terminal functions of ML-HCM-LL-LCPs and WSP react with the metal ions on the particle surfaces.

When using magnetic ion complexes as magnetic entities, the complexation reaction can take place either before or after the introduction of the ML-HCM-LL- LCPs. If the complexation occurs before the ML-HCM-LL-LCP grafting, the system must be purified to remove free magnetic ions, using for example dialysis to avoid the complexation of magnetic ions by the carboxylic function of ML-HCM-LL-LCPs. Another alternative is to use LCPs bearing protective groups on the carboxylic functions (ex. alkyl ester groups).

Any combination of embodiments separately presented hereinabove as preferable, advantageous or generally applicable in the context of the invention is itself, as a combination of features, to be construed as within the teaching of the invention, unless the features concerned are said to be mutually exclusive or are clearly understood by one skilled in the art to be mutually exclusive in context.

The full context of all documents referred to in the present application is incorporated herein by reference.

The following worked examples illustrate the functioning of the present invention but are not to be considered as limiting the scope of the invention. Examples

PHYSICAL MEASUREMENTS AND INSTRUMENTATION

- Infrared spectra in the range 500-4000cm ^_1 were recorded on a Perkin Elmer spectrophotometer (spectrum 65 FT-IR with Universal ATR Sampling Accessory).

- XRD analysis was performed with an X-ray powder diffractometer (Model X'Pert PRO MPD) in Bragg Brentano geometry (Θ/Θ) with Cu-K _a i _a 2 or Co-K _ala2 radiation and combined with a fast detector based on real time multiple strip technology (X'Celerator). Phase identification was carried out using the EVA software (Bruker) and the JCPDS-ICDD PDF-2 database.

- Dynamic Light Scattering (DLS) experiments and Zeta potential measurements were carried out using a MALVERN zetasizer nano ZS.

- Transmission Electron Microscopy (TEM) was performed on a JEOL 2010F transmission electron microscope operating at 200 kV.

- Samples were prepared by evaporating a drop of the particle dispersions in water on a 5θΑ thick carbon-coated cooper grid (300 mesh). - Relaxivity measurements at 7T magnetic field strength were performed at a Bruker Biospec horizontal bore magnet (Bruker Biospin, Ettlingen, Germany).

Samples were diluted in saline into five different concentrations and filled in 2 mL tubes. These diluted concentrations were used for relaxivity measurements at all field strengths.

The prepared tubes (described above) with 5 different concentrations of the same sample were placed in a dedicated holder (Figure 14A) together with a reference tube filled with saline. The holder were filled with distilled water and placed in icocenter of the magnet. Tl values were measured by using a series of spin echo images with varying repetition times (RAREVTR pulse sequence; echo time (TE)=9 ms, repetition time (TR)= 15, 20, 50, 100, 200, 400, 1000, 2000, 5000, 20000 ms, RARE factor 2, field of view (FOV) 65 x 50 mm, 128 x 128 matrix, slice thickness 3 mm. T2 values were measured by using a multi spin multi echo sequence (MESME, TR= 10000 ms, TE= 11, 22, 33, 44, 55, 66, 77, 88, 99, 110, 121, 132, 143, 154, 165, 176 ms) with identical geometry as the Tl measurements. Tl and T2 values were calculated from the signal intensities in the tubes using ATLAB. A map of the calculated Tl values of tubes filled with different five different concentrations of Ti02-bisphosphonate-PEG-DOTA-Gd and one tube with saline are shown in Figure 14B.

The relaxivity values were calculated by a linear fit of the relaxation rates

(inverse Tl and inverse T2) as function of concentrations using Excel (Office 2003, Microsoft, USA). Such a linear fit is illustrated in Figure 14C. Both per paramagnetic/super paramagnetic ion ([mM ^' V ¹ ] and per particle relaxivities rates ([uM ^' V ¹ ]) were calculated.

