METHOD OF INTERFERING WITH FORMATION OF α-ANTICHYMOTRYPSIN-β-PROTEIN COMPLEX AND SYNTHETIC PEPTIDES FOR USE THEREIN
Background
Alzheimer's disease is a degenerative disorder of the central nervous system that results in a progressive loss of memory and other intellectual functions, such as reasoning, orientation, and judgement (R. Katzman, Banbury Report 15: Biological Aspects of Alzheimer's Disease, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, (1983)).
Alzheimer's disease occurs in sporadic and familial forms, and in the United States, affects about 600 people for every 100,000. A characteristic aspect of the neuropathology of the disease is the
occurrence of proteinaceous deposits referred to as "amyloid" in the cores of brain lesions called neuritic or senile plaques, as well as in cerebral blood vessels. The "amyloid" deposits are generally defined as 6-10 nm protein filaments with certain staining properties (Abraham, C.R. et al., Cell,
52:487-501 (1988)).
Amyloid deposits are also found in the brains of aged humans, although not as extensively as in Alzheimer's disease. Further, Down's syndrome patients more than 30 or 40 years old invariably develop the symptoms and neuropathology characteristic of Alzheimer's disease.
One component of the amyloid deposits was identified as A4 amyloid or y9-protein (β-protein) and is 42 amino ac ids long ( G l enner , G . G and C . G . Wong , Biochem. Biophys. Res. Commun. 120 : 885 - 890 (1984)). This protein is apparently derived from a larger membrane-spanning precursor protein which is alternately spliced to yield several products
(Selkoe, D.J., Science, 248: 1058-1060). These observations suggested that the amyloid deposits in Alzheimer's disease could result from abnormal expression or pos ttranslational modification or processing of a normal molecule. Also intriguing was the finding that the gene encoding the amyloid protein precursor is located on chromosome 21, suggesting a common cause for the the deposits observed in Down syndrome, caused by trisomy of chromosome 21, and Alzheimer's disease.
As mentioned above, some cases of Alzheimer's disease appear to be familial, and are inherited in an autosomal dominant fashion. Linkage analysis in four families pointed to a lesion on the long arm of chromosome 21 (St. George-Hyslop, P.H. et al.,
Science, 238: 664-660 (1987)), which correlated well, with the mapping data and similarities between Down syndrome and Alzheimer disease. Recently,
hereditary cerebral hemorrhage with amyloidosis of Dutch origin was reported to be linked to the APP gene, and a point mutation in the coding region of the gene was identified (Van Broeckhoven, C. et al.. Science, 248:1120-1122 (1990); Levy, E. et al., Science, 248 :1124-1126 (1990)). Patients with this disease have a form of the β-protein in amyloid deposits in meningeal and cerebral blood vessels.
However, other studies reported linkage of familial Alzheimer's disease to a locus on chromosome 21 distinct from the amyloid precursor protein (APP) gene (Tanzi, R.E. et al, Nature, 329:156-157 (1987); Van Broeckhoven, C. et al., Nature,
329:153-155 (1987)). Furthermore, there was no evidence of duplication of the APP gene in cases of familial or sporadic disease. In fact, studies of some families reportedly indicate no linkage to chromosome 21 ( Schellenberg, G.D., Science,
241:1507-1510, (1988). These data suggest that there may be genetic heterogeneity in the cause of inherited forms of Alzheimer's disease, and other locations for the disease gene have been proposed, such as chromosome 14 (Weitkamp, L.R., Amer. J. Hum.
Genet. 35:443-453 (1983)).
Thus, other components of the proteinaceous deposits in Alzheimer's disease may also be of interest and may provide clues to the cause or progress of the disease. In fact, a second component of the amyloid deposits has been characterized as α1-antichymotryps in (ACT), which, interestingly, is located on chromosome 14. Abraham et al.
reported the identification of the serine protease inhibitor ACT in amyloid deposits in Alzheimer's disease brain. (Abraham, C.R. et al., Cell
52:487-501 (1988).
