Disease

The present invention relates to compounds for use in the prevention and treatment of Alzheimer's Disease (AD) .

Amyloid-β peptide (Aβ) plays a central role in the neuropathology of Alzheimer's disease (AD) (Roher et al 1993; PNAS 90:10836) . Familial forms of the disease have been linked to mutations in the amyloid precursor protein (APP) and the presenilin genes. Disease-linked mutations in these genes result in increased production of the 42-amino acid form of the peptide (Aβ42), which is the predominant form found in the amyloid plaques of Alzheimer's disease. An animal model for the disease is commercially available. The PDAPP transgenic mouse, which over-expresses mutant human APP (in which the amino acid at position 717 is F instead of V) , progressively develops many of the neuropathological hallmarks of Alzheimer' s disease in an age- and brain-dependent manner (Games et al 1995; Nature 373:523) .

A promising strategy has been developed using the apheresis technology by removing Aβ42 specifically from the individual having or being at risk of AD (PCT/AT2004/000311) .

Vaccination studies with a "normal", not mimotope-based vaccine have already been performed. Transgenic animals were immunized with aggregated Aβ42, either before the onset of AD-type neuropathologies (6 weeks) or at an older age (11 months) : Immunization of young animals prevented the development of plaque formation, neuritic dystrophy and astrogliosis . Treatment of older animals markedly reduced AD-like neuropathologies. This experimental vaccination approach induced the development of antibodies against Aβ42 able to cross the blood-brain barrier and attack amyloid plaques (Schenk et al 1999; Nature 400:173). The plaques are subsequently removed by several mechanisms, including Fc-receptor mediated phagocytosis (Bard et al 2000; Nature Med 6:916) . This vaccine was also able to delay memory deficits (Janus et al 2000; Nature 408:979).

A highly promising immunization therapy for AD has been in clinical trials since late 1999. Immunization is presumed to trigger the immune system to attack the plaques and clear these deposits from the affected human brain, although the precise mechanism underlying needs to be characterized in more detail.

These clinical trials were conducted by the pharmaceutical company Elan in conjunction with its corporate partner, American Home Products (therapeutic vaccine AN-1792, QS21 as adjuvant). Phase I trials were successfully completed in 2000. Phase II trials were begun late 2001 to test efficacy in a panel of patients with mild to moderate AD.

Now these phase II trials have been permanently discontinued due to neuroinflamination in several patients (Editorial 2002 "Insoluble problem?" Nature Med 8:191) . The symptoms included aseptic meningoencephalitis leading to the immediate halt of these world-wide trials. In the worst case scenario, affected patients will be shown to have mounted an autoimmune response - a risk inherent in many immunotherapies . Autoimmune complications could have been anticipated given the ubiquity of APP, which of course bears antigenic determinants in common with its proteolytic product. More recently, additional studies concentrated on the nature of aggregated Aβ42 immunization-induced antibodies (in humans and mice) revealing that most antibodies recognize a small domain between amino acid 4 and 10 of Aβ42 (Aβ4-10) . The mouse antibodies were able to block Aβ fibrillogenesis and disrupted pre-existing Aβ fibres (McLaurin et al 2002; Nature Med 8:1263) . Of note, the human antibodies do not react with APP exposed on the surface of cells or any other non-aggregated proteolytic product of the precursor (Hock et al 2002; Nature Med 8:1270) . A clear difference was observed between human and mouse sera: In contrast to human antibodies, mouse antibodies detect monomeric, oligomeric, and fibrillar Aβ . This is of importance and may be a prerequisite for the therapeutic potency since evidence is accumulating that small oligomers of Aβ, which are not recognized by human anti-Aβ, are the major toxic players in

the disease (Walsh et al 2002; Nature 416:535) . Thus, a potential new strategy is the immunization with a vaccine containing β-amyloid amino acids 4-10 (instead of aggregated Aβ42) . Despite unknown efficacy this strategy may also face autoimmune problems since patients shall be directly immunized with a (linear B cell) "self" epitope.

Therefore, there is still an urgent need for new and alternative strategies for the treatment or the prevention of AD which can be used additionally or alternatively with other methods .

The present invention therefore relates to the use of a compound having the formula (I)

R _l X _l X ₂ X ₃ X ₄ X ₅ XgX ₇ X ₈ R ₂ ( I ) ₁ wherein

X ₂ is H;

X ₃ is I, L, V or A;

X ₄ is I, L, V or A;

X ₅ is I, L, V or A;

X ₆ is E or D;

X ₇ is F or Y;

X ₈ is I, L, V or A; and Xi to X ₈ are covalently linked by peptide bonds;

Ri and R ₂ are suitable N- and C-terminal groups of the peptide defined by X ₁ to X ₈ ; for the preparation of a medicament for the prevention or treatment of Alzheimer's Disease (AD) in a patient. The amino acid residues of the formula (I) are given in the single letter code, i.e. H is His, I is lie, L is Leu, V is VaI, A is Ala, E is GIu, D is Asp, F is Phe and Y is Tyr.