Ex vivo magnetic resonance imaging was performed on a Bruker Biospec horizontal bore magnet (Bruker Biospin, Ettlingen, Germany), on mouse brain after direct injection of 2 μΙ_ 0.2225 mM TiO ₂ -bisphosphonate-PEG-DOTA-Gd. Synthesis Example 1: Synthesis of gadolinium phosphate rod particles (GdP0 ₄ ) and functionalization with Bisphosphonate-PEG (GdP0 ₄ -PEG)

Gadolinium phosphate nanoparticles were prepared according to the following process. A solution of 1.49 g GdCI ₃ , 6 H ₂ 0 (4 mmol) in 10 mL water is slowly added under stirring to a solution of 0.55 g (4 mmol) NaH ₂ P0 ₄ in 10 mL water previously heated at 70 °C. The solution is stirred at 70 °C for 5 H. The nanoparticles are purified through precipitation in acetone and dispersed in water. A stable colloidal suspension is obtained. The prepared particles are nanorods (GdP0 ₄ ). The nanoparticles were characterized using Transmission Electron Microscopy (TEM) (the TEM images are shown in Fig.l) and Dynamic Light Scattering (DLS). The TEM combined with DLS shows that the particles present a good monodispersity with size around 100 nm length and 20 nm diameter.

These nanoparticles were functionalized with Bisphosphonate-PEG ligands as schematically illustrated on Fig.2, in order to stabilize them in aqueous media (GdP0 ₄ -PEG). Typically, 1 g of Bisphosphonate-PEG (2 mmol, Mw = 500, Ci ₅ H ₃₄ 0i ₄ P ₂ , purchased from Surfactis Technologies, France) is dissolved in 10 mL water and filtered (0.45 m nylon filter). 5 mL of aqueous solution of nanoparticles (gadolinium phosphate nanoparticles previously prepared, 4 mmol) is added to the biphosphonate under vigorous stirring. The suspension is sonicated and becomes transparent. The mixture is heated at 70°C for 1 H, and the particles are purified through dialysis.

The surface modification and the presence of biphosphonate-PEG groups are evidenced using FT-IR as shown on Fig.3 and quantified using ICP. The GdP0 ₄ crystalline structure is confirmed by the powder X ray diffraction patterns.

Relaxivities are respectively: 2 and 11 mM ( _GCI ^S ^"1 for ri and r ₂ at 7T

211 and 1158 μΜ particles) ^" ^ ^"1 for ri and r ₂ at 7T Synthesis Example 2: Synthesis of spherical gadolinium fluoride particles (GdF ₃ ) and functionalization with Bisphosphonate-PEG (GdF ₃ - PEG)

Gadolinium fluoride nanoparticles were prepared according to the following process. 0.45 g of hydrofluoric acid (9 mmol HF) is mixed with 16 mL of N,N- dimethylformamide. A solution of 1.49 g (4 mmol) GdCI ₃ , 6 H ₂ 0 in 3 mL methanol is slowly added under stirring. A transparent solution is obtained. This solution is heated at 170 °C for 16 hours. A stable colloidal suspension is obtained. The nanoparticles are purified through precipitation in acetone and then dispersed in water. The prepared particles are spherical with a diameter of 20 nm (GdF ₃ ). The nanoparticles were characterized using Transmission Electron Microscopy (the TEM images are shown on Fig.4) and Dynamic Light Scattering (DLS). The TEM combined with DLS shows that the particles present a good monodispersity with a size around 20 nm.

These nanoparticles were functionalized with Bisphosphonate-PEG ligands

(BP-11-104) in order to stabilize them in aqueous media (GdF ₃ -PEG). Typically, 1 g of Bisphosphonate-PEG (2 mmol) is dissolved in 10 mL water and filtrated (0.45 pm nylon filter). 5 mL of aqueous solution of nanoparticle GdF ₃ (4 mmol) is added to the biphosphonate under vigorous stirring. The suspension is sonicated and becomes transparent. The mixture is heated at 70 °C for 1 H, and the particles are purified through dialysis.

The surface modification and the presence of biphosphonate-PEG groups are evidenced using FT-IR and quantified using ICP. The GdF ₃ crystalline structure is confirmed by the powder X ray diffraction patterns shown on Fig.5.

