Novel Amylin Analogues Reduce Amyloid-β Cross-Seeding Aggregation and Neurotoxicity

Abstract

Background:

Type 2 diabetes related human islet amyloid polypeptide (hIAPP) plays a dual role in Alzheimer’s disease (AD). hIAPP has neuroprotective effects in AD mouse models whereas, high hIAPP concentrations can promote co-aggregation with amyloid-β (Aβ) to promote neurodegeneration. In fact, both low and high plasma hIAPP concentration has been associated with AD. Therefore, non-aggregating hIAPP analogues have garnered interest as a treatment for AD. The aromatic amino acids F23 and I26 in hIAPP have been identified as the key residues involved in self-aggregation and Aβ cross-seeding.

Objective:

Three novel IAPP analogues with single and double alanine mutations (A1 = F23, A2 = I26, and A3 = F23 + I26) were assessed for their ability to aggregate, modulate Aβ oligomer formation, and alter neurotoxicity.

Methods:

A range of biophysical methods including Thioflavin-T, gel electrophoresis, photo-crosslinking, circular dichroism combined with cell viability assays were utilized to assess protein aggregation and toxicity.

Results:

All IAPP analogues showed significantly less self-aggregation than hIAPP. Co-aggregated Aβ₄₂-A2 and A3 also showed reduced aggregation compared to Aβ₄₂-hIAPP mixtures. Self- and co-oligomerized A1, A2, and A3 exhibited random coil conformations with reduced beta sheet content compared to hIAPP and Aβ₄₂-hIAPP aggregates. A1 was toxic at high concentrations compared to A2 and A3. However, co-aggregated Aβ₄₂-A1, A2, or A3 showed reduced neurotoxicity compared to Aβ₄₂, hIAPP, and Aβ₄₂-hIAPP aggregates.

Conclusion:

These findings confirm that hIAPP analogues with non-aromatic residues at positions 23 and 26 have reduced self-aggregation and the ability to neutralize Aβ₄₂ toxicity. This warrants further characterization of their protective effects in pre-clinical AD models.

Keywords

Aggregation Alzheimer’s disease amylin analogue amyloid-beta human islet amyloid polypeptide (hIAPP)

INTRODUCTION

Dementia is the leading cause of disability in the ageing population [1, 2]. Dementia is an umbrella term for a number of neurological conditions, of which the major symptom includes a generalized decline in brain function [1]. There are over 100 diseases that cause dementia [1]. The most common causes of dementia include AD, vascular dementia, and dementia with Lewy bodies [1]. AD is the most prevalent form of dementia and currently affects around 30–35 million people worldwide [3, 4]. It is a chronic neurodegenerative disorder characterized by senile amyloid plaques, inflammation, oxidative stress, and extensive cell loss in the brain [1, 2]. Similar to AD, type 2 diabetes (T2D) is another chronic metabolic disorder that currently affects around 380 million people worldwide [5]. Multiple clinical and epidemiological studies have shown that there is a strong link between AD and T2D, indicating that in comparison to non-diabetic individuals, patients with T2D are more prone to developing AD [6 –8].

AD and T2D share many pathological features including amyloidogenic protein accumulation, oxi-dative stress, inflammation, and metabolic dysfunction [9]. Amyloidogenic protein deposition is one of the main pathogenic markers observed in both AD and T2D patients [7, 10]. The term ‘amyloid’ is generally associated with disease and refers to any protein that has the capacity to misfold and aggregate into oligomeric intermediates and form insoluble fibrils that deposit in the affected tissues and organs [11]. The main amyloidogenic protein in the senile plaques found in the AD brain is Aβ [2, 12], whereas the amyloidogenic protein that deposits in the pancreas of T2D patients is hIAPP [9]. Neuronal cell death in the brain of AD patients and beta-cell death in the pancreas of T2D patients is associated with the deposition of aggregated forms of these amyloidogenic proteins, Aβ and hIAPP, respectively [13, 14].

