A Systematic Review of Antiamyloidogenic and Metal-Chelating Peptoids: Two Structural Motifs for the Treatment of Alzheimer’s Disease

Alzheimer’s disease (AD) is an incurable form of dementia affecting millions of people worldwide and costing billions of dollars in health care-related payments, making the discovery of a cure a top health, societal, and economic priority. Peptide-based drugs and immunotherapies targeting AD-associated beta-amyloid (Aβ) aggregation have been extensively explored; however, their therapeutic potential is limited by unfavorable pharmacokinetic (PK) properties. Peptoids (N-substituted glycine oligomers) are a promising class of peptidomimetics with highly tunable secondary structures and enhanced stabilities and membrane permeabilities. In this review, the biological activities, structures, and physicochemical properties for several amyloid-targeting peptoids will be described. In addition, metal-chelating peptoids with the potential to treat AD will be discussed since there are connections between the dysregulation of certain metals and the amyloid pathway.


Introduction
Alzheimer's disease (AD) is an incurable form of dementia afflicting an estimated 5.5 million Americans [1]. By 2050, AD will affect an estimated 13.8 million patients [1] and cost the United States as much as $1.1 trillion [2], representing a growing health, societal, and economic concern. Four of the five AD drugs currently on the market (donepezil, rivastigmine, galantamine, and a donepezil-memantine combination drug) inhibit the enzyme acetylcholinesterase (AChE) in order to mitigate cholinergic dysfunction and subsequent neurodegeneration and memory decline [3,4]. Memantine targets the N-methyl-D-aspartate (NMDA) receptor, which, at increased levels, can lead to deterioration in cognitive function [5]. Since marketed AD drugs treat symptoms rather than the underlying disease pathology, there is a dire need for new AD drugs with different mechanisms of action.
From 2002 to 2012, 99.6% of AD drugs that entered clinical trials failed, which is a higher rate of failure than either oncology or cardiovascular drugs [6,7]. Recent high-profile clinical trial failures have slowed progress toward a cure and cast doubt on one of the most widely studied hypotheses for AD pathology, the amyloid hypothesis [8,9]. Challenges surrounding AD drug discovery include mis-diagnosis of clinical trial patients [10][11][12], a lack of reliable biomarkers [10], and an incomplete understanding of the complex and multifaceted disease pathology involving tau hyper-phosphorylation, Aβ aggregation, cholinergic dysfunction, metal dysregulation, inflammation, and oxidative stress [3,6,13,14], which are often interconnected. In addition, AD drugs must penetrate the blood-brain barrier (BBB), further limiting the number of viable therapeutic options [15].
In recent years, peptoids (N-substituted glycine oligomers; [16]) have emerged as a promising class of peptidomimetics [17] with a range of properties, including anti-cancer [18] and antibacterial must penetrate the blood-brain barrier (BBB), further limiting the number of viable therapeutic options [15].
In recent years, peptoids (N-substituted glycine oligomers; [16]) have emerged as a promising class of peptidomimetics [17] with a range of properties, including anti-cancer [18] and antibacterial activities [19,20]. Peptoids have several advantages over peptides as drug candidates, including their ease of synthesis via either solid-phase synthesis (SPS; [16,21,22]) or solution-phase methods [23] and their nonimmunogenic nature [24]. Since peptoids are synthesized from primary amine building blocks, of which thousands are commercially available, tremendous side-chain diversity can be readily achieved [25], and their secondary structures are tunable [20]. Furthermore, compared to peptide-based drugs and other biologics, linear and/or cyclic peptoids have more favorable pharmacokinetic (PK) properties, such as longer biological half-lives, due to their unnatural structures, and higher membrane permeabilities, due to a limited number of hydrogen-bond donors (HBDs; Figure 1) [26][27][28][29]. Together, these properties make peptoids viable drug candidates for any disease, but especially for AD and other CNS diseases in which BBB permeability is critical to drug success. This review presents the structures, methods used, biological activities, and physicochemical properties for various antiamyloidogenic peptoids. In addition, several metal-chelating peptoids are presented that could prove useful against AD, since there are connections between the dysregulation of certain metal ions and the amyloid pathway [6,30] (vide infra).
The medicinal properties of peptide and peptoid foldamers, including antibacterial properties and their abilities to disrupt protein-protein interactions, have been reviewed [19,20]. The potential of peptoids as AD diagnostic agents, including for the detection of Aβ1-42, has been reviewed by Kodadek and Vanderstichele [31] and will not be discussed here.