Relaxivities are respectively: 0.55 and 55 mM (G _C I^S ^"1 for ri and r ₂ at 7T

190 and 19200 μΜ (particles) ^"1 ^ ¹ for n and r ₂ at 7T Synthesis Example 3 : Synthesis of spherical titanium oxide particles Ti0 ₂ - bisphosphonate-PEG-DOTA-Gd

- Synthesis of titanium oxide particles (T1O2)

Titanium isopropoxide (0.11 mol) and isopropanoi (0.20 mol) were first collected under inert atmosphere (argon) and introduced in a funnel tightly capped. This mixture is then introduced dropwise into a large quantity of ultrapure water (13.9 mol) with vigorous stirring. It formed in the middle a white precipitate. After 20 min of stirring, a few milliliters of concentrated nitric acid 69 % (0.022 mol) are introduced. The solution is then distilled to remove all isopropanoi. After 24 hours at reflux, an aqueous solution of titanium dioxide nanoparticles is obtained.

- Functionalization of titanium oxide particles with bisphosphonate-PEG-NH ₂ The structure of bisphosphonate-PEG-NH ₂ is presented hereafter :

In 10 mL aqueous solution of 0.80 g of bisphosphonate-PEG-NH ₂ (2 mmol, MW = 353; C8H21NO10P2, purchased from Surfactis Technologies, France), 10 mL of a Ti0 ₂ suspension (4 % (4 mmol)) is added under stirring. The suspension is sonicated and stirred at 70 °C for 1 h. The nanoparticles are purified through dialysis and dispersed in 12 mL DMF. - Clicking of DOTA-NHS on Ti0 ₂ - bisphosphonate- PEG-NH ₂ particles

The structure of DOTA NHS (DOTA-N-hydroxysuccinimide) is presented hereafter:

4 ml (6.7 mmol) of colloidal suspension of PEG-Ti0 ₂ in DMF is prepared. To this solution is added 0.16 mL (1.3 mmol) of triethylamine and 3.5 mL (1.3 mmol) of a DOTA-NHS solution in DMF. The suspension is stirred at room temperature overnight. The nanoparticles are purified through dialysis and dispersed in 3 ml water.

- Complexation of gadolinium ions on T1O2- bisphosphonate- PEG-DOTA

with Gadolinium ions

A solution of 0.494g (1.33 mmol) GdCI ₃ , 6 H ₂ O in 1 ml water is added under stirring to the obtained suspension of TiO ₂ -bisphosphonate-PEG-NH ₂ . 5 drops of NaOH 1M are added to the suspension. The mixture is stirred overnight. The nanoparticles are purified through dialysis and dispersed in water.

The functionalization of the surface is confirmed by FT-IR and quantified using ICP. Results of ICP give a ratio of 2 Gd for 1 DOTA-NHS. Some Gd ions are complexed by the phosphonates at the surface of the nanoparticles. The structure of the T ^' iO ₂ -bisphosphonate-PEG-NH2 NPs is schematically represented on Fig.6. Using phosphonate or polyphosphonate functions allow complexation of magnetic cation with phosphorus atoms near the particle surface, exalting the magnetic properties.

Relaxivities are at 7T.

Fig.7 illustrates how this contrast agent with high ri and moderate r ₂ relaxivities are visible in MR images as both positive and negative contrast depending on the imaging sequence used for acquisition. Images were obtained in an ex vivo mouse brain after direct injection of 2 pL 0.2225 mM Ti0 ₂ - bisphosphonate-PEG-DOTA- Gd. A) In a Tl weighted spin echo image the MNPs appear as positive contrast while the same MNPs appears as negative contrast in a Tl-weighted gradient-echo image (B) due to T2* effects. Due to the high r ₂ the signal are negative in both T2-weighted gradient echo (C) and T2-weighted spin echo (D) images.

This experiment, results of which are shown in Figure 7, shows that the magnetic nanoparticles, or non-magnetic nanoparticles surrounded by magnetic complexes, used in the invention are promising for use in magnetic imaging. The species used here do not contain an oligothophene (LCP) moiety.