Aβ is a 39–43 amino acid (AA) long peptide generated by the sequential proteolytic cleavage of amyloid-β protein precursor (AβPP) by beta-site amyloid precursor protein cleaving enzyme 1 (BACE1) at the N-terminus, followed by γ-secretase at the C-terminus [12, 15]. The normal physiological function of Aβ is not fully understood but current evidence suggests that it is responsible for controlling synaptic activity, glutamate release, and anti-microbial activity [16]. Although Aβ₄₀ is known to be the most abundant form of Aβ, the 42 AA long form (Aβ₄₂) with the two additional hydrophobic residues (I41 and A42) at the C-terminus has increased propensity to aggregate and is known to be the most neurotoxic form and the main protein component in the senile plaques observed in the AD brain [17 –19]. It is also suggested that AD brain features increased accumulation of longer forms such as Aβ₄₂ and Aβ₄₃, whereas cognitively intact individuals tend to generate more Aβ₃₇, Aβ₃₈, and Aβ₄₀ [86]. Aβ can mediate neurotoxicity through multiple mechanisms and some of the prominent pathways may include disruption of neuronal receptor activity and ion channels causing ion dysregulation, mitochondrial dysfunction, and oxidative stress eventually leading to cell death [20 –24]. Aβ can be present in multiple structural isoforms and its soluble oligomers are known to be the most neurotoxic, closely correlating with the clinical symptoms of AD [25].

The pancreatic amyloid protein hIAPP is generated by the proteolytic cleavage of pre-pro IAPP (89 AA) that generates pro-hIAPP (67 AA), which is then further cleaved to a 37 AA peptide [26]. Unlike Aβ, hIAPP has well-established physiological roles and is known to regulate post-prandial glucose levels, gastric emptying, and leptin signaling in the brain [8]. hIAPP is co-secreted with insulin at a 1:20 ratio [8]. Along with hyperinsulinemia, hyperamylinemia is a characteristic feature of early T2D [8]. This is considered to be one of the main causes of hIAPP aggregation, deposition, and beta cell death in the pancreas [8]. Similar to Aβ, hIAPP induces cell death by similar mechanisms including inflammation, oxidative stress, and mitochondrial dysfunction [8]. More importantly, hIAPP is known to play a central role, both as a pathogenic risk factor and as a treatment strategy in AD and T2D.

hIAPP is highly amyloidogenic and is similar to Aβ in that they share 50% sequence similarity and 25% identity [27 –29]. Both Aβ and hIAPP are neurotoxic. hIAPP is even referred to as a second amyloid in AD [9, 29], mainly due to its ability to cross the blood-brain barrier [30] and deposit in the brains of both AD and T2D patients [19 , 32]. It has also been identified in cerebral capillary walls and in the cerebrospinal fluid of AD and T2D patients [31, 32]. Moreover, postmortem studies of the brain tissue of AD patients have identified co-localized hIAPP/Aβ deposits [19, 31]. Brain tissue analysis of T2D patients has also found hIAPP and irregular hIAPP/Aβ deposits in the grey matter of the temporal lobe [9]. Consistent with this, in vitro studies found that hIAPP deposits in the brain interact with Aβ through a cross-seeding mechanism and worsen AD pathology [33]. Animal studies with a transgenic rat model over-expressing hIAPP found increased levels of hIAPP accumulation in the brain [30 , 34]. Furthermore, transgenic hIAPP mice injected with preformed hIAPP and Aβ fibrils were shown to trigger the formation of amyloid in both the brain and pancreas, potentially through homo/heterologous seeding mechanisms [35]. It has been suggested that the neurotoxic properties of hIAPP may be similar to Aβ [30]. A recent study demonstrated that hIAPP promotes oligomerization of Aβ and the formation of large heterocomplex aggregates and small oligomers [29]. In comparison to the self-aggregation of Aβ and hIAPP, co-oligomerized Aβ₄₂-hIAPP mixtures exhibit amorphous structure and increased neuronal cell death by 3-fold [29]. It was suggested that the neurotoxic properties of Aβ-hIAPP aggregates include the ability to damage and permeabilize cells more strongly than hIAPP aggregates [29, 36]. Collectively, all these studies show that hIAPP can promote Aβ aggregation and neurotoxicity, and that their cross-seeding interactions in the brain are one of the main pathways underlying the risk of neurodegeneration and AD in T2D [29 , 38].

Cross-seeding between IAPP and Aβ occurs via the interaction of IAPP amyloidogenic core residues with residues 16–22 in Aβ [69 , 76]. Molecular dynamics (MD) simulation studies have also determined the ability of IAPP and Aβ to misfold into similar β-strand-turn-β-strand “U-bend” structures that promote efficient cross-seeding capacity [29, 69]. Other nuclear magnetic resonance NMR studies and MD simulations studies have suggested that cross-seeding of IAPP and Aβ leads to the formation of a polymorphic single layer conformation [69, 77]. The mechanism through which proteins aggregate is quite diverse, ranging from a nucleation-dependent mechanism, a nucleation-independent mechanism, primary nucleation, secondary nucleation, and fragmentation and branching [68 , 72–74]. The type of mechanism preferred is dependent on the concentration of the peptide, but most amyloid fibril aggregates are known to follow a nucleation-dependent aggregation mechanism [68 , 72–74]. Aggregation of amyloidogenic proteins is highly chaotic and this may be related to the formation of micelle like oligomers at specific critical concentrations and temperatures [87].