Summary of the Invention
This invention relates to a novel class of synthetic peptides or peptide-like compounds which mimic a component of a specific complex between the Alzheimer's β-protein and ACT, and which are useful to interfere with
formation of the complex. The invention also relates to a method of treating an individual in whom such complexes form, resulting directly or indirectly in an abnormal condition or disease state, and particularly to a method of treating an individual with Alzheimer's disease. The synthetic peptides of the present invention, which inhibit complex formation between the Alzheimer's β-protein and ACT by binding to the β-protein or to ACT, can be administered to an individual in such a manner as to interfere with the ACT-β-protein interaction, and in sufficient quantity so as to have the desired effect (i.e., reduction of complex formation and the abnormal disease state). Brief Description of the Drawings
Figure 1 is the amino acid sequence of
Alzheimer amyloid β-protein (top row) and the active sites of the following five serine proteases:
cytotoxic T cell protease; cathepsin G; mast cell protease; trypsin; and chymotrypsin.
Figure 2 is a graphic representation of the protease regulatory activity of Alzheimer amyloid β-protein. De tai l e d Descri p t ion o f the I nvent io n
The protease inhibitor α1-antichymotrypsin (ACT) and the 42-aa β-protein are integral components of the brain amyloid deposits of Alzheimer's disease, Down's syndrome, and normal aging. This indicates that there is a special affinity between ACT and the β-protein, perhaps essential to amyloid formation. A basis for this association is
suggested by the similarity of the N-terminus of β-protein to the active site of serine proteases.
As described herein, in vitro experiments demonstrate that ACT and β-protein form a complex that reflects the specificity and stability of a protease-inhibitor interaction. These results suggest a model for the amyloid filament and a physiological function for the β-protein. They also make it possible to design or select synthetic peptides which can be introduced into cells in order to interfere with
(reduce or prevent) binding of the two components of the complex. Such synthetic peptides include peptides and peptide-like compounds (e.g., modified or derivatized peptides) which interfere with interaction of ACT with β-protein, particularly peptides which "mimic" the active (or binding) site of serine proteases. Such peptides can be short peptides in which the amino acid sequence is sufficiently homologous with the sequence of the binding site of a serine protease or with the N-terminus of β- protein as shown in Figure 1, that they bind with ACT. Alternatively, such peptides can include, in addition to the sequence sufficiently homologous with the serine protease binding site or β-protein region, other amino acids (e.g., one or more amino acids at either or both ends of the binding site sequence). According to the method of the present invention, a compound of the present invention is introduced into an individual in such a manner that it interferes with formation of the ACT-β-prote in complex. In the present method of interfering with complex formation, a compound is administered in sufficient quantity and by an appropriate route to have a therapeutic effect.
The following is a description of the
demonstration that ACT and β-protein interact with specificity to form a stable complex in vitro and of the compounds which can be used to interfere with the interaction. Such compounds and the method by which they are administered are also described.
The present invention can be used in the treatment of individuals, such as those with
Alzheimer's disease, in whom complex formation would otherwise occur. Demonstration of Formation of ACT-β-Protein Complex
The extracellular amyloid filaments found in the plaques and blood vessels of Alzheimer's disease, Down's syndrome, and normal aging, contain two proteins, both intimately associated with the filaments - the 42 amino acid β-protein and the serine protease inhibitor, α1-antichymotrypsin
(Glenner, G . G . and C.W. Wong, Biochem. Biophys. Res.
Comm., 122:885 (1984); C.L. Masters, et al., EMBO J. 4:2757 (1985); D.J. Selkoe, et al., J. Neurochem.
46:1820 (1986); D.J. Selkoe et al., Science, 235:873 (1987); Abraham, C.R. and H. Potter, Bio/technology 7:147 (1989); D.J. Selkoe, Ann. Rev. Neurosci.
