Review pubs.acs.org/acscombsci
Rationally Designed Peptides and Peptidomimetics as Inhibitors of Amyloid‑β (Aβ) Aggregation: Potential Therapeutics of Alzheimer’s Disease Deepti Goyal,* Suniba Shuaib, Sukhmani Mann, and Bhupesh Goyal* Department of Chemistry, School of Basic and Applied Sciences, Sri Guru Granth Sahib World University, Fatehgarh Sahib-140406, Punjab, India ABSTRACT: Alzheimer’s disease (AD) is a progressive neurodegenerative disease with no clinically accepted treatment to cure or halt its progression. The worldwide effort to develop peptide-based inhibitors of amyloid-β (Aβ) aggregation can be considered an unplanned combinatorial experiment. An understanding of what has been done and achieved may advance our understanding of AD pathology and the discovery of effective therapeutic agents. We review here the history of such peptide-based inhibitors, including those based on the Aβ sequence and those not derived from that sequence, containing both natural and unnatural amino acid building blocks. Peptide-based aggregation inhibitors hold significant promise for future AD therapy owing to their high selectivity, effectiveness, low toxicity, good tolerance, low accumulation in tissues, high chemical and biological diversity, possibility of rational design, and highly developed methods for analyzing their mode of action, proteolytic stability (modified peptides), and blood−brain barrier (BBB) permeability. KEYWORDS: Alzheimer’s disease (AD), amyloid aggregation, amyloid-β (Aβ) peptide, peptide inhibitors, peptidomimetics
■
INTRODUCTION It has been more than 100 years since Alzheimer’s disease (AD) was first characterized by Alois Alzheimer, a Bavarian psychiatrist, in 1907.1 However, because of the lack of effective treatment, AD2 remains epidemic in the 21st century with no clinically accepted treatment to cure or halt its progression.3 Most cases of AD are sporadic and occur in the aging population [late-onset AD (LOAD)], but in 1−2% of cases, AD presents as an autosomal dominant trait in families [familial AD (FAD) or early onset familial AD (EOFAD)]. AD is characterized by loss of short-term memory, disorientation, and impairment of judgment and reasoning. Patients lose their ability to communicate, fail to recognize their near and dear ones, and become bed-ridden at the final stages of AD. This imposes huge social and economic burdens on patients and their families.4 Approximately 47 million people worldwide are living with dementia.4a This number is expected to increase to 75 million by 2030, and 135.5 million by 2050, with much of the increase in China, India, and their South Asian and western Pacific neighbors.5 The direct cost of dementia in 2015 worldwide was estimated at US $818 billion, and the average annual cost per person with dementia is calculated to be about US $17 500. So, there is an urgent need for additional research in this area. Although the exact cause of AD is not clear, neuropathology, genetics, and transgenic modeling studies have revealed that neurofibrillary tangles and amyloid deposits in the brain (hippocampus, amygdala, and association neocortex) of the patient are the two pathological hallmark of the disease.6 Neurofibrillary tangles are aggregates of paired helical filaments composed of abnormally phosphorylated and β-folded tau © 2017 American Chemical Society
protein (the hydrophilic protein that is expressed in six human isoforms of 352 to 441 amino acid residues).7 Amyloid plaques are dense and insoluble extracellular deposits of the Aβ oligopeptide (39 to 43 amino acid residues), which is highly prone to aggregation.8 Especially important is the Aβ42 sequence, which undergoes a conformational transition from α-helix to β-sheet that exposes hydrophobic amino acid residues and promotes its aggregation to form large intermediates like protofibrils, fibrils, and plaques around neuronal cells.9 A large number of studies have been performed to predict the folding of high-energy disordered polypeptide chains to their thermodynamically stable native state that will help in the elucidation of protein misfolding mechanism.10 Aβ results from proteolysis of the amyloid precursor protein (APP) [a transmembrane integral glycoprotein in the central nervous system] (Figure 1). During the amyloidogenic pathway, β-secretase first cleaves the APP between Met671 and Asp672, followed by the action of γ-secretase to produce Aβ40 and Aβ42 in a 9:1 ratio.11 The subsequent aggregation of the Aβ42 and Aβ40 peptides to oligomers or fibrils is important for the development of the disease.12 Aβ42 is more prone to oligomerization as the C-terminus of Aβ42 is more rigid than the C-terminus of Aβ40, the latter therefore being the more abundant but less toxic form.13 These results led to the foundation of “amyloid cascade hypothesis”. According to this hypothesis accumulation of synaptotoxic and neurotoxic Aβ42 Received: July 31, 2016 Revised: December 25, 2016 Published: January 3, 2017 55
DOI: 10.1021/acscombsci.6b00116 ACS Comb. Sci. 2017, 19, 55−80
Review
ACS Combinatorial Science
molecule interactions occur over much smaller contact surfaces (about 300−1000 Å2).29 (ii) Regions of protein−protein interactions are often relatively featureless.30 (iii) Interacting protein surfaces are often highly plastic in nature.31 However, the discovery that N-terminus,32 hydrophobic core, or selfrecognition sequence (emerging as a key area for these studies),33 the hinge or turn region,34 and the C-terminus35 of Aβ peptide sequence (Figure 2) are the sites responsible for Aβ fibril formation has revived the idea that small molecules can act as inhibitors of this particular protein−protein interaction.