Many other factors like the presence of tau, synuclein, and lipids are known to play an important role in promoting the aggregation of amyloidogenic proteins like hIAPP and Aβ [66 , 87]. Recent MD simulation studies have suggested that Aβ/hIAPP monomers have the tendency to interact with lipid membranes through an absorption-insertion process and change their structure into beta-sheet rich aggregates [69, 78]. α-synuclein (αS) is a 140 AA protein that is present in the brain at high concentration during adulthood [67]. Studies have found that monomers of αS and IAPP cross-seed and form coaggregates upon incubation [67, 69]. The aggregation-prone region of αS that is responsible for the formation of fibrillary aggregates is called the nonamyloid-beta component (NAC) [67]. MD simulation studies have revealed that NAC fibrils cross-seed with Aβ by inducing the formation of a fibrillary Aβ structure with two U-turns connecting three beta strands [69, 79]. The hydrophobic properties of the NAC fibril sequence have been shown to promote the stability of Aβ and IAPP [69, 80]. Interestingly, αS is also known to act as a seeding agent to aggregate and cross-seed unrelated proteins like Aβ and tau, a process known as proteinopathic cross-seeding [66 , 71]. Tau is also known to co-aggregate with Aβ through the interaction of beta strands on Aβ and tau [69 , 81–85]. The hydrophobic region KLVFFA of Aβ has been found to strongly interact with the hydrophobic region VQIINK of tau through the interaction of intermolecular beta sheets [70]. Together, all these studies reveal that amyloids change structure upon aggregation and the various factors that influence amyloid aggregation.

Although the above studies show a gain of toxic function for hIAPP, it is deficient in T2D and loss of its normal function as a metabolic hormone contributes to the disease pathology. This occurs because increased hIAPP aggregation and deposition in the pancreas leads to a decline in peripheral hIAPP [8, 30]. hIAPP deficiency can also be due to beta cell death, which further lowers its synthesis [8]. Like T2D, low hIAPP is being proposed to be a risk factor for cognitive decline and AD [39]. A genome-wide interaction study showed that a hIAPP gene polymorphism is associated with cognitive impairment and brain amyloid burden in AD [39]. A more recent study examined the association of plasma hIAPP with dementia or AD risk in the Framingham Heart Study (FHS), a large population-based, multigeneration cohort [40]. The association between brain volume and plasma hIAPP with AD onset was U-shaped and nonlinear: participants with either a low or an extremely high concentration of plasma hIAPP had a higher rate of AD incidence [40]. Overall, these clinical studies show that hIAPP is helpful for the ageing brain, but an exceedingly high plasma hIAPP levels may lead to aggregation and lose its protective function posing a risk for AD development [39, 40].

In vitro studies and animal models have provided further evidence of the benefits of hIAPP in AD [8, 41]. AD mouse model studies have shown an improvement in cognitive ability, reduction of Aβ levels in the brain and decreased pathological features upon treatment with peripheral hIAPP [41 –44]. hIAPP and non-amyloidogenic rat amylin (rIAPP) also reduce Aβ₄₂ oligomer-induced cell death in neuronal cells [29]. Overall, these findings have led to interest in developing non-aggregating IAPP peptide analogues as a treatment for AD. Pramlintide, a hIAPP analogue with proline mutations at residues A25P, S28P, and S29P, is an FDA approved T2D drug that has garnered interest in AD [45 –47]. In AD animal models, pramlintide administration has been shown to promote Aβ efflux and reduce Aβ plaques in the brain [41 , 45–47]. However, recent studies have found that pramlintide can form fibrils and cross-seed with Aβ, which has raised concern about its use as a treatment for AD [48]. Many other hIAPP analogues are also in development and which are less prone to aggregation [49]. Aggregation studies and MD simulations have identified the IAPP region 15–26 as the “core region” for its self- and hetero-assembly with Aβ_40 (42) [27 , 51]. Within this region, F15, L16, F23, and I26 are potentially considered to be the key residues involved in this process [51]. Of these four residues, interaction of Aβ₄₀ with alanine-substituted peptides (A23 and A26) were found to have strongly reduced affinities, being 200- and 44-fold lower than IAPP, respectively [51]. In contrast, A15 and A16 did not have a strong effect in reducing the ability of IAPP to bind to Aβ₄₀ [51]. These findings suggest that IAPP residues F23 and I26 are the primary surface hot spots for the interaction of IAPP with Aβ [51]. Consequently, aromatic hydrophobic residues (F23 and I26) within this hotspot are deemed to be the key molecular determinants of the ability of hIAPP to aggregate with itself and with Aβ [27 , 51]. Nuclear magnetic resonance studies have also shown that these aromatic residues are involved in the formation of toxic Aβ and IAPP heterocomplexes [29]. However, it is yet to be determined whether hIAPP analogues with substitutions in these key aromatic residues have reduced self-aggregation and Aβ₄₂ cross-seeding interaction, as well as whether they exhibit toxicity in neuronal cells. This leads to the primary hypothesis of this work: whether hIAPP analogues with modifications in these aromatic residues have reduced self-aggregation and co-aggregation with Aβ.