12:493 (1989); B. Muller-Hill, Ann. Rev. Biochem. 58:287 (1989); Neve, R.L. and H. Potter, in
Molecular Genetic Approaches to Neuropsychiatric
Disease, J. Brosius and R. Fremeau, Eds. (Academic Press, San Diego, in press); C.R. Abraham, et al . , Cell, 52:487 (1988); C.R. Abraham, et al.,
Neuroscience, 32:715 (1989)). Recently, it was found that the major component of the vascular amyloid in the Dutch variant of hereditary cerebral
hemorrhage with amyloidos is (HCHWA-D) is the β-protein, not cystatin C, as in the Icelandic version (HCHWA-I), although the angiopathy appears similar (J . Ghiso, et al., Proc. Natl. Acad. Sci. 83:2974 (1986); S.G. van Duinen, et al., Proc. Natl Ac ad . S c i 84 : 5 991 ( 1987 ) ) . When b r a in s e c t i ons from individuals with this disease were analyzed by immunolabeling, ACT was also found to be present (M.M. Picken, et al., Am. J. Pathol., 134:749
(1989)). In contrast, amyloid deposits found in other diseases do not contain either β-protein or ACT. The biochemical characteristics of ACT and the β-protein suggests a basis for this special association. First, ACT is a serine protease inhibitor that functions by acting as a pseudosubs tr ate and binding covalently to its target protease to form a long-lived complex (J. Travis, et al., Biochemistry, 17:5651 (1978)). Second, an inspection of the sequence of the β-protein reveals a region near the N-terminus which shows a striking homology to one segment of the active site of serine proteases, including the key serine amino acid (Figure 1) (E. Roberts, Neurobiol. Aging, 7:561 (1986); Travis, J. and G.S. Salve sen, Ann. Rev. Biochem, 52:655 (1983); G. Salvesen, et al., Biochemistry, 26:2289 (1987)). Thus, it seemed possible that ACT and β-protein might be able to form a complex by virtue of a protease inhibitor-like interaction and that this complex contributes to the stability of Alzheimer amyloid filaments. This hypothesis was tested, as described below. Results showed that, in fact, the β-protein is able to specifically bind stably to the inhibitory active site of ACT.
As described in Example 1, various synthetic peptides were tested for their effect on the
inhibition of chymotrypsin by ACT in vitro.
Figure 2 shows the results of assessment of protease activity, carried out with the synthetic peptides. As shown in Figure 2, when the ACT/chymotrypsin molar ratio was approximately 1:1, the ACT inhibited over 90 percent of the chymotrypsin activity. However, when ACT was pre-incubated with an approximately four-fold molar excess of synthetic peptides corresponding to amino acids 1-12 or 1-28 of the β-peptide prior to the addition of chymotrypsin, the inhibitory activity of ACT was substantially reduced and the chymotrypsin reaction rate increased 2 to 8-fold. In contrast, preincubation with even a 10-fold molar excess of a peptide corresponding to amino acids 258-277 of the β-protein precursor (which shows no similarity to the active site of serine proteases), failed to interfere with ACT. These data indicate that peptides showing similarity to the region around the key serine in the active site of serine proteases (Figure 1), and in particular the Alzheimer amyloid, β-protein, are able to interfere with the inhibitory function of a serine protease inhibitor, ACT. The specificity of the interaction indicates that it is occurring at the inhibitory active site of ACT. The fact that ACT still shows a substantial ability, even in the presence of the peptide, to inhibit chymotrypsin, probably reflects the fact that the binding of a protease inhibitor to its target involves many contacts with amino acids in the full
protease active site which the small peptides naturally lack. Assessment of the Stability and Specificity of Formation of the α1-ACT-β Protein Complex
Unlike true substrates that form a transient covalent intermediate with the protease through the hydroxyl group of its reactive serine, and then become cleaved as the bond breaks and the protease resumes its active state, protease inhibitors become cleaved, but only very slowly release their attachment to the protease (J . Travis, e t a l . , Biochemistry, 17:5651 (1978)). Thus, the inhibitors essentially inactivate the protease in a suicidal, stoichiometric interaction. The fact that serine protease inhibitors form stable complexes with their target proteases and that a serine protease inhibitor, ACT, is an integral component of the insoluble Alzheimer amyloid deposits, suggests that this protein might be incorporated into the filaments through a stable inhibitor-protease interaction.
Therefore, the stability of the interaction between α1- antichymotryps in and the β-protein was assessed. As described in Example 2, radio - iodinated peptides corresponding to amino acids 1-12 and 1-28 of the β-protein and the unrelated segment 258-277 of the β-protein precursor were prepared, incubated in the presence of ACT under various conditions, and the mixture elec trophoresed on SDS polyacrylamide gels. In the absence of ACT, the low M.W. peptides
migrated rapidly at the dye font. However, in the presence of ACT, a radioactive band was generated at a position corresponding to a few thousand M.W.
larger than the ACT protein. Thus, the interaction first detected in Figure 2 between ACT and the β-protein can be sufficiently stable to resist denaturation by boiling in SDS and β-mercaptoethanol. The indication in Figure 2 that the
ACT- -β-protein interaction occurs at the protease inhibitory site of ACT was confirmed by the fact that the addition of chymotrypsin to ACT prior to the radioactive peptide prevented the formation of the ACT-peptide complex. Heat denaturation of ACT prior to the addition of the peptide also prevented complex formation.