Synthesis Example 4: Complexation of Gd ⁺³ ions on the surface of bisphosphonate-PEGylted gadolinium phosphate particles (GdP0 ₄ -PEG) of example 1. Improvement of relaxivity properties

GdP0 ₄ -PEG powder comes from the freeze-drying of the particles of example 1. 300 mg (1 mmol) of GdP0 ₄ -PEG containing 0.05 mmol of bisphosphonate-PEG is dispersed in 10 mL of water. 0.5 mL (0.15 mmol) of Gd ³⁺ aqueous solution (GdCI ₃ , 0.3 M) is added. The suspension is stirred at room temperature overnight. The nanoparticles are purified through dialysis and dispersed in water. Gadolinium ions are quantified using ICP and 3.22 pmol Gd per mg NPs sample versus 2.92 pmol Gd per mg NPs sample before Gd ³⁺ adding are obtained. Some Gd ions are complexed by the phosphonates at the surface of the nanoparticles, with the same structure shown in Fig.6 for Ti0 ₂ particles.

Relaxivities are respectively: 3 and 19 mM (cd^S ^"1 for ri and r ₂ at 7T

310 and 2126 μΜ (particles) ^"1 ^ ¹ for n and r ₂ at 7T There is an increasing of per particle relaxivities, around 50 % for ri and 100 % for r ₂ .

Example 1 : Coupling of oligothiophenes (LCPs) and bisphosphonate- PEG to GdF ₃ particles

Preparation of ML-HCM-LL-LCP species

The following worked syntheses illustrate synthesis methods described more generally in Schemes 2, 3 and 1 hereinabove.

LCP-2

To a solution of LCP-1 (1 mmol) and DMF (0.15 mmol) in toluene (5 mL) thionylchloride (1.3 mmol) was added. The reaction was refluxed for 1 h followed by concentration to dryness. The acid chloride and diisopropylethylamine was dissolved in DMF and the amine (1.1 mmol) was added to the solution. After 2 hours the reaction mixture was diluted with toluene and washed with HCI (1M, aq) and brine. The crude product was purified by flash chromatography (toluene/ethyl acetate 4:1 to 2:1 to 1:1) to give LCP-2 in 92% yield.

LCP-3

LCP-2 (1 mmol) was dissolved in methylene chloride/triflouroacetic acid (TFA, 4:1, 5 mL). After 1 h the solution was concentrated in vacuo and dissolved in DMF. Diisopropylamine was added until basic whereupon diethylphosphonoacetic acid (1.1 mmol) and HATU (1.1 mmol) were added. After 4 h the reaction mixture was concentrated and purified by flash chromatography methylene chloride/methanol (30:1) to give product LCP-3 in 74 % yield.

LCP-4 (LCP a) LCP-3 (1 mmol) was dissolved and dioxane (5 ml_) and NaOH (1 M, aq., 3 mmol) was added. After 2 h the mixture were evaporated in vacuo. The product was dissolved in acetonitrile and trimethylsilylbromide (3 mmol) was added. When no starting material was seen on HPLC the solvents were evaporated and the product was dialyzed in water with a cutoff of 500 D to give LCP 4 in 96% yield.

LCP-5

LCP-1 (1 mmol) was dissolved in DMF (5 ml_). Tert-Butyl 2,2,2-trichloroacetimidate (2 mmol) and BF3*Et ₂ 0 (0.30 mmol) were added. After 30 min the reaction was concentrated in vacuo and purified on reversed phase chromatography (acetonitrile/water 4:1 to 9:1) to give the product quantitatively.

LCP-6

LCP-5 (1 mmol) was dissolved in DMF (5 mL) and cooled to -15 °C. N- bromosuccinimide (1 mmol) was added slowly and the reaction mixture was allowed to reach room temperature. After 2 h the solution was diluted with toluene and washed with brine to give LCP-6 quantitatively.