Figure 1. Proteolysis of APP by the amyloidogenic pathway.
oligomers cause AD.12 Oligomeric Aβ peptides interact with neurons and glial cells leading to activation of pro-inflammatory cascades such as mitochondrial dysfunction and increased oxidative stress, impairing of intracellular signaling pathway, deregulation of calcium metabolism, and finally induces neuronal apoptosis and cell death. On the basis of these studies, efforts have been made to design inhibitors that can (i) block the expression of APP, (ii) prevent the proteolytic cleavage of APP into Aβ, or (iii) clear different Aβ aggregates (monomers, oligomers, or fibrils) from the brain.14 However, none of these strategies has produced an effective clinical agent to date, for several reasons. Inhibiting γ-secretase is problematic as it acts on many physiologically essential substrates (e.g., notch signaling protein); for example, its inhibition has been found to give rise to impaired lymphocyte differentiation and altered intestinal goblet cell structure.15 β-secretase has a large pocket so its inhibition needs large molecules, which can have difficulties passing through the blood-brain barrier (BBB).16 Therefore, many compounds targeting β- or γ-secretase17 have failed at different levels of clinical trials. Tarenflurbil, semagacestat, and avagacestat are γ-secretase inhibitors, but for all, studies have been discontinued after phase II or III trials.18 The β-secretase 1 (BACE-1) inhibitor LY2886721 failed in a phase II trial because of liver toxicity.19 Controlling the production of Aβ monomers is also hazardous because these molecules are useful for neurons.20 Chelation therapies21 once attracted considerable attention, however, the only known metal chelators under clinical trials, PBT1 and PBT2 (designed to disrupt interactions between Aβ and metal), have failed.22 Therapies based on tau protein regulation23 have not yielded many therapeutically important molecules, phenothiazine methylene blue (which acts as an inhibitor of tau protein) being the only one under phase III clinical trials.24 Antibodybased therapies for the treatment of AD have not been very successful, largely due to the failure of antibodies to cross the BBB because of their large size.25 Bapineuzumab, Solanezumab, Crenezumab, and Ponezumab are monoclonal antibodies, which failed to show clinical efficacy in phase II or III clinical trials.26 An alternative disease-modifying therapy for AD involves the search for compounds that can inhibit Aβ aggregation,27 since blocking this process should not lead to mechanism-based toxicity. However, designing small molecular inhibitors that disrupt protein−protein interactions is a challenge for three main reasons. (i) Protein−protein interactions occur over large surface areas (typically 1500−3000 Å2),28 whereas protein−small
Figure 2. Subregion in Aβ42 (the regions are color coded as mentioned).