In this study, three novel IAPP analogues containing single and double point mutations at F23 and I26 [50, 51] were developed to address the above hypothesis. These two AAs were substituted with alanine as it is non-bulky and chemically inert [52]. Previous studies suggest that the alanine mutation improved conformational stability, as they reduced aggregation and formed less beta sheets and amorphous aggregates. It is likely that not all AA replacements have the same effect on function or structure of the protein. The effect of the AA substitutions may vary depending on how similar or dissimilar the replaced AAs are based on physico-chemical properties, size, charge, as well as on their position in the sequence or the structure. For example, it is likely that substitution with a hydrophobic AA such as phenylalanine or valine would increase the propensity of self-aggregation of the peptide and its cross-seeding with Aβ and hIAPP, as compared to alanine [50, 51]. IAPP analogue one (A1) has a single point mutation at F23, IAPP analogue two (A2) has a single point mutation at I26, and IAPP analogue three (A3) has a double point mutation at both F23 and I26 (Fig. 1, Supplementary Table 1). The main aim of this study was to determine the ability of these novel hIAPP analogues to aggregate, reduce Aβ₄₂ oligomer formation, and exhibit toxicity in comparison to hIAPP. To address this, a range of biophysical methods were used including Thioflavin-T (Th-T) fluorescence, sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) and western immunoblotting of photo-crosslinked peptides, circular dichroism (CD), as well as colorimetric mitochondrial dehydrogenase assay (MTS) to assess cell viability.

Fig. 1

Aβ₄₂, hIAPP, and IAPP analogues amino acid sequence. The residues highlighted in purple represent the two key aromatic residues of hIAPP. The residues highlighted in yellow represent the alanine substitution in the key residues of novel hIAPP analogues [51].

MATERIALS AND METHODS

Peptide preparation

Synthetic peptides (Aβ₄₂, hIAPP, A1, A2, and A3) were purchased from the ERI Amyloid Laboratory, LLC, USA. Each of these dry peptides were weighed to 1.8 mg and dissolved in 500μl of 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP; Sigma-Aldrich, USA) at room temperature (RT) for 1 h. Following this, the 500μl peptide dissolved in HFIP was separated into 180μg aliquots and left to dry overnight at RT. Subsequently, the dry peptide films were stored at -20°C for future use.

For Th-T aggregation studies, the frozen peptide films were dissolved in 8μl of dimethyl sulfoxide (DMSO; Merck Millipore, USA), centrifuged for 1 min at 13,000 g and then bath-sonicated at 18°C for 15 min using the Bioruptor® Plus sonication device (Diagenode Incorporated, USA). Following this, the peptide solution was brought to a final concentration of 100μM by the addition of 392μl ice-cold Ham’s F12 phenol red-free media (F-12; Atlanta Biologicals, USA). For western immunoblotting and CD spectroscopy experiments, the frozen peptide films were dissolved in 8μl of 20 mM sodium hydroxide (NaOH) and 392μl of phosphate buffered saline (PBS, pH 7.4) after centrifugation and sonication.

Thioflavin-T aggregation assay

Peptides were prepared as described above and their aggregation kinetics was assessed independently and in combination (at a 1:1 ratio) with Aβ₄₂, hIAPP at 20μM concentration. Twenty μl of the prepared independent or combination peptides was added to each well in a black-clear bottom 96-well plate along with 80μl of 6μM Th-T solution (Sigma-Aldrich, USA). A control measurement was also performed with 20μl DMSO and F-12 solution. The plates were then incubated for 22 h in a Perkin-Elmer plate reader with the fluorescence measured every hour. Fluorescence readings in the plate reader were measured with emission and excitation maxima at 490 and 450 nm, respectively. The experiment was repeated three times with four replicates of each of the independent and combination peptides.