2.1.Overview
The amyloid hypothesis, first reported in 1984 [32,33], states that accumulation of Aβ peptide (specifically the more neurotoxic Aβ1-42 peptide) oligomers, aggregates, and fibrils plays a role in AD and can trigger a host of effects, including neuroinflammation, taupathy, and neuronal cell death [34]. This hypothesis has replaced the cholinergic hypothesis as the predominant foundation for the development of new AD therapies. Emerging evidence suggests soluble aggregates rather than fibrils are responsible for AD-associated neurotoxicity, which has led to a shift in approach in anti-amyloid drug design [35,36]. Although the importance of the amyloid hypothesis in AD drug discovery has been questioned [8,9], Aβ, and its associated genes, proteins, and pathways, remains the most extensively studied and pursued AD target [34].
Several peptide-based drugs have been explored to target AD-associated Aβ aggregation [37]. Due to their low BBB penetrabilities, short half-lives, and low bioavailabilities, many of these This review presents the structures, methods used, biological activities, and physicochemical properties for various antiamyloidogenic peptoids. In addition, several metal-chelating peptoids are presented that could prove useful against AD, since there are connections between the dysregulation of certain metal ions and the amyloid pathway [6,30] (vide infra).
The medicinal properties of peptide and peptoid foldamers, including antibacterial properties and their abilities to disrupt protein-protein interactions, have been reviewed [19,20]. The potential of peptoids as AD diagnostic agents, including for the detection of Aβ 1-42 , has been reviewed by Kodadek and Vanderstichele [31] and will not be discussed here.

Overview
The amyloid hypothesis, first reported in 1984 [32,33], states that accumulation of Aβ peptide (specifically the more neurotoxic Aβ 1-42 peptide) oligomers, aggregates, and fibrils plays a role in AD and can trigger a host of effects, including neuroinflammation, taupathy, and neuronal cell death [34]. This hypothesis has replaced the cholinergic hypothesis as the predominant foundation for the development of new AD therapies. Emerging evidence suggests soluble aggregates rather than fibrils are responsible for AD-associated neurotoxicity, which has led to a shift in approach in anti-amyloid drug design [35,36]. Although the importance of the amyloid hypothesis in AD drug discovery has been questioned [8,9], Aβ, and its associated genes, proteins, and pathways, remains the most extensively studied and pursued AD target [34].
Several peptide-based drugs have been explored to target AD-associated Aβ aggregation [37]. Due to their low BBB penetrabilities, short half-lives, and low bioavailabilities, many of these compounds are not ideal clinical candidates. Unfortunately, in some cases, attempts to improve bioavailability and stability resulted in decreased inhibition of Aβ aggregation [38]. Amyloid immunotherapies have shown promise, and some have reached clinical trials, but, to date, none have reached the market, in some cases due to adverse effects [39]. Immunotherapies do not penetrate the BBB due to their large size, polarity, and large number of HBDs and hydrogen-bond acceptors (HBAs), further limiting their therapeutic potential [15]. Considering the limitations of current amyloid-targeting therapies, peptoids, which have more favorable PK properties, represent a promising therapeutic class. Section 2 presents various antiamyloidogenic peptoids. Peptoid structures are presented in Figure 2, methods and reported biological activities are in Table 1, and physicochemical properties are summarized in Table 2.

Peptoid-Based Aβ 1-42 -Aggregation Inhibitors
Yang et al. recently discovered the nanomolar Aβ 1-42 inhibitor amyloid inhibitory peptoid 1 (AIP1; Figure 2). AIP1 increased the lag time to form the critical nucleus for fibril formation and demonstrated in vitro BBB permeability in line with that of previously reported Aβ-aggregation inhibitors ( Table 1) [40]. In addition to providing a viable lead for AD, this work presents an approach for the discovery of additional peptoid-based Aβ 1-42 aggregation inhibitors.
Bezprozvanny and coworkers reported monomer inhibitor of amyloid 1 (IAM1; Figure 2) that selectively suppressed Aβ 1-42 aggregation and could be used to detect Aβ 1-42 [41,42]. The authors also investigated the dimer (IAM1)2, which had an even higher Aβ 1-42 affinity than the monomer and demonstrated neuroprotective properties ( Table 1). The peptoid building blocks in this study were chosen to increase structural diversity and water solubility yet allow for Aβ 1-42 -peptoid interactions and suitable BBB penetrability.