The specificity of the α1- ACT - β protein interaction was demonstrated by incubating the 1-28 peptide with other proteins (e.g., BSA, CPK, and
CA) , with which it failed to form stable complexes.
In sum, the work described indicates that the two components of the Alzheimer amyloid deposits¬- -ACT and the β-protein- - can associate in vitro to form an SDS-stable complex. The interaction is specific for both the peptide and the ACT protein, and likely occurs between the active protease inhibitor site of ACT and the N-terminus of the β-protein, which resembles the active site of serine proteases.
These results suggest a model for the structure of the Alzheimer amyloid filaments. The hydrophobic C-terminal portion of β-protein molecules making up the filament would be tucked into the interior, while the hydrophilic central segment (amino acids 12-28) would form the surface of the filament. The protease active site-related amino acids at the N-terminus would form an arm projecting from the
surface of the filament and available for binding by ACT. However, the latter seems at least possible, inasmuch as the β-protein alone can form filaments ΪB Yi^££ having a β-pleated sheet conformation (D.A. Kirschner, et al., Proc. Natl . Acad. Sci., 83 :
(1986); E.M. Castano, e t al., Biochem. Biophys. Res.
Comm., 141:782 (1986); D.A. Kirschner, et al., Proc. Natl. Acad. Sci., 84:6953 (1987)). However, these filaments can easily be solubilized and therefore must lack a key component or structural conformation characteristic of true Alzheimer amyloid filaments. It is possible that the binding of a β-protein core filament to the relatively protease - res is tant ACT provides the required extra stability.
These results also suggest a potential biological function for the β-protein. Since the β-protein can competitively interact with the active site of a serine protease inhibitor in vitro, it might be expected to be able to play a similar role in vivo. By decoying protease inhibitors (including the
Kunitz-type inhibitor in the β-protein precursor), the β-protein, though not itself a protease, would serve as a protease enhancer - - effect ively increasing proteolytic activity in its vicinity. Overproduction of the β-protein, as would result, for
instance, from increased proteolytic degradation of the β-protein precursor, could thus lead to a
further increase in protease activity, with
progressively adverse consequences.
Thus, as described herein, it has been shown that, in vitro, a serine protease inhibitor (ACT) interacts specifically with a serine protease-like target pro te in (Alzhe imer β-protein) at a region of
the latter which bears striking sequence homology to the active site of serine proteases, resulting in formation of a stable ACT-β-protein complex. The work described herein provides a reasonable molecular mechanism for formation of the insoluble protein filaments that comprise the amyloid deposits of Alzheimer's disease.
As a result of the discovery of the specific interaction between ACT and Alzheimer β-protein, it is possible to design or select compounds which are useful to interfere with (reduce or prevent) this ACT-β-protein interaction. The discovery has also made possible a method of preventing the interaction. The compounds and method are particularly useful in reducing formation of ACT-β- protein complexes in individuals in whom such complexes form and result, directly or indirectly, in an abnormal or undesirable condition or a disease state. For example, such compounds can be used to reduce (totally or partially) or prevent formation of ACT-β-protein complexes in individuals who have Alzheimer's disease or would, without appropriate treatment, develop Alzheimer's disease.
Compounds of the present invention can be used, to interfere with binding of ACT and Alzheimer's β -protein, either by binding to ACT, thus preventing formation of the ACT- β-protein complex, or by bind- ing to β-protein, also preventing formation of the ACT-β protein complex. For example, a peptide corresponding to all or a portion of the amino acid sequence of a serine protease, such as all or a por- tion of the sequence of the Alzheimer's β-protein can be used.
Alternatively, a synthetic peptide which
mimics the amino acid sequence of the inhibitory active site of ACT can be used to interfere with ACT-β-protein complex formation. This type of peptide will bind to the β-protein, resulting in the production of a synthetic peptide- β-protein complex and, in essence, will "tie up" β-protein, precluding the β-protein from interacting with ACT.
The peptide compounds of the present invention may have additional effects resulting from their influence, direct or indirect, on the protease inhibitor function of ACT. ACT is a serine protease inhibitor and Alzheimer's β-protein has been shown to inhibit ACT's ability to inhibit a serine protease (i.e., chymotrypsin). Thus, because it decoys protease inhibitors, β-protein, although not itself a protease, enhances protease activity by effectively increasing proteolytic activity (i.e., by reducing ACT's ability to exhibit serine proteases).