SDS-PAGE and western immunoblotting analysis using photo-crosslinked peptides

Independent and combination (1:1 ratio) peptides were prepared as described above and were incubated at 4°C for 24 h before performing photo-induced cross-linking (PICUP). Following incubation, 50μl of each of the independent and combination peptides were dissolved in 5μl of reaction buffer (2.5μl of 20 mM ammonium persulfate (APS) in 10 mM sodium phosphate buffer (pH 7.4) and 2.5μl of 1 mM tris(2,2-bipyridyl) dichlororuthenium (II) hexahydrate (RuBpy; Sigma-Aldrich, USA)). The peptide solution was then irradiated using a Samsung S8 camera via 8 one-second pulses at a 5 cm distance. Immediately after irradiation, the reaction mixture was quenched using 22.5μl 2x Sample Reducing Agent (ThermoFisher Scientific, USA) and 22.5μl 2x NuPAGETM LDS buffer (ThermoFisher Scientific, USA) containing 5% β-mercaptoethanol (Sigma-Aldrich, USA). To determine whether cross-linking was successful, an uncross-linked Aβ₄₂ solution was also prepared with both the reaction and quenching agents but the solution was not irradiated.

The cross-linked peptides and the uncross-linked Aβ₄₂ were then analyzed using gel electrophoresis and western immunoblotting techniques. Following cross-linking, the peptides were heat treated at 96°C for 10 min. Subsequently, 900 ng of each peptide was loaded into the wells in Bolt^TM 4–12% Bis-Tris Gel (ThermoFisher Scientific, USA) alongside 6μl of Pre-Strained Protein standers (Novex Sharp). 1X MES Running Buffer was then added and the gel was set to run at 115 V for 1 h, after which the gel was transferred on to a nitrocellulose membrane using the Trans-Blot Turbo Transfer System (Bio-Rad). After transfer, the membrane was blocked in 20 ml of 5% milk in TBS for 1 h. Following this, the 5% milk was discarded, and the membrane was incubated at 4°C with 5μl anti-amyloid-β antibody [WO-2] or anti-IAPP antibody [T-4157] and 10 ml of 0.5% milk in TBS-T for 24 h. The following day, the membrane was washed with 1X TBS-T (3 times for 8 min each) and incubated for 1 h in 10 ml of 0.5 % milk in TBS-T with 2μl of secondary (2°) anti-mouse antibody for the WO-2 treated membrane and 10μl 2° anti-rabbit antibody for the T-4157 treated membrane. Later, the membranes were washed first with TBS-T and then with TBS (two times for 8 min each). Finally, 1 ml of ECL Western Blotting Detection Reagent 1 and 2 (Amersham) was added to the membranes and imaged using a Fusion FX Spectra 6.0 EDGE chemiluminescence imaging system.

Cell viability analysis in SH-SY5Y neuronal cells following peptide treatment

Cell toxicity assays were performed using a previously established protocol [29, 53]. In brief, SH-SY5Y cells were cultured, plated into 96-well plates (with a seeding density of 20,000 cells/well) and incubated for 24 h in a 37°C 5% CO2 incubator. Along with this, the independent and combination (1:1 ratio) peptides were prepared as described above and were incubated at 4°C for 24 h to enable the formation of oligomers [88]. After 24 h, but before the peptide treatment, the media in the wells were replaced with fresh DMEM (1X, without F-12) treatment media (Gibco) containing L-Glutamine and 1% FBS. Then, the cells were treated with 10, 20, and 30μM concentration of independent peptide, and 20μM of co-oligomerized peptide combinations (1:1 ratio). This was further incubated for 48 h. Subsequently, 20μl of MTS (soluble tetrazolium) assay was added to each well and incubated for 4 h before assessing cell viability. MTS assayed plates were then placed in a Perkin-Elmer plate reader (absorbance at 450 nm) to quantify viable cells. The experiment was repeated three times with 5 replicates of each of the independent and combination peptide cell treatment.

Circular dichroism spectroscopy

The independent and co-oligomerized (1:1 ratio) peptides required for CD spectroscopy were prepared as described above and incubated at 4°C for 24 h. After incubation, 200μl of each peptide was placed in a 1 mm quartz cuvette (Hellma) and loaded on to a Jasco J-810 spectropolarimeter. Measurements of NaOH and PBS buffer were also taken for subtracting background. The instrument was optimized to measure the samples at RT with a 200–260 nm wavelength range, using three accumulations, a 3-s response time, 1 nm bandwidth and 1 nm data pitch. CD data was analyzed using the JFIT program [54] by converting the spectral absorbance into percentage of secondary structure content present in each peptide.