Peptoid-Based Aβ 1-40 -Aggregation Inhibitors
Servoss et al. designed peptoids to mimic residues 16-20 of Aβ 1-40 (KLVFF) [43,44], the hydrophobic region responsible for aggregation [45,46]. One peptoid in particular, JPT1 (Figure 2), adopts a stable polyproline type-I (PPI)-like helical structure, which enhances interactions with Aβ and suppresses peptide aggregation through pi-pi stacking interactions between the peptoid and the phenylalanine residues of the Aβ 1-40 section. This peptoid suppressed Aβ 1-40 aggregation in a dose-dependent manner, decreased fibril formation, and decreased the lag time to form fibrils (Table 1). A follow-up study investigated the impact of side-chain chirality and placement on Aβ 1-40 -aggregation suppression [47]. Interestingly, by removing the aromatic side chains or moving the chiral aromatic side chains, inhibition of Aβ aggregation was still maintained; however, the morphology and size of the aggregates changed compared to the original JPT1 analog. The authors proposed that off-pathway aggregates were formed in order to surpass/suppress fibril formation. These studies have enhanced the understanding of Aβ-small molecule interactions and could allow for more effective design of new amyloid therapies. Although the BBB permeabilities of JPT1 and related analogs were not assessed, the lipophilic nature of these Aβ-aggregation inhibitors and the limited number of HBDs could result in suitable BBB permeabilities, although the molecular weights (~1000 Da; Table 2) are outside of the typical range for CNS drugs [48].
Kirshenbaum and Sadowski reported a large library of stable peptide/peptoid hybrids that suppress Aβ 1-40 fibrillization and oligomerization in vitro and in vivo and mitigate Aβ-mediated NMDA receptor loss (Table 1); these molecules could also be used in AD diagnosis [49]. The approach involved replacing specific amino acid residues of Aβ 1-42 (or sections thereof) with peptoids or other unnatural amino acids in order to improve PK properties. These peptoids have properties consistent with BBB-permeable molecules (hydrophobic nature, limited number of HBDs, and high plasma stability), making them viable therapeutic candidates. Furthermore, these molecules exhibited suitable BBB penetrabilities, as evidenced by detection in cerebrospinal fluid after intravenous (iv) administration [49]. Structures of antiamyloidogenic peptoids AIP1 [40], IAM1 [41,42], JPT1 [43,44,47], and CPO_Aβ17-21P [50]. Protonation states are the same as those shown in the original references.

Peptoids Targeting the Apolipoprotein E4-Aβ Interaction
Apolipoprotein E4 (apoE4), a protein that is a genetic risk factor in late-stage AD, plays a role in Aβ accumulation, plaque formation, and tau pathology. ApoE/apoE4 can interact with specific residues of Aβ to promote aggregation; the apoE4 allele also decreases Aβ brain clearance [34]. As such, the Aβ-apoE interaction has been explored as a target for novel AD therapies. Kwon, Wisniewski, and coworkers discovered CPO_Aβ17-21P ( Figure 2) from a small library of cyclic and linear peptoids; this peptoid inhibited apoE4-mediated Aβ aggregation, had a favorable safety profile, improved cognitive function in a transgenic mouse model, and did not trigger a neuroinflammatory response (Table 1) [50]. Furthermore, the fairly low molecular weight (~700 Da; Table 2), hydrophobicity, and cyclic nature of this compound [27,28] could result in favorable PK properties. The BBB permeability of CPO_Aβ17-21P was not explicitly explored, although in vivo efficacy suggests suitable permeability [50].
Aβ1-42 aggregation inhibition could be achieved by binding to the N-terminus of Aβ1-42 [40], binding to the hydrophobic β-turn-β-sheet region (residues 18-42) of the peptide, or interactions with the C-terminus [42]. Bezprozvanny et al. proposed that the Aβ1-42 selectivity of IAM1 was due to interactions between the hydrophobic isoleucine and alanine residues at the C-terminus of the peptide, potentially due to the increased rigidity of this region of Aβ1-42 compared to Aβ1-40 [42,51]. According to docking studies, the Aβ1-42-binding activity of AIP1 could be due to π-π stacking and hydrogen-bonding interactions between the peptoid and Aβ1-42 and the peptoid binds to the N-terminus of Aβ1-42 [40].
Modulation of Aβ1-40 aggregation can occur through π-π stacking interactions between the KLVFF portion of the peptide and aromatic peptoid residues, which may be responsible for the activity of JPT1 [44,47]. Since JPT1 was not assessed as an Aβ1-42 inhibitor, it is unclear whether the peptoid would exhibit Aβ1-40-selectivity or dual inhibition. The amyloid-targeting peptoids herein exhibit a range of structures, selectivities, and activities to expand the arsenal of potential AD drugs.  [40], IAM1 [41,42], JPT1 [43,44,47], and CPO_Aβ17-21P [50]. Protonation states are the same as those shown in the original references.

