Synthetic immunogenic but non-deposit-forming polypeptides and peptides homologous to amyloid beta, prion protein, amylin, alpha-synuclein, or polyglutamine repeats for induction of an immune response thereto

The present application is a divisional of U.S. patent application Ser. No. 10/301,488, filed Nov. 21, 2002, which claims priority under 35 U.S.C. 119(e) of U.S. Provisional Application No. 60/331,801, filed Nov. 21, 2001, both of which are hereby incorporated by reference in their entirety.

The experiments performed in this application were supported in part by the National Institutes of Health, Grant Nos. AG08721, AR02594, AG17617, AG20245, AG02594, AG05891, and AG20197. The U.S. Government has a paid-up license in this invention and the right in limited circumstances to require the patent owner to license others on reasonable terms as provided for by the terms of each respective grant listed above.

1. Field of the Invention

The present invention relates to the field of amyloid β peptides, prion protein, amylin, α-synuclein, and polyglutamine repeats and methods for inducing an immune response to amyloid β peptides and amyloid deposits, prion proteins and deposits, amylin and amylin fibrils, filaments and fibrils containing α-synuclein, or protein aggregates containing polyglutamine repeats. The invention also relates to a method of treating or preventing amyloid plaque forming diseases or amyloidosis by the use of a synthetic immunogenic but not deposit forming polypeptide homologous to the protein which forms the amyloid plaque.

2. Description of the Background Art

Alzheimer\'s disease (AD) is the most common form of late-life dementia in adults (Soto et al., 1994), constituting the fourth leading cause of death in the United States. Approximately 10% of the population over 65 years old is affected by this progressive degenerative disorder that is characterized by memory loss, confusion and a variety of cognitive disabilities. Neuropathologically, AD is characterized by four major lesions: a) intraneuronal, cytoplasmic deposits of neurofibrillary tangles (NFT), b) parenchymal amyloid deposits called neuritic plaques, c) cerebrovascular amyloidosis, and d) synaptic and neuronal loss. One of the key events in AD is the deposition of amyloid as insoluble fibrous masses (amyloidogenesis) resulting in extracellular neuritic plaques and deposits around the walls of cerebral blood vessels. The major constituent of the neuritic plaques and congophilic angiopathy is amyloid β (Aβ), although these deposits also contain other proteins such as glycosaminoglycans and apolipoproteins.

Aβ is a 4.1-4.3 kDa hydrophobic peptide that is codified in chromosome 21 as part of a much longer amyloid precursor protein APP (Muller-Hill et al., 1989). The APP starts with a leader sequence (signal peptide), followed by a cysteine-rich region, an acidic-rich domain, a protease inhibitor motif, a putative N-glycosylated region, a transmembrane domain, and finally a small cytoplasmic region. The Aβ sequence begins close to the membrane on the extracellular side and ends within the membrane. Two-thirds of Aβ faces the extracellular space, and the other third is embedded in the membrane (Kang et al., 1987 and Dyrks et al., 1988). Several lines of evidence suggest that amyloid may play a central role in the early pathogenesis of AD.

Evidence that amyloid may play an important role in the early pathogenesis of AD comes primarily from studies of individuals affected by the familial form of AD (FAD) or by Down\'s syndrome. Down\'s syndrome patients have three copies of the APP gene and develop AD neuropathology at an early age (Wisniewski et al., 1985). Genetic analysis of families with hereditary AD revealed mutations in chromosome 21, near or within the Aβ sequence (Forsell et al., 1995), in addition to mutations within the presenilin 1 and 2 genes. Moreover, it was reported that transgenic mice expressing high levels of human mutant APP progressively develop amyloidosis in brain (Games et al., 1995). These findings appear to implicate amyloidogenesis in the pathophysiology of AD. In addition, Aβ fibrils are toxic in neuronal culture (Yankner et al., 1989) and to some extent when injected into animal brains (Sigurdsson et al., 1996 and 1997).