With the compounds according to the present invention, a "peri-

pheral sink" effect for reducing Aβ brain load can be achieved. The "peripheral sink" effect as mechanism to reduce Aβ brain load was originally described by DeMattos et al in 2001 (PNAS 98:8850) . It was shown that the monoclonal antibody (rtiAb) 266 (anti-Aβ, domain 13-28) rapidly sequestered all plasma Aβ present in APP-transgenic (PDAPP) mice and caused a large accumulation of brain-derived Aβ in the plasma. Furthermore, chronic parenteral treatment with rtiAb 266 resulted in substantial suppression of Aβ deposition in the brain hereby suppressing AD- like pathology by altering Aβ clearance from CNS to plasma. No binding of rtiAb 266 to Aβ deposits in the brain could be detected. Administration (i.v.) of 500μg rtiAb 266 to PDAPP mice resulted in a 1000-fold increase in plasma Aβ with a maximum level by day 4. Parenteral administration of rtiAb 266 altered the equilibrium of Aβ between the central (CNS) and peripheral (plasma) compartments . Previous studies have demonstrated rapid transport of exogenous Aβ between CNS and plasma. Recently, efficient receptor-mediated transport mechanisms for Aβ at the blood brain barrier have been reported.

DeMattos et al confirmed their own results in 2002 (Science 295:2264, Journal of Neurochem 81:229). Furthermore, they demonstrated that the magnitude of plasma Aβ increase was highly correlated with amyloid burden in the hippocampus and cortex of PDAPP mice. Matsuoka et al described a similar effect when either gelsolin or the ganglioside GMl was taken to sequester plasma Aβ (Matsuoka et al 2003, Journal of Neuroscience 23:291) . Gelsolin and GMl were especially effective when relatively young mice were used for experiments . It was speculated that compounds that specifically bind Aβ in the periphery (blood) may be clinically efficacious .

It has been described (PCT/AT2004/000311) that the small 17mer peptide 1208 (FGFPEHLLVDFLQSLSC) is able to bind Aβ peptides in vitro. An in vivo effect or any pharmaceutical effects if administered to an individual, especially to human individuals having or being at risk of developing AD, has not been demonstrated.

With the present invention, the compounds according to formula (I) which have been developed from peptide FGFPEHLLVDFLQSLSC and its derivatives comprising comparable in vivo effects on Aβ42 (i.e. with respect to peripheral sink efficacy) are provided to be administered to individual, especially to human individuals having or being at risk of developing AD, in an amount effective to reduce Aβ load in brains. The administration is performed by making use of the peripheral sink effect. Any pharmaceutical preparation comprising compound (I) is therefore preferably administered peripherally. In order to achieve a long term effect of the peripheral sink, which is in many cases preferred, a retard or deposit form of compound (I) may be provided and administered to the patient. Such retard or deposit forms enabling slow release are well available to the skilled man in the art and may easily be optimised depending on the specific physico- chemical and pharmacokinetic properties of a given compound according to formula (I) .

Usually, Ri and R ₂ are not critical for the present invention; these residues should be present in a form not significantly disturbing (spatially and physiologically) the effect of the core of the compound of the formula (I) with Aβ42 in the body. Therefore, Ri and R ₂ should not be too bulky. As a rule of thumb, Ri and R ₂ are selected to result in a molecule having the same or similar binding characteristics (binding affinity, binding strength) to Aβ42 as peptide 1208 (a binding affinity which is in the same order of magnitude or better, preferably not less than 50 % of peptide 1208) . Preferably, R ₁ and R ₂ are amino acid residues, peptides or polypeptides covalently linked to the peptide defined by Xi to X ₈ by peptide bonds . The compound according to formula (I) is then a peptide, preferably consisting of the 20 naturally and universally occurring amino acid residues.

If the compound according to formula (I) comprises an N- or C- terminal cystein (C) moiety, the pharmacological handling of the compound, the production of the compound and the binding effectiveness to Aβ42 may be significantly improved. It is therefore preferred that Ri and/or R ₂ comprise an N- or C-terminal cystein

(C ) moiety .