This view of the interaction between ACT and the β-protein suggests alternate outcomes from administration of the two types of peptide. In the case where the peptide binds to ACT to inhibit complex formation, the peptide may have an
inhibitory effect on ACT, resulting in enhancement of protease activity, in a manner similar to that suggested for the β-protein itself. In the case in which the synthetic peptide mimics the inhibitory active site of ACT and binds to the β-protein, one might expect the opposite outcome. That is, the peptide would complex with the β-protein, preventing its interaction with ACT, and thus blocking the postulated protease enhancing effect of the
β-protein. ACT then would be left free to exercise its function as a protease inhibitor.
The discovery described herein also makes it possible to identify ACT-β-protein complexes in tissue obtained from an individual, such as in biopsy tissues obtained from an individual suspected of having abnormally high levels of the complexes. This can be used, for example, to determine the presence or absence and, if desired, the quantity of ACT-β-protein complexes in brain tissue obtained at autopsy.
A synthetic peptide of the present invention can be administered in a physiologically acceptable carrier (e.g., an appropriate buffer or physiologic saline solution). It can be administered by any route via which it is possible to deliver a
therapeutically effective quantity or dose to the individual in a form available to have the desired effect (reduction of complex formation). For example, a synthetic peptide of the present
invention can be administered parenterally (e.g., intravenously or intramuscularly) in a composition which protects the peptide from degradation.
A synthetic peptide of the present invention can be made using known techniques, such as
recombinant/genetic engineering techniques or chemical synthesis.
EXAMPLE 1 Assessment of the Protease Regulatory
Activity of Alzheimer β-Protein
Various synthetic peptides were tested for their effect on the inhibition of chymotrypsin by ACT in vitro. The synthetic peptides tested were:
1. a peptide corresponding to amino acids 1-28 of the N-terminal portion of the Alzheimer β protein (See Figure 1).
2. a peptide corresponding to amino acids 1-12 of the N-terminal portion of the Alzheimer β protein (See Figure 1, first 12 amino acids); and
3. a control peptide corresponding to amino acids 258-277 of the Alzheimer β-protein precursor, which does not show any similarity to the serine protease active region sequence. The activity of chymotrypsin was measured by cleavage of the chromogenic substrate Succinyl-Ala-Ala-Pro-Phe-nitroanilide. Results showed that chymotrypsin activity is reduced about 90 percent in the presence of an approximate equal molar
concentration of ACT. If a four-fold molar excess of either of the two synthetic peptides from the N-terminal portion of the β-protein (corresponding to amino acids 1-28 and 1-12 respectively) are added to ACT prior to the protease assay, they interfere with the inhibitory activity of ACT. In contrast, the control peptide corresponding to amino acids 258-277 of the β-protein precursor does not modulate ACT's ability to inhibit chymotrypsin.
Results are shown in Figure 2. The data shown for each graph represent the average of four
independent assays in which 0.5 (left) or 0.6
(right) μg of ACT were incubated for 2 min ± peptide
in 10 μl .1M phosphate buffer, pH 7, at 20°C prior to the addition of .3 μg of chymotrypsin. After a further incubation of 2 min, the reaction volume was increased to 0.8 ml, 15 μl of 2 mg/ml substrate was added and the reaction followed at OD 405 for five minutes. The graphs compare the relative slopes of the reaction curves (all straight lines) normalized to the reaction rate of chymotrypsin alone. The peptides did not have any independent effect on chymotrypsin, nor did they show protease or esterase activity of their own.
EXAMPLE 2 Assessment of the Stability of
Interaction Between α-antichymotrypsin and β-protein
The ability of synthetic fragments of β-protein to inhibit ACT (Figure 2) indicates that at least a transient complex can form between the two proteins.
The stability of such complexes can be demonstrated by incubating ACT with raαiolabeled β-protein fragments under various conditions, and then analyzing the complex formation by polyacrylamide gel electrophoresis. In some experiments, a protein cross-linking agent (DSS) was used to stabilize the complex prior to gel electrophoresis. However, even in the absence of cross-linking, a complex is formed which is stable to boiling in SDS and β-mercaptoethanol. The addition of an equal molar amount of chymotrypsin to ACT blocks the active site and prevents the complex formation. Denaturing the ACT protein by heat also prevents the complex formation. Immunostaining of the blotted protein with antibodies to α1 - antichymo tryps In indicated that neither
the heat nor the chymotrypsin treatment destroyed the ACT. Thin layer chromatography confirmed that the amount of chymotrypsin added to the ACT was not sufficient to digest the peptide.