Data analysis

Analyses of raw data and plots were performed using Microsoft Excel and obtained with GraphPad Prism 8, respectively. Th-T fluorescence of each peptide was blank corrected by subtracting the baseline value (t = 0) from each of the readings. ImageLab software was used to quantify the bands of the peptides from SDS-PAGE and western immunoblotting. Background values were subtracted from the readings and then the fold change relative to Aβ₄₂ was calculated by dividing the band intensity of hIAPP or IAPP analogues by the band intensity of Aβ₄₂. SH-SY5Y cell viability percentage (%) was calculated by dividing the sum (upon addition of 5 replicates of each peptide and subsequent division by 5) of the raw fluorescence data of each peptide by the sum of control and then multiplying it by 100. Th-T fluorescent aggregates, immunoblot band intensities, and cell toxicity data was plotted using Prism and were statistically analyzed using one-way ANOVA (with Tukey’s post-hoc test). All the data is presented as a mean+/- standard error of mean (SEM), and the p-value was set to < 0.05 to determine the significance of the data.

RESULTS

Aggregation of IAPP analogues and cross-seeding interaction with Aβ₄₂

Loss of the protective function of the pancreatic hormone hIAPP is implicated in AD and its replacement is gaining interest as an intervention strategy. hIAPP has a high propensity to self-aggregate and to seed Aβ₄₂ aggregation [29]. Therefore, non-aggregating analogues of hIAPP are being developed for management and treatment of T2D and AD. An important consideration for hIAPP analogues for AD is reduced self-aggregation and attenuated co-aggregation or cross-seeding with Aβ₄₂. The aromatic residues phenylalanine (F23) and isoleucine (I26) in the 37 AA hIAPP play a key role in self-aggregation and co-aggregation with Aβ. However, whether hIAPP analogues with substitutions in these key aromatic residues have reduced aggregation, Aβ₄₂ cross-seeding interaction, and toxicity in neuronal cells had not been previously investigated. In this study, three hIAPP analogues (A1, A2, and A3) containing single- and double-point mutations at phenylalanine (F23) and isoleucine (I26) (Fig. 1) were characterized.

Firstly, the physicochemical properties of A1, A2, and A3 were compared to hIAPP using the ExPASy ProtParam server (Supplementary Table 1). Notably, all the IAPP analogues showed similar characteristics, except GRAVY, which was expectedly lower compared to hIAPP. To investigate whether these hIAPP have reduced self-aggregation and co-aggregation with Aβ₄₂, we initially assessed the aggregation of Aβ₄₂, hIAPP and analogues A1, A2, and A3 individually and in combination with Aβ₄₂ using Th-T assays. Th-T is a widely used biophysical method to measure the aggregation kinetics of peptides [55]. Th-T assay contains a benzothiazole fluorescence dye that binds to β-sheets in amyloid fibrils and acts as a marker for amyloid aggregates [55].

We first determined aggregation kinetics of freshly prepared Aβ₄₂, hIAPP, and hIAPP analogues A1, A2, and A3 incubated with Th-T at 0 and every 1 h thereafter up to 22 h. (Fig. 2A, B). Values from 0 to 22 h were normalized by subtracting the baseline value (t = 0) from all values to represent relative increase in fluorescence. In comparison to all the other peptides, the self-aggregation of hIAPP increased steeply from the 3-h time point and plateaued after 18 h (Fig. 2A, B). In contrast, a gradual increase in the self-aggregation of Aβ₄₂ was observed from 11 h (Fig. 2A). The self-aggregation of A1, A2, and A3 was significantly reduced compared to that of hIAPP (Fig. 2A–C), confirming previous studies demonstrating the role of these aromatic residues in mediating hIAPP aggregation [50 –52]. Notably, both the single mutant analogues A1 and A2 and the double mutant analogue A3 showed similar Th-T aggregation profiles (Fig. 2B). We also quantified the total Th-T fluorescence of the peptides after 22 h incubation to compare the levels of aggregation at the end of the kinetics experiment. The self-aggregation of hIAPP was observed to be 30-fold (72065 arbitrary units; AU) higher than that of Aβ₄₂ (2391 AU) at 22 h (Fig. 2C). On the other hand, the self-aggregation of A1, A2, and A3 were 878-fold (82 AU), 54-fold (1348 AU), and 61-fold (1169 AU) lower than hIAPP, respectively, at 22 h (Fig. 2C).