Furthermore, several other pieces of evidence suggest that the deposition of Aβ is a central triggering event in the pathogenesis of AD, which leads subsequently to NFT formation and neuronal loss. The amyloid deposits in AD share a number of properties with all the other cerebral amyloidoses, such as the prion related amyloidoses, as well as the systemic amyloidoses. These characteristics are: 1) being relatively insoluble; 2) having a high β sheet secondary structure, which is associated with a tendency to aggregate or polymerize; 3) ultrastructurally, the deposits are mainly fibrillary; 4) presence of certain amyloid-associated proteins such as amyloid P component, proteoglycans and apolipoproteins; 5) deposits show a characteristic apple-green birefringence when viewed under polarized light after Congo red staining. The same peptide that forms amyloid deposits in AD brain was also found in a soluble form (sAβ) normally circulating in the human body fluids (Seubert et al., 1992 and Shoji et al., 1992). Zlokovic et al. (1994), reported that the blood-brain barrier (BBB) has the capability to control cerebrovascular sequestration and transport of circulating sAβ, and that the transport of the sAβ across the BBB was significantly increased when sAβ was perfused in guinea pigs as a complex with apolipoprotein J (apoJ). The sAβ-apoJ complex was found in normal cerebrospinal fluid (CSF; Ghiso et al., 1994) and in vivo studies indicated that sAβ is transported with apoJ as a component of the high density lipoproteins (HDL) in normal human plasma (Koudinov et al., 1994). It was also reported by Zlokovic et al. (1996), that the transport of sAβ across the BBB was almost abolished when the apoJ receptor gp330 was blocked. It is believed that the conversion of sAβ to insoluble fibrils is initiated by a conformational modification of the 2-3 amino acid longer soluble form. It has been suggested that the amyloid formation is a nucleation-dependent phenomena in which the initial insoluble “seed” allows the selective deposition of amyloid (Jarrett et al., 1993).

Peptides containing the sequence 140 or 142 of Aβ and shorter derivatives can form amyloid-like fibrils in the absence of other protein (Soto et al., 1994), suggesting that the potential to form amyloid resides mainly in the structure of Aβ. The relation between the primary structure of Aβ and its ability to form amyloid-like fibrils was analyzed by altering the sequence of the peptide. Substitution of hydrophilic residues for hydrophobic ones in the internal Aβ hydrophobic regions (amino acids 17-21) impaired fibril formation (Hilbich et al., 1992), suggesting that AD assembly is partially driven by hydrophobic interactions. Indeed, larger Aβ peptides (Aβ1-42/43) comprising two or three additional hydrophobic C-terminal residues are more amyloidogenic (Jarrett et al., 1993). Secondly, the conformation adopted by Aβ peptides is crucial in amyloid formation. Aβ peptides incubated at different pH, concentrations and solvents can have either a mainly α-helical, random coil, or a β-sheet secondary structure (Barrow et al., 1992; Burdick et al., 1992 and Zagorski et al., 1992). The Aβ peptide with α-helical or random coil structure aggregates slowly; Aβ with β-sheet conformation aggregates rapidly (Zagorski et al., 1992; Soto et al., 1995 and Soto et al., 1996). The importance of hydrophobicity and β-sheet secondary structure on amyloid formation also is suggested by comparison of the sequence of other amyloidogenic proteins.

Analysis of Aβ aggregation by turbidity measurements indicates that the length of the C-terminal domain of Aβ influences the rate of Aβ assembly by accelerating nucleus formation (Jarrett et al., 1993). Thus, the C-terminal domain of Aβ may regulate fibrillogenesis. However, in vitro modulators of Aβ amyloid formation, such as metal cations (Zn, Al) (Bush et al., 1994 and Exley et al., 1993) heparin sulfate proteoglycans, and apoliprotein E (Strittmatter et al., 1993) interact with the 12-28 region of Aβ. Moreover, mutations in the APP gene within the N-terminal Aβ domain yield analogs more fibrillogenic (Soto et al., 1995 and Wisniewski et al., 1991). Finally, while the C-terminal domain of Aβ invariably adopts a β-strand structure in aqueous solutions, environmental parameters determine the existence of alternative conformation in the Aβ N-terminal domain (Barrow et al., 1992; Soto et al., 1995 and Burdick et al., 1992). Therefore, the N-terminus may be a potential target site for inhibition of the initial random coil to β-sheet conformational change.