LCP-7

LCP-6 (1 mmol) K ₂ C0 ₃ (6 mmol), 5-Carboxythiophene-2-boronic acid (3 mmol) and PEPPSI-IPr (0.05 mmol) were dissolved in 5 mL toluene/methanol (1:1) and heated to 80 °C. After 20 minutes the reaction mixture were diluted with acetic acid and ethyl acetate and washed with brine several times to give LCP-7 in 67 % yield.

LCP-8

LCP-7 (1 mmol), diisopropylethylamine (3 mmol), propargylamine (2 mmol) and HATU (1.2 mmol) were dissolved in DMF. After two hours the reaction mixture were diluted with chloroform and washed with brine 3 times. The product was purified by flash chromatography using toluene/ethyl acetate/chloroform (80:7:7) as the eluent to give the protected product in 66% yield. The protected product was dissolved in chloroform/TFA (4:1). After 1 h the solvents were removed in vacuo and the product was dissolved in water and NaOH (3 mmol). After 3 h the solution was lyophilized to give LCP-8 in 66% yield over three steps.

1. Synthesis of In-Br.

To 1.7 ml of bromoethanol (24 mmol) and 3.5 mL of triethylamine (24 mmol) in lOOmL dichloromethane in ice bath, slowly added bromo-isobutyryl bromide (3.5 mL, 29 mmol). The solution was allowed to warm to room temperature and kept under agitation overnight. The reaction was stopped after 18 h and the solution was washed with 1M hydrochloric acid several times and one time with distilled water. Dried over Na ₂ SO ₄ and solvent was removed by evaporation. Clear oil consistent with In-Br by NMR ( Η and ¹³ C) was used further without any purification (6.04 g, 92 %).

2. Synthesis of Initiator (In-P03Et ₂ ).

Compound In-Br and triethylphosphate (9.5 mL, 55 mmol) were heated at 150 °C in a round bottom flask equipped with condenser under inert atmosphere. After 16h the reaction was stopped and excess triethylphosphate was evaporated under high vacuum. The residue was purified by column chromatography using silica gel eluting with dichloromethane/ethyl acetate mixture (1:1) to give light yellow oil of In-P0 ₃ Et ₂ (3.98 g, 54 %).

3. Polymerization of Hydroxyethylacrylate (HEA) by In-P0 ₃ Et ₂ (Br-p(HEA) _n - P0 ₃ Et ₂ )

In a typical example, initiator (In-P0 ₃ Et ₂ ) (0.71 g, 2.15 mmol) and HEA (12.5 g, 107.5 mmol) and bpy ligand (0.67 g, 4.3 mmol) were mixed and degassed for 10 minutes. CuBr catalyst (0.31 g, 2.15 mmol) were added to this degassed solution, and stirred at 85 °C, affording a dark brown solution. Polymerization occurred immediately, leading to an increase in viscosity of the solution. The polymerization was quenched in liq. N ₂ after 11 min, the solution was diluted with deionized water and then transferred to dialysis membrane (spectra/por membrane, MWCO=1000) for dialysis. The outer phase was replaced at 6-h intervals with fresh water during dialysis (2 days). The polymer solution was filtered and lyophilized to yield Br- p(HEA) _n -P0 ₃ Et ₂ . 4. Deprotection reaction of phosphonate ester (Br-p(HEA) _n -P0 ₃ H ₂ ).

In a typical reaction, a solution of 0.70 g of polymer (M _n = 3400) in anhydrous DMF (lOmL) at 0 °C, added excess amount of trimethylsilyl bromide (2.4 ml_, 18.0 mmol) drop-wise and the solution was slowly warmed to room temperature. After 6 h the mixture was kept in ice and bath and methanol (10 ml_) was added to it. After 10 min of stirring, the solution was transferred to a dialysis tube and dialyzed against deionized water (same way as before) for 2 days to remove the organic part. The solution was filtered and lyophilized to obtain the polymer. 5. Synthesis of azide terminated polymer (N ₃ -p(HEA)n-P0 ₃ H2)

In a lOOmL flask, 0.40 g of the polymer was dissolved in DMF (5mL) and NaN ₃ (0.04 g, 0.61 mmol) was added to it. The mixture was kept stirring for 24 h at 50 °C. After cooling to room temperature, the mixture diluted with water and dialysis was performed (same way as before) followed by lyophilization to yield the compound.