On the basis of these observations, various small molecule inhibitors36 (e.g., tramiprosate,37 inositol,38 curcumin,39 and resveratrol40) have been reported and are under different levels of clinical trials. However, a majority of these agents bind to Aβ42 with low affinity and without selectivity. Peptide-based inhibitors provide a reasonable alternative to chemical pharmaceuticals. The chief advantages of peptide inhibitors are (i) their potential for high specificity and low toxicity, (ii) the possibility of rational design, (iii) highly developed methods for analyzing their mode of action, (iv) better chance to cross the BBB, (v) low accumulation in tissues, (vi) high chemical and biological diversity, (vii) the potential to increase affinity for the target by modifications in the peptide sequence, and (viii) decreased susceptibility of modified peptides41 to proteases. These factors have contributed to the fast growth of peptides and peptide derivatives as drug candidates. At present, more than 60 therapeutic peptides are on the market, approximately 140 are in clinical trials, and at least 500 are undergoing preclinical evaluation.42 This review gathers a large amount of scientific activity in this area,43 attempting to provide the reader with primary literature over the past two decades, specific examples of important types of peptide-based inhibitors, and an appreciation of the potential advantages and future directions in AD therapy. The inhibitors covered here are divided into categories based on their relevance to the natural Aβ sequence and whether or not they contain modified amino acids (Scheme 1). To aid in appreciating the evolution of peptide-based inhibitors of Aβ aggregation, references in each section are arranged in a chronological order.
■
PEPTIDE INHIBITORS OF AMYLOID-β (Aβ) AGGREGATION Aβ Sequence Derived Peptides. Peptides Based on the Central Hydrophobic Core (CHC) Sequence [Aβ(16−20) or Aβ(16−22)]. Natural Amino-Acid-Containing Peptides. The 56
DOI: 10.1021/acscombsci.6b00116 ACS Comb. Sci. 2017, 19, 55−80
Review
ACS Combinatorial Science
partial eight residue peptides derived from Aβ42 and evaluated their ability to inhibit acid-induced aggregation of Aβ42.46 Peptides Aβ(15−22), Aβ(16−23) and Aβ(17−24) were found to be potent inhibitors of acid-induced aggregation of Aβ42. In 1996, Soto et al. synthesized a 11 residue peptide [inhibitor of Aβ fibrillogenesis (iAβ11)] which is homologous to Aβ(17− 21) and with a similar degree of hydrophobicity, but with a very low propensity to adopt β-sheet conformation.47 A molar excess of iAβ11 (RDLPFFPVRID) was required to inhibit amyloidogenesis and partially disassembled preformed Aβ fibrils in vitro. Later in 1998, Soto et al. reported a 5 residue β-sheet breaker (BSB) peptide LPFFD (iAβ5) which inhibited Aβ aggregation, disassembled preformed fibrils in vitro and prevented neuronal death.48 On injection into the rat brain model of amyloidosis iAβ5 showed reduction in Aβ protein deposition and completely blocked the formation of amyloid fibrils. In another report,49 the neuroprotective effect of the chronic intraperitoneal administration of iAβ5p (acetyl-LPFFD-amide) on the rat behavioral deficit induced by the intra hippocampal Aβ-fibrils injection was evaluated. Partial reduction of amyloid deposits and a decreased astrocytic response around the injection site was observed after one month. In 2004, Datki et al. designed a pentapeptide LPYFD-amide50 based on iAβ5 by replacing one Phe by Tyr and the C-terminalCOO− anion with CONH2. Peptide LPYFD-amide significantly decreased neurite degeneration, tau aggregation, and cell viability reduction induced by Aβ42. In another report, Eisel et al. investigated the neuroprotective properties of the pentapeptide LPYFD-amide in vitro, as well as its memory-preserving capacity against Aβ42 induced learning deficits in vivo.51 In 2014, Minicozzi et al. reported a related peptide, Ac-LPFFN-NH2, as an effective inhibitor of Aβ40 aggregation by stabilization of the native and nonaggregative α-helical conformation of Aβ40.52 Moreover, the peptide delayed fibril formation by increasing the lag phase in a fluorescence assay.
Scheme 1. Classification of Peptide Inhibitors of Aβ Aggregation Considered in the Present Review
central hydrophobic core (CHC) region of Aβ is also known as the key amyloidogenic sequence within the peptide Aβ that is responsible for aggregation. It is the “self-recognition sequence” or the “nucleation site”, that is, KLVFF.44,45 Since peptide-based fibrillogenesis is a process of self-association, a sensible approach is to design inhibitors that are somewhat homologous to the presumed site at which self-association occurs. Various CHCbased inhibitors discussed in this section are listed in Table 1. Thyberg and co-workers identified the key regions in the Aβ peptide that are important for Aβ aggregation by systematic substitution of Lys16, Leu17, and Phe20 by Ala in the KLVFF sequence and designed short peptides based on the hydrophobic core (KLVFF) of Aβ as potent inhibitors of amyloid polymerization.44 Peptide AcQKLVFFNH2 inhibited the formation of amyloid-like fibrils of the full-length Aβ. The association of KLVFF with homologous sequences in Aβ [i.e., Aβ(16−20)] lead to the formation of an antiparallel β-sheet structure which was stabilized by the interaction between Lys, Leu and Cterminal Phe as indicated by molecular modeling studies.45 In another report, Matsunaga and co-workers synthesized a series of
Table 1. Peptide Inhibitors Based on the Central Hydrophobic Core (CHC) Sequence [Aβ(16-20) or Aβ(16-22)] Containing Natural Amino Acids S. no.