Peptoids Targeting the Apolipoprotein E4-Aβ Interaction
Apolipoprotein E4 (apoE4), a protein that is a genetic risk factor in late-stage AD, plays a role in Aβ accumulation, plaque formation, and tau pathology. ApoE/apoE4 can interact with specific residues of Aβ to promote aggregation; the apoE4 allele also decreases Aβ brain clearance [34]. As such, the Aβ-apoE interaction has been explored as a target for novel AD therapies. Kwon, Wisniewski, and coworkers discovered CPO_Aβ17-21P ( Figure 2) from a small library of cyclic and linear peptoids; this peptoid inhibited apoE4-mediated Aβ aggregation, had a favorable safety profile, improved cognitive function in a transgenic mouse model, and did not trigger a neuroinflammatory response (Table 1) [50]. Furthermore, the fairly low molecular weight (~700 Da; Table 2), hydrophobicity, and cyclic nature of this compound [27,28] could result in favorable PK properties. The BBB permeability of CPO_Aβ17-21P was not explicitly explored, although in vivo efficacy suggests suitable permeability [50].
Aβ 1-42 aggregation inhibition could be achieved by binding to the N-terminus of Aβ 1-42 [40], binding to the hydrophobic β-turn-β-sheet region (residues 18-42) of the peptide, or interactions with the C-terminus [42]. Bezprozvanny et al. proposed that the Aβ 1-42 selectivity of IAM1 was due to interactions between the hydrophobic isoleucine and alanine residues at the C-terminus of the peptide, potentially due to the increased rigidity of this region of Aβ 1-42 compared to Aβ 1-40 [42,51]. According to docking studies, the Aβ 1-42 -binding activity of AIP1 could be due to π-π stacking and hydrogen-bonding interactions between the peptoid and Aβ 1-42 and the peptoid binds to the N-terminus of Aβ 1-42 [40].
Modulation of Aβ 1-40 aggregation can occur through π-π stacking interactions between the KLVFF portion of the peptide and aromatic peptoid residues, which may be responsible for the activity of JPT1 [44,47]. Since JPT1 was not assessed as an Aβ 1-42 inhibitor, it is unclear whether the peptoid would exhibit Aβ 1-40 -selectivity or dual inhibition. The amyloid-targeting peptoids herein exhibit a range of structures, selectivities, and activities to expand the arsenal of potential AD drugs. Table 1. Methods used and bioactivities reported for example antiamyloidogenic peptoids.