Although it is in many instances preferred that the compound according to formula (I) is provided as a small molecule drug, other forms of the compound may also be provided in larger form, e.g. coupled to larger molecules or even bound to (small) surfaces thereby forming a suspension to be applied to the individual. This might be specifically chosen for retard forms wherein the compound according to formula (I) is slowly released by e.g. hydrolysation of a linker bond to such larger molecules or surfaces in vivo. In such cases, Ri and/or R ₂ are designed to have suitable linker groups or other functional chemical groups enabling a covalent or strong electrostatic binding to the larger molecules or surfaces . According to a preferred embodiment of the present invention Ri and/or R ₂ comprise chemical linker groups for the attachment of pharmaceutical carriers or pharmaceutically active compounds . On the other hand, R ₁ and/or R ₂ may also be a simple chemical group, such as methyl, ethyl, acetyl, formyl, and the like, or even a hydrogen atom (depending on the ionic state of the terminal amino acid group) , thereby forming an N-terminus and/or a C-terminus with Xi (NH ₂ or NH ₃ ⁺ ) and/or X ₈ , (COO ^" or COOH) respectively.

In preferred compounds according to formula (I), Ri is or comprises an amino acid or a peptide selected from the group P, FP, GFP, FGFP, CP, CFP, CGFP, CFGFP or CFGFP. In preferred compounds according to formula (I), R ₂ is or comprises an amino acid or a peptide selected from the group Q, QS, QSL, QSLS, QC, QSC, QSLC or QSLSC.

As already mentioned, the medicament according to the present invention is provided as a slow release drug with respect to the compound according to formula (I) . Preferred forms of administration comprise intravenous, subcutaneous, oral, intramuscular or mucosal administration. Suitable pharmaceutical carriers, additives, formulations, dosages and application means are available to the man skilled in the art.

Preferably, the medicament according to the present invention contains a peptide selected from FGFPEHLLVDFLQSLS, GFPEHLLVD-

FLQSLS, FPEHLLVDFLQSLS, PEHLLVDFLQSLS, EHLLVDFLQSLS, FGFPEHLLVD- FLQSL, FGFPEHLLVDFLQS, FGFPEHLLVDFLQ, FGFPEHLLVDFL, FGFPDHLLVDFLQSLS, GFPDHLLVDFLQSLS, FPDHLLVDFLQSLS, PDHLLVD- FLQSLS, DHLLVDFLQSLS, FGFPDHLLVDFLQSL, FGFPDHLLVDFLQS, FGFP- DHLLVDFLQ, FGFPDHLLVDFL, GFPDHLLVDFLQSL, FPDHLLVDFLQS, GFPDHLLVDFLQS, PDHLLVDFLQ, FPDHLLVDFLQ, DHLLVDFLQ, GFPDHLLVDFLQ, GFPEHLLVDFLQSL, FPEHLLVDFLQSL, GFPEHLLVDFLQS, PEHLLVDFLQSL, GFPEHLLVDFLQSL, PEHLLVDFLQ, FGFPEHLLVEFLQSLS, GFPEHLLVEFLQSLS, FPEHLLVEFLQSLS, PEHLLVEFLQSLS, PEHLLVEFLQSL, FGFPEHLLVEFLQSL, FGFPEHLLVEFLQS, FGFPEHLLVEFLQ, FGFPEHLLVEFL, GFPEHLLVEFLQSL, FPEHLLVEFLQSL, FGFPEHLLVEFLQSLS, FFPEHLLVEFLQSLS, FGFPEHLLVE-

FLQSS , FGPEHLLVEFLQSLS , a peptide wherein X ₃ X ₄ X ₅ i s LLV, ILV, ALV,

LLL, III, AAA, VW, LW, LVL, VLL, IVL, LLA, LLI, LIV, LAV, LLA, HA, IAI, AVA, AAV OR VAA, FGFPEHLLVEYLQSLS, FGFPEHLLVDYLQSLS, FGFPDHLLVEYLQSLS, GFPEHLLVDYLQSLS, FPEHLLVDYLQSLS, PEHLLVDYLQSLS, FGFPEHLLVDYLQSL, FGFPEHLLVDYLQS, FGFPEHLLVDYLQ, FGFPEHLLVDYL, GFPEHLLVDYLQSL, FPEHLLVDYLQSL, FPEHLLVDYLQS, PEHLLVDYLQSL, PEHLLVDYLQS, PEHLLVDYLQ, FGFPEHLLVDFIQSLS, FG-