sequence/name of peptides investigated
key findings
refs
2.
ALVFF, KAVFF, KLAFF, KLVAF, KLVFA, AcQKLVFFNH2 Aβ(15−22), Aβ(16−23), Aβ (17−24)
Identification of the key region in Aβ being responsible for the aggregation and synthesis of small peptide inhibitors capable of binding full length Aβ Eight residue peptides inhibited acid induced aggregation of Aβ42
3.
iAβ11 (RDLPFFPVRID)
4.
iAβ5 (LPFFD)
5.
iAβ5p (Ac-LPFFD-amide)
6.
LPYFD-amide
7. 8.
LPYFD-amide Ac-LPFFN-NH2
9.
KLVFFK, KLVFFKK, KLVFFKKK, KLVFFKKKK, KLVFFKKKKK, KLVFFEEEE, KLVFFSSSS cyclo(17, 21)-[Lys17, Asp21]Aβ(1−28)
Peptide has a very low propensity to adopt β-sheet conformation and inhibited amyloidogenesis and partially disassembled preformed Aβ fibrils in vitro β-Sheet breaker (BSB) peptide disassembled preformed fibrils in vitro and prevented neuronal death Chronic intraperitoneal administration of iAβ5p in the rat model showed partial reduction of amyloid deposits formed and a decreased astrocytic response around the injection site Peptide significantly decreased neurite degeneration, tau aggregation, and cell viability reduction induced by Aβ42 Peptide could preserve memory by reverting Aβ42 oligomer-induced learning deficits Peptide stabilizes the native and nonaggregative α-helical conformation of Aβ40 and also delayed the fibril formation by increasing the lag phase in fluorescence assay Peptides accelerated Aβ aggregation protects against Aβ toxicity by accelerating association of soluble oligomers into fully formed fibrils and thus reducing the concentration of putative toxic intermediates Cyclic peptide showed no aggregation or toxicity in vitro and inhibited Aβ(1−28) and Aβ40 aggregation and toxicity by disrupting the amyloid fibrils into short fibrils and amorphous aggregates These were effective inhibitors of the Aβ fibril formation; however, only OR2 inhibited Aβ oligomer formation RR effectively inhibited aggregation of Aβ40 via multiple hydrophobic and electrostatic interactions and hydrogen bonding by casing three important subregions of Aβ40
Thyberg et al.44,45 Matsunaga et al.46 Soto et al.47
1.
10. 11. 12.
OR1 (RGKLVFFGR), OR2 (RGKLVFFGR−NH2) RR (Ac-RYYAAFFARR-NH2)
57
Soto et al.48 Inestrosa, Soto et al.49 Datki et al.50 Eisel et al.51 Minicozzi et al.52 Murphy et al.53 Kapurniotu et al.54 El-Agnaf et al.55 Yuan et al.56
DOI: 10.1021/acscombsci.6b00116 ACS Comb. Sci. 2017, 19, 55−80
Review
ACS Combinatorial Science
Table 2. Peptide Inhibitors Based on the Central Hydrophobic Core (CHC) Sequence [Aβ(16-20) or Aβ(16-22)], Containing Modified Amino Acids S. no.
sequence/name of peptides or peptidomimetics investigated
1.
PPI-368 [cholyl-(LVFFA)−OH], PPI-433 [cholyl-(lvffa)−OH], PPI457 [cholyl-(lvffa)-NH2] 125 I-YiAβ11, 125I-PUT-YiAβ11, 125IPUT-D-YiAβ11 NH2-K(Me-L)V (Me-F)F(Me-A) ECONH2 inL, inD, and inrD
2. 3. 4. 5. 6.