The emerging picture from studies with synthetic peptides is that Aβ amyloid formation is dependent on hydrophobic interactions of Aβ peptides adopting an antiparallel β-sheet conformation and that both the N- and C-terminal domains are important for amyloid formation. The basic unit of fibril formation appears to be the conformer adopting an antiparallel β-sheet composed of strands involving the regions 10-24 and 2940/42 of the peptide (Soto et al., 1994). Amyloid formation proceeds by intermolecular interactions between the β-strands of several monomers to form an oligomeric β-sheet structure precursor of the fibrillar β-cross conformation. Wood et al., (1995) reported the insertion of aggregation-blocking prolines into amyloid proteins and peptides to prevent aggregation of such proteins and peptides. In this manner, the authors suggest that novel proteins can be designed to avoid the problem of aggregation as a barrier to their production without affecting the structure or function of the native protein. Thus, Wood et al. seek to produce novel proteins that would not aggregate during recombinant protein production and purification by inserting aggregation/blocking prolines into these novel peptides.

To date there is no cure or effective therapy for reducing a patient=s amyloid burden or preventing amyloid deposition in AD, and even the unequivocal diagnosis of AD can only be made after postmortem examination of brain tissues for the hallmark neurofibrillary tangles (NFT) and neuritic plaques. However, there are an increasing number of publications outlining strategies for the treatment of Alzheimer\'s disease. Amyloid-related therapeutic strategies include the use of compounds that affect processing of the amyloid-β precursor protein (APP; Dovey et al., 2001), that interfere with fibril formation or that promote fibril disassembly (Soto et al., 1998; Sigurdsson et al., 2000; and Findeis, 2000).

Heparin sulfate (glycosoaminoglycan) or the heparin sulfate proteoglycan, perlecan, has been identified as a component of all amyloids and has also been implicated in the earliest stages of inflammation-associated amyloid induction. Kisilevsky et al. (1995) describes the use of low molecular weight (135-1,000 Da) anionic sulfonate or sulfate compounds that interfere with the interaction of heparin sulfate with the inflammation-associated amyloid precursor and the β-peptide of AD. Heparin sulfate specifically influences the soluble amyloid precursor (SAA2) to adopt an increased β-sheet structure characteristic of the protein-folding pattern of amyloids. These anionic sulfonate or sulfate compounds were shown to inhibit heparin-accelerated Alzheimer\'s Aβ fibril formation and were able to disassemble preformed fibrils in vitro as monitored by electron micrography. Moreover, when administered orally at relatively high concentrations (20 or 50 mM), these compounds substantially arrested murine splenic inflammation-associated amyloid progression in vivo in acute and chronic models. However, the most potent compound, poly-(vinylsulfonate), was acutely toxic.

Anthracycline 4′-iodo-4′-deoxy-doxorubicin (IDOX) has been observed clinically to induce amyloid resorption in patients with immunoglobin light chain amyloidosis (AL). Merlini et al. (1995), elucidated its mechanism of action. IDOX was found to bind strongly via hydrophobic interactions to two distinct binding sites (Scatchard analysis) in five different tested amyloid fibrils, inhibiting fibrillogenesis and the subsequent formation of amyloid deposits in vitro. Preincubation of IDOX with amyloid enhancing factor (AEF) also reduced the formation of amyloid deposits. Specific targeting of IDOX to amyloid deposits in vivo was confirmed in an acute murine model. This binding is distinct from heparin sulfate binding as removal of the glycosaminoglycans from extracted amyloid fibrils with heparinases did not modify IDOX binding. The common structural feature of all amyloids is a β-pleated sheet conformation. However, IDOX does not bind native amyloid precursor light chains which suggests that the β-pleated sheet backbone alone is not sufficient to form the optimal structure for IDOX binding, and that it is the fibril cross-β-sheet quaternary structure that is required for maximal IDOX binding. It has been found that the amount of IDOX extracted from spleens is correlated with amyloid load and not circulating serum precursor amyloid levels. IDOX, however, is also extremely toxic.