FPEHLLVDFAQSLS, FGFPEHLLVDFVQSLS, FGFPEHLLVDFX ₈ QSLS, wherein X ₈ is defined as in claim 1; GFPEHLLVDFX ₈ QSLS, wherein X ₈ is defined as in claim 1; FPEHLLVDFX ₈ QSLS, wherein X ₈ is defined as in claim 1; PEHLLVDFX ₈ QSLS, wherein X ₈ is defined as in claim 1; FG- FPEHLLVDFX ₈ QSL, wherein X ₈ is defined as in claim 1; FGFPEHLLVD- FX ₈ QS, wherein X ₈ is defined as in claim 1; FGFPEHLLVDFX ₈ Q, wherein X ₈ is defined as in claim 1; FGFPEHLLVDFX ₈ , wherein X ₈ is defined as in claim 1; fragments of these peptides having comparable binding strength to Aβ42; and/or derivatives of these peptides comprising one amino acid exchange in X ₁ to X ₈ as defined in claim 1, and/or at least one amino acid exchange, deletion or insertion in Ri and R ₂ , wherein said exchange, deletion or insertion is not negatively affecting binding strength to Aβ42.

According to another aspect, the present invention provides a pharmaceutical preparation comprising compounds having the formula (I) as defined above and a pharmaceutically acceptable car-

rier .

Moreover, the present invention provides for compounds having the formula (I) as defined above, preferably the compounds of formula (I) except peptide 1208 (FGFPEHLLVDFLQSLSC) and except the peptide with the sequence FGFPEHLLVDFLQSLS . Both the WO 99/66957 A2 and the US 2005/0025782 Al mention a multitude of T helper cell antigens, including the CETP (cholesteryl-ester- transportprotein) antigen FGFPEHLLVDFLQSLS.

The present invention is further described in the following example and the figures, yet without being restricted thereto.

Fig. 1 shows competition of Aβ42 binding to 1208-BSA;

Fig. 2 shows the effect of peptide 1208 in vivo;

Fig. 3 shows the effect of peptide 1208 on Aβ42 brain load.

Examples :

1. : Binding competition studies

With the present examples, binding studies of peptide 1208 and N-terminally truncated versions of peptide 1208 are performed as described in WO2004/062556 A.

The following peptides were tested with respect to binding competition: peptide competition

FGFPEHLLVDFLQSLSC high

CFGFPEHLLVDFLQSLS high

GFPEHLLVDFLQSLSC high

PEHLLVDFLQSLSC intermediate

HLLVDFLQSLSC no

Truncations of several N- and C-terminal amino acid residues in peptide 1208 can still bind Aβ peptides. Figure 1 shows that soluble peptide 4071 (amino acid sequence identical with peptide 1208) strongly competes with Aβ42 for binding to BSA-coupled, plate-bound peptide 1208) and, thus, inhibits binding of Aβ42 to 1208-BSA. Peptide 4074 (1 N-terminal amino acid missing, GFPEHLLVDFLQSLSC) is competing in a similar way, peptide 4075 (3 N-terminal amino acids missing, PEHLLVDFLQSLSC) is competing worse, peptide 4076 (5 N-terminal amino acids shorter, HLLVDFLQSLSC) is basically not competing (like an unrelated control peptide) with Aβ42 for binding to plate-bound 1208-BSA.

Furthermore, peptide 1208 is not only binding Aβ42 in ELISA, but also specifically detects Aβ deposits in the brain of aged APP- transgenic mice.

2. : Binding Aβ peptides in vivo

It was also shown that the small peptide 1208 is also binding Aβ peptides in vivo (see Figure 2) : APP-transgenic mice were either injected with PBS (mouse 75, lOOμl i.v.) or with peptide 1208 (mouse 83, lOOμg peptide 1208 in lOOμl PBS i.v.). The blood levels of Aβ42 were investigated 1 h after injection: Whereas the level of peptide Aβ42 was hardly affected in the mouse injected with PBS only, the amount of Aβ42 was about 8-fold increased in the mouse injected with peptide 1208 (in PBS) .

3. : Peripheral sink effect of the present invention in mice

Next, the effect of repetitive injections of peptide 1208 on Aβ42 brain load of APP-transgenic mice was investigated (see

Figure 3) . APP-transgenic mice were either injected 5x times within 3 months with PBS (mouse 26, lOOμl i.v.) or with peptide 1208 (mouse 82, lOOμg peptide 1208 in PBS i.v.). 24 h after the last injection mice were killed and the level of total Aβ40/42 brain load determined. Figure 3 shows a strong qualitative decrease of Aβ40/42 plaque burden (as measured in a non-quantitative specific ELISA) .

Taken together, these data show that peptide 1208 ("depletin") and its analogues and derivatives are able to sequester Aβ peptides in the bloodstream, hereby influencing the Aβ blood-brain equilibrium: The efflux from Aβ peptides from brain to blood finally decreases the overall Aβ brain load.