6. Synthesis of q-FTAA-p(HEA)n-P0 ₃ H ₂ . 0.2 g of polymer (N ₃ -p(HEA)n-P0 ₃ H2) and q-FTAA (0.053 g, 0.067 mmol) were dissolved in a 10 mL mixture of water/ dioxane (1:1, v/v) containing in flask fitted with a rubber septum. The solution was degassed by bubbling argon, and then CuS0 ₄ (0.033 g, 0.134 mmol) and ascorbic acid (0.35 g, 2.0 mmol) were added. The mixture was stirred at 70 °C for 48 h under argon atmosphere. Then the mixture was cooled to room temperature and transferred to dialysis tube (MWCO=2000). Dialysis was performed for 3 days against water to completely remove the excess q-FTAA molecule, Cu and the organic part. The solution was then filtered through a celite plug and lyophilized to get the q-FTAA-p(HEA) _n - P0 ₃ H ₂ .

Coupling method

Samples were prepared using 0.5 mmol of coupling molecules (95 % bisphosphonate-PEG and 5 % ML-HCM-LL-LCPs as a mole percentage for 1 mmol of GdF ₃ particles as obtained at the beginning of Synthesis Example 2, which corresponds to excess relative to the amount needed for a full surface coverage on the particles. After modification, the anchored particles were dialyzed to remove unreacted coupling molecules.

In a typical experiment, bisphosphonate-PEG (0.475 mmol, Mw = 500, C15H34O14P purchased from Surfactis Technologies, France, with 7 units of ethylene oxide) is dissolved in 5 mL of water and filtered (0.45 pm nylon filter) and mixed with 0.025 mmol of LCPs in the form of ML-HCM-LL-LCP units in 0.5 mL of water. The mixture is heated to 70 °C. A suspension of 1 mmol of GdF ₃ particles in 2 mL of water is added to this mixture, under vigorous stirring, and is heated at 70 °C for 1 H in daylight. The resulting suspension is purified through dialysis. Relaxivity measurement

Grafting is performed on GdF ₃ with two different LCPs, the first one with 95 % bisphosphonate-PEG and 5 % short spacer (LCP-a) and the second with 98 % bisphosphonate-PEG and 2 % longer spacer (LCP-b) represented hereafter.

Results show almost no change in ri (0.55 mM^S ^"1 , at 7T) while a significant increase in r ₂ using b-LCP (55 to 158 mM^S ^"1 , at 7T) was observed compared to naked or PEGylated particles without LCPs. The increased r ₂ values are probably due to the increased particle size inducing a reduced particle tumbling rate.

Example 2: Staining and selectivity of GdF ₃ -bisphosphonate-PEG-LCP to amyloid-beta fibrils

An in vitro test was performed with GdF ₃ -bisphosphonate-PEG without LCP (control) and GdF ₃ -bisphosphonate-PEG-LCP (a and b, respectively with short and long spacer) on cells positive and negative. As shown in Fig. 8 in the case of LCP-a (short spacer), only association of GdF ₃ -bisphosphonate-PEG-LCP-a and the positive cells gives a fluorescence signal, characteristic of the binding of LCPs-a to fibrils.

Fig. 9 shows TEM images of the association of GdF ₃ -bisphosphonate-PEG alone with fibrils on the one hand, and GdF ₃ -bisphosphonate-PEG-LCP with fibrils on the other hand in the case of LCP-a (short spacer). This shows a good selectivity and affinity to in vitro formed Αβ fibrils.

Amyloid plaques are stained by GdF ₃ -bisphosphonate-PEG-LCP (short and longer spacer) as well as with LCPs alone. LCPs and GdF ₃ -bisphosphonate-PEG-LCP bind in a comparable fashion to the plaques as shown on Fig. 9 when LCPs-a (short spacer) are used.