Name/Description
Methods Used Bioactivities References

Connections between Metal Dysregulation, the Amyloid Pathway, and AD Pathology
The dysregulation of Fe, Cu, and Zn plays a documented role in AD, and this dysregulation could be linked to other AD pathologies, including Aβ aggregation and tau hyperphosphorylation [6]. Amyloid plaques contain abnormally high levels of these three metals, and Cu-amyloid complexes have been linked to reactive oxygen species (ROS) generation, mitochondrial dysfunction, and neuronal cell death [6]. In addition, Fe-amyloid complexes can trigger oxidative stress, and an imbalance of all three metal ions can interfere with Aβ degradation [30]. Despite the role of these metals in AD, metal chelators are rarely explored as AD therapies. In addition to addressing metal dysregulation, these drugs could feasibly mediate AD-associated Aβ aggregation, ROS generation, inflammation, and oxidative stress. Developing a relatively nontoxic drug that is selective for the metal ions of interest, can extract metal ions from amyloid plaques, and can penetrate the BBB is a formidable challenge [6]. In Section 3, various metal-chelating peptoids are presented, with a focus on peptoids that bind to AD-associated metals. The structures of these peptoids are presented in Figure 3, methods used and reported biological activities are in Table 3, and physicochemical properties are summarized in Table 4.
Of particular interest are the selective Cu(II)-binding peptoids since this metal plays an important role in AD [6]. For example, helix HQT i + 3 ( Figure 3) binds Cu(II) selectively when in a mixture of other metal ions such as Zn(II), Fe(III), Co(II), Ni(II), and Mn(II) [58]. This peptoid can also simultaneously bind two different metal ions, such as Cu and Zn, further enhancing the abilities of these molecules to mimic the complex processes of a natural protein. Metal-binding studies revealed that the selectivity and cooperativity are from a combination of helicity and pre-organization of the ligands. The peptoid could also adjust its binding mode (intramolecular vs intermolecular) in response to changes in metal ion concentrations, for example [58].
Varma and coworkers reported a calix [4]arene peptoid tetramer that binds Cu(II) selectively in the presence of various other metal cations (as perchlorates), including Cr(III), Mn(II), Fe(II), Co(II), Cu(II), Zn(II), Hg(II), Ag(I), Pb(II), Cd(II), Na, K, Ca, and Al(III) ( Table 3; Figure 3) [59]. One of the peptoids was reported to have high selectivity for Cu(II) based on no change in the observed UV band in the presence of other metal ions. The numbers of HBDs and HBAs as well as the molecular weights of these calix [4]arene derivatives, which range from~1600 to 2000 Da, could limit the potential of these molecules as CNS drugs (Table 4). Table 3. Methods used and bioactivities reported for example metal-chelating peptoids.
Some peptoids exhibited Zn-binding affinity at least one order of magnitude higher than that of various metal ions. [60] Benzyloxyethyl-based peptoids free and immobilized on a chitosan film with antioxidant and Fe-chelating properties 2,2-Diphenyl-1-picrylhydrazyl (DPPH) and hydroxyl radical procedures were used to assess antioxidant activities; Fe-chelating properties investigated using an EDTA-competition assay.
Two of the free peptoids exhibited concentrationdependent radical scavenging effect of up to 80% at 5 g/L. [61] Screening for selective Ni(II)-binding peptoids SPS used to synthesize peptoids; high-throughput, bench-top X-ray fluorescence screening (with ICP-MS and a colorimetric assay for validation) used to screen peptoids for Ni(II)-binding.
Two of the peptoids bind Ni(II) in the presence of other metal ions. [62] Screening for iron-and copper-binding peptoids Split-pool SPS used to synthesize peptoid library; 19 F NMR spectroscopy used to screen for metal-binding properties. Results were validated using UV titration.

Zn(II)-Binding Peptoids
Zuckermann and coworkers designed heat-resistant two-helix bundles for tight Zn binding to mimic nature [60]. In the presence of the metal ions Mg(II), Mn(II), Ca(II), Ni(II), Co(II), Cu(II), and Cd(II), one of the peptoids exhibited a Zn(II)-binding affinity at least one order of magnitude higher than that of the other metal ions. The Zn affinities for some of the peptoids reported approach that of a Zn-binding protein [60]. This work provides a platform for future screening of peptoid-based metal-binders. The molecular weights (>4000 Da), hydrophilicities, and number of HBDs of these molecules could preclude BBB penetrability, however (Table 4).

Fe(II)-Binding Peptoids
Taillefumier and coworkers reported a series of free and chitosan-bound benzyloxyethyl-based peptoids with both antioxidant and iron-chelating properties for use in food products [61]. Interestingly, binding the peptoids to the chitosan surface resulted in an enhancement of the antioxidant activity of the chitosan itself, although the free peptoids also exhibited some antioxidant properties (Table 3; [61]). Considering the predominant medical use of chitosan films as implants [65,66], the unbound peptoids reported in this study will likely be more useful for AD therapy.