PPI-1019 {D-(H-[(Me-L)-VFFL]NH2)} klvffa, kklvffa, kfvffa, kivffa, kvvffa
7.
pgklvya, kklvffarrrra, kklvffa
8. 9.
six copy conjugate of kffvlk X0-X6, DDX1, DDX2, DDX3
10. 11.
peptide 1 KLVFFKKKKKK hybrids
12.
SEN304 {D-[(chGly)-(Tyr)-(chGly)(chGly)-(mLeu)]-NH2} AMY-1, AMY-2
13. 14. 15. 16.
K4 peptide 2 (KLVF-ΔA-I-ΔA), peptide 3 (KF-ΔA-ΔA-ΔA-F) conjugate 4-5
17. 18. 19.
peptide 6−11 peptide 12−15 iAβ5-PEG3200, iAβ5-PEG5000
20.
Th-NT, Th-CT, Th-SC
21.
rGffvlkGr-NH2, rGffvlkGr-1,5diaminopentane RI-OR2 (Ac- rGffvlkGr-NH2), RIOR2-TAT (AcrGffvlkGrrrrqrrkkrGy-NH2) peptides 16−21
22. 23. 24. 25. 26. 27. 28. 29. 30. 31.
Fc-KLVFF Fc-KLVFFK6 LPfFFD-PEG, LVFfFD-PEG, LVfFFD-PEG cyclic peptides 22−25 poly(LVFF-co-β-amino ester) copolymer P2, P3, P4, P5, P6 LK7 (Ac-LVFFARK-NH2), LK7@ PLGA-NPs AuNPs@POMD-pep
key findings
refs
PPI-433 and PPI-457 retained inhibitory activity like PPI-368 and were both stable in monkey cerebrospinal fluid for 24 h
Findeis et al.57
125
Poduslo et al.58
I-PUT-D-YiAβ11 partially inhibited Aβ-fibril formation and showed enhanced BBB permeability and proteolytic degradation N-methylated peptides inhibited fibril formation by binding to the growth site of Aβ nuclei and prevented the propagation of network of hydrogen bonds Both inrD and inD inhibited Aβ42 aggregation more effectively than inL, inrD decreases Aβ42 cytotoxicity to a greater extent than inL and inD, also both inD and inrD are stable to proteases N-methylated peptide inhibited Aβ aggregation and toxicity; The D-amino acid peptides show better inhibitory activity against L-Aβ and L-KLVFF act as a more potent inhibitor of D-Aβ aggregation Two D-peptides (pgklvya and kklvffa) were found capable of improving survival in the transgenic C. elegans Six-copy conjugate of kffvlk showed 10000-fold greater affinity than the monomer peptide ffvlk Because of hydrophilic units attached to the KLVFF sequence, they show suppression of toxicity of Aβ and DDX3 was found to be the most effective inhibitor of Aβ aggregation The ester bonds in peptide 1 disrupted the hydrogen bonds and inhibited the amyloid fibril formation Hybrid peptides increased the solvent surface tension and greatly enhanced the Aβ-aggregation leading to the formation of nontoxic oligomers Peptide SEN304 inhibited Aβ aggregation by binding directly to Aβ42, delayed β-sheet formation and promotes aggregation of toxic oligomers into a nontoxic form Designed peptides bind strongly to fibrils and simultaneously inhibited further assembly by providing a blocking face The better inhibition shown by K4 was accredited to its multivalent nature Peptides 2 and 3 were protease stable and inhibited aggregation of Aβ to different extents Conjugates 4 and 5 were proved to be more effective than curcumin in preventing the Aβ fibril formation Peptides 6, 10 and 11 were potent inhibitors of Aβ40 aggregation in comparison to iAβ5p The peptides 12 and 13 were the better inhibitors due to the presence of sulfonamide group Peptides showed improved stability in the bloodstream as compare to the parent peptide iAβ5, without affecting fibrillogenesis inhibition activity Trehalose-conjugated peptides showed the inhibitor effected at the nucleation phase of Aβ42 aggregation process Peptides have increased efficacy in the inhibition of aggregation and toxicity of Aβ RI-OR2-TAT showed enhanced binding affinity toward Aβ42 amyloid fibrils owing to the presence of positively charged amino acid residues on the TAT portion; RI-OR2 is highly resistant to proteolysis and stable in human plasma and brain extracts N-methylation at every second amino acid in short peptide sequences (