We next determined the aggregation profiles of Aβ₄₂ co-oligomerized with-hIAPP, A1, A2, or A3 to assess if these hIAPP peptides altered the time-course aggregation profile of Aβ₄₂ (Fig. 3A, B). Aβ₄₂ was co-incubated with hIAPP, A1, A2, or A3 in a 1:1 ratio in the presence of Th-T followed by time-based fluorescence analysis, as described before. The Aβ₄₂-hIAPP cross-aggregation profile was compared with that of the self-aggregation of Aβ₄₂, and the cross-aggregation profiles of the analogues were compared with those of Aβ₄₂-hIAPP and Aβ₄₂. In comparison to that of Aβ₄₂, the aggregation profile of Aβ₄₂-hIAPP markedly increased from 0 to 20 h and seemed to plateau after 20 h (Fig. 3A). In particular, the aggregation of Aβ₄₂-hIAPP was 3-fold higher than Aβ₄₂ at 7 h (Fig. 3A). In contrast to Aβ₄₂ and Aβ₄₂-hIAPP, co-oligomerized Aβ₄₂-A1, Aβ₄₂-A2, and Aβ₄₂-A3 demonstrated a marked reduction in their aggregation profiles (Fig. 3A, B). In comparison to Aβ₄₂, aggregation of Aβ₄₂-A1, Aβ₄₂-A2, and Aβ₄₂-A3 was reduced by 8-fold, 5-fold, and 2-fold, respectively, at 22 h (Fig. 3B, D). Notably during the early time points, Aβ₄₂-A3 showed increased aggregation compared to Aβ₄₂ (5–15 h) (Fig. 3A, B). Based on the total Th-T fluorescence quantification at 22 h, the aggregation of Aβ₄₂-A1, Aβ₄₂-A2, and Aβ₄₂-A3 was significantly lower compared to Aβ₄₂-hIAPP (Fig. 3A, B). This demonstrates that all three analogues have reduced co-aggregation with Aβ₄₂, as compared to hIAPP. Although, A1, A2, and A3 showed similar aggregation profiles as individual peptides, distinct differences in aggregation were observed when they were co-oligomerized with Aβ₄₂ (Fig. 3A, B). It seems that A3 showed increased Aβ₄₂ cross-seeding ability compared to A1 and A2. Nevertheless, all hIAPP analogues exhibited lower Aβ₄₂ cross-seeding ability compared to hIAPP. (Fig. 3D). To confirm the changes observed in the Aβ₄₂ aggregation profiles observed with the hIAPP analogues, denaturing SDS-PAGE followed by immunoblotting were used to detect the formation of oligomers.

Fig. 2

Self-aggregation kinetics of Aβ₄₂, hIAPP, A1, A2, and A3. A) Self- aggregation kinetics of Aβ₄₂ and hIAPP (20μM) along with DMSO + F-12 control assessed over 22 h period by Thioflavin-T (Th-T) fluorescence assay. Th-T fluorescence is represented as arbitrary units (AU). B) Individual aggregation kinetics of A1, A2, and A3 (20μM) assessed over 22 h period by Th-T assay. C) Total fluorescence quantification (at 22 h). Data represented as a mean±SEM (n = 3, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001).

Fig. 3

Aggregation kinetics of co-oligomerized Aβ₄₂, hIAPP, and IAPP peptides. A) Aggregation kinetics of co-oligomerized Aβ₄₂-hIAPP mixture, DMSO + F-12 control, (1:1 ratio, 20μM) along with individual Aβ₄₂ control assessed over 22 h period by Th-T fluorescence assay. B) Aggregation kinetics of co-oligomerized Aβ₄₂-A1, Aβ₄₂-A2, Aβ₄₂-A3, and Aβ₄₂ peptides (1:1 ratio, 20μM) assessed over 22 h period by Th-T fluorescence assay. C) Aggregation kinetics of co-oligomerized hIAPP-A1, hIAPP-A2, hIAPP-A3, and hIAPP peptides (1:1 ratio, 20μM) assessed over 22 h period by Th-T fluorescence assay. D) Total fluorescence quantification (at 22 h). Data represented as a mean±SEM (n = 3, ^*p < 0.05, ^****p < 0.0001).