Screening Approaches for the Discovery of Metal-Binding Peptoids
With the capabilities to rapidly generate large compound libraries, high-throughput screening methods to search for peptoids with biologically relevant activities, including metal-binding capabilities, are essential for the discovery of peptoid-based therapies. Kirshenbaum et al. recently reported a bench-top X-ray fluorescence method to screen for metal-binding properties [62], which was validated using a colorimetric assay and ICP-MS. Using this assay, the detection of emission lines can determine metal-binding events. This method requires very small sample sizes and has pmol detection limits, making it an excellent candidate for on-bead analysis of peptoids generated via SPS. Furthermore, the analysis time is on the order of seconds, making this a high-throughput technology. The side chains of these peptoids mimicked histidine, glutamic acid, serine, tyrosine, and tryptophan residues to allow for metal binding, with some non-natural side chains, including pyridylalanine. Metal binding depended on the sequence and structure of the peptoid [62]. Although the authors specifically used this technology to screen for selective Ni(II)-binding peptoids, this approach could be used to search for peptoids that bind metal ions more prominently linked to AD pathology, viz Cu, Fe, and Zn. 19 F NMR spectroscopy, with aryl fluorides as tags, was used to screen a library of 90 peptoids for Fe-and Cu-binding properties [63,64]. Primary amines containing hydroxyl, imidazole, and carboxyl groups were used as building blocks to attempt to mimic the active site of serine proteases [64]. The metal-binding properties of hits were further investigated using UV-vis titration (Table 3). Selective Cu(II)-and Fe(III)-binding peptoids, some with K D values in the micromolar range, were discovered from this screening [64]. Advantages of this approach include the wide availability of the instrumentation and the high sensitivity and large chemical shift dispersion of 19 F NMR spectroscopy.

Comparative Structure-Activity Relationship for Metal-Chelating Peptoids
As shown in Figure 3, common peptoid ligands for coordination to metals such as Fe and Cu include HQ (e.g., 7mer-HQ2), terpyridine (e.g., Helix HQT i + 3), and pyridine derivatives (e.g., PentA) as well as residues to mimic natural amino acids such as histidine or serine, with HQ ligands being the most prevalent among the Cu-binders [55,58]. In addition, chiral aromatic and aliphatic residues were often incorporated to induce helicity, an important property linked to the selectivities observed for some of the metal-binders. Hydrophilic Nsmp groups also promoted water solubility, which could result in more favorable PK properties.

Advantages of Metal-Chelating Peptoids
Many of these peptoids selectively chelate metals and have modifiable secondary structures and PK properties, thereby expanding their therapeutic potential. Furthermore, the ease of peptoid synthesis via the submonomer approach [16,21,22], the tremendous structural diversity possible, and the water solubilities of some of these peptoids, including Maayan's Nsmp-containing peptoids [55], provide additional advantages. In order to explore the full potential of these molecules as AD therapeutics, their stabilities and BBB penetrabilities should be explored. The metal-chelating peptoids herein have a reasonable balance of hydrophobic and hydrophilic groups; large molecular weights and large numbers of HBDs and HBAs could limit the potential of some of these molecules as CNS drugs (Table 4). Ultimately, the expanded knowledge of peptoid sequence-structure-function relationships will allow for more effective design of peptoid-based metal-chelators for specific applications, such as for the treatment of AD.
Considering the complex and polyetiological nature of AD, the one-target one-drug paradigm seems insufficient [67], and multifunctional drugs hold promise. Since tremendous structural diversity can be achieved with peptoids, these molecules can be readily modified to incorporate multiple functionalities. For example, a metal-chelating group could be incorporated into an Aβ-aggregation inhibitor, further expanding the potential of peptoid-based AD drugs.
Out of the 105 potential AD drugs currently in the pipeline [68], none of them are peptoids. With additional advances in the field of peptoids, this should soon change. The biggest hurdles that lie ahead include establishing full PK and in vivo activity profiles, including brain bioavailabilities, for AD-targeting peptoids, continuing to explore the impact of peptoid structure on biological function, and further studying the impact of peptoid sequence on secondary and tertiary structures [69,70]. As the field of biomimetic foldamers continues to advance [71], so will peptoid-based drug design.