Example 3 : Coupling of LCPs to Ti0 ₂ -bisphosphonate-PEG-DOTA-Gd

There are many ways to fit together different parts of this particle, and one of these ways is presented hereafter.

a) Synthesis of titanium oxide particles (Ti0 ₂ )

Ti0 ₂ particles are synthesized with the previously described method in Synthesis Example 3.

b) Co-functionalization of Ti0 ₂ particles with bisphosphonate-PEG-NH ₂ (BP- 01-302) and LCP

5 ml_ (0.62 mmol) of an aqueous solution of bisphosphonate-PEG-NH ₂ (Mw = 353, C ₈ H ₂₁ NOioP2, purchased from Surfactis Technologies, France) is mixed with 0.3 mL (0.013 mmol) of an aqueous solution of LCPs-a (short spacer). 7 mL (3.72 mmol) of a Ti0 ₂ suspension is added under stirring. The suspension is sonicated and stirred at 70 °C for 2 h. By this means particles are functionalized with 2% LCPs and 98 % of bisphosphonate-PEG-NH ₂ . The nanoparticles are precipitated with acetone, and washed three times with water and three times with DMF. Any fluorescence is observed in supernatant solution after washing. Particles are dispersed in 15 mL of DMF. c) Clicking of DOTA-NHS on Ti0 ₂ - bisphosphonate-PEG-NH ₂ -LCP

To this suspension is added DMF solution of 3 mL (0.5 mmol) of DOTA-NHS and 0.075 mL of triethylamine. The suspension is stirred at room temperature overnight. The nanoparticles are precipitated with acetone, and washed three times with DMF and three times with water. The particles are dispersed in 10 mL of water. d) Complexation of gadolinium ions on Ti0 ₂ - bisphosphonate-PEG-DOTA with Gadolinium ions

To this suspension is added under stirring, 0.32g (0.86 mmol) of GdCI ₃ , 6 H ₂ O in 2 ml water, and 5 drops of NaOH 1M. The mixture is stirred overnight. The nanoparticles are precipitated with acetone, and washed five times with water. The particles are dispersed in 10 mL of water.

Example 4 : Staining and selectivity of Ti0 ₂ nanoparticles coupled with LCPs (Ti0 ₂ -PEG-LCP) to amyloid-beta fibrils

Selectivity and affinity of TiO ₂ -PEG-LCP according to example 7 b) to Αβ-is shown in Fig. 10 in the case of LCPs-a. Amyloid plaques are stained by TiO ₂ -PEG-LCP as well as with LCPs alone. LCPs and TiO ₂ -PEG-LCP bind in a comparable fashion to the plaques as shown on Fig. 11. Example 5 : Luminescent spectroscopic properties of multifunctionalized GdF ₃ -bisphophonate-PEG-LCP

Fluorescence measurements shown on Fig. 12 are performed on the particles obtained in example 6. QE (quantum efficiency) spectral properties are very similar for NPs with short, longer spacer and LCP alone. Interestingly, the case with the long spacer (LCP-b) has approximately the same QE with and without GdF ₃ -bisphophonate-PEG. The shorter variant (LCP-a) shows a decrease in QE when attached to the GdF ₃ -bisphophonate-PEG. The proximity of the a-LCP when shortly bound to the GdF ₃ -bisphophonate-PEG makes the chromophores (LCP) vicinity quenching from each other.

Example 6 : Probes survive in vivo and pass the Blood-Brain-Barrier

After two consecutive in-vivo injections of 100 pi of GdF ₃ -PEG-LCP-b longer spacer (6.6 μΜ), or 35 μΜ GdF ₃ -PEG-LCP-a short spacer in 48 h, simple fluorescent microscope was used for analysis. All samples tested can cross the blood brain barrier (BBB) and label cerebral Αβ plaques as shown on Fig. 13. The distribution pattern assessed by fluorescence microscopy in Figure 13 and Figure 14 also clearly showed LCP staining of the plaque from i.v. injected mice. ICP-MS on brain hemispheres fresh frozen 7 days after injection showed elevated concentrations of Gd in the brains and with a tendency towards higher concentration among APPPS1 mice (Figure 15).