In addition, we determined the aggregation profiles of hIAPP co-oligomerized with A1, A2, or A3 to assess if these hIAPP analogues altered the time-course aggregation profile of hIAPP (Fig. 3C). hIAPP was co-incubated with A1, A2, or A3 in a 1:1 ratio in the presence of Th-T followed by time-based fluorescence analysis, as described above. The cross-aggregation profiles of the analogues were compared with the aggregation profile of hIAPP. In comparison to hIAPP, the aggregation profile of hIAPP-analogues markedly reduced from 0 to 15 h (Fig. 3C). Specifically, in comparison to hIAPP, aggregation of hIAPP-A1, hIAPP-A2, and hIAPP-A3 was reduced by 19-fold, 4-fold, and 1-fold, respectively, at 22 h (Fig. 3C, D). Notably during the 13–22 h time points, hIAPP-A3 showed increased aggregation compared to hIAPP (Fig. 3C). However, based on the total Th-T fluorescence quantification at 22 h, the aggregation of hIAPP-A1, hIAPP-A2, and hIAPP-A3 was significantly lower compared to hIAPP (Fig. 3D). This demonstrates that all three analogues have reduced co-aggregation with hIAPP. It seems that A3 exhibits increased hIAPP cross-seeding ability compared to A1 and A2. Nonetheless, all hIAPP analogues exhibited lower hIAPP cross-seeding ability compared to hIAPP alone (Fig. 3D).

Co-oligomerization and size distribution of Aβ₄₂-IAPP analogues mixtures

Previous studies have demonstrated that hIAPP promotes Aβ₄₂ oligomerization and formation of oligomer and large amorphous aggregates [29]. While hIAPP acted like a seed for Aβ₄₂ aggregation, the above findings demonstrate that the hIAPP analogues with aromatic residues substituted with alanine reduce Aβ₄₂ aggregation (Fig. 3). We next determined if hIAPP analogues reduce Aβ₄₂ aggregation by altering the formation of its oligomers using PICUP. This technique has been previously utilized to covalently cross-link and stabilize metastable protein complexes that are prone to disaggregation by SDS and heat-denaturing treatments [39, 56]. A representative image of the western immunoblots of cross-linked (CL) and uncross-linked (UCL) Aβ₄₂ samples is shown in Fig. 4A. The UCL Aβ₄₂ displayed monomer and dimer as the most prominent bands detected using Aβ antibody WO2. By contrast, the CL Aβ₄₂ showed monomer (4.6 kDa), dimer (10 kDa), trimer (13 kDa), tetramer (16.5 kDa), pentamer (> 20 kDa), hexamer (> 20 kDa), higher molecular weight oligomers (HMWA; 45–160 kDa), and large aggregates (LA; > 250 kDa), demonstrating that PICUP stabilizes metastable proteins that are susceptible to disaggregation by heat denaturation and SDS-PAGE methods [29, 56].

Fig. 4

Oligomerization and size distribution of cross-linked Aβ₄₂-IAPP and Aβ₄₂-IAPP analogue mixtures. A) Western immunoblotting analysis of un-crosslinked (UCL) and cross-linked (CL) Aβ₄₂ peptide at 24 h time period was analyzed using WO2 (an Aβ antibody). B) Western immunoblotting analysis of CL co-oligomerized Aβ₄₂ and Aβ₄₂-hIAPP mixture (1:1 ratio, 900 ng). C) CL co-oligomerized Aβ₄₂, Aβ₄₂-A1, Aβ₄₂-A2, and Aβ₄₂-A3 mixture at 22 h period analyzed using WO2 antibody.

To assess if hIAPP analogues altered Aβ₄₂ oligomer formation, co-oligomerized mixtures of Aβ₄₂-hIAPP, Aβ₄₂-A1, Aβ₄₂-A2, and Aβ₄₂-A3 were cross-linked using PICUP, followed by SDS-PAGE, western immunoblotting analysis (Fig. 4), and densiometric analysis to quantify the Aβ₄₂ oligomer species (Fig. 5). As previously shown [29], CL Aβ₄₂-hIAPP mixtures displayed large insoluble aggregates that were unresolved in the wells (Fig. 4B). Large aggregates in CL Aβ₄₂-A1 mixtures were reduced by 2.4-fold compared to CL Aβ₄₂-hIAPP (Fig. 5A, B), although this change was not significant. Notably, these large aggregates were significantly absent in CL Aβ₄₂-A2 and CL Aβ₄₂-A3 compared to CL Aβ₄₂-hIAPP (Fig. 4C, 5E). CL Aβ₄₂-hIAPP showed increased levels of hexamers, but this was not significant (Fig. 5A). CL Aβ₄₂-A1 mixtures showed an increase in large aggregates and hexamers, but this was not significant compared to CL Aβ₄₂ (Fig. 5B). CL Aβ₄₂-A2 mixtures showed an increase in hexamers, but again this was not significant compared to CL Aβ₄₂ (Fig. 5C). Consistent with our above findings in Th-T assays (Fig. 3B, D), all CL Aβ₄₂-A1, A2, or A3 mixtures showed significant reduction in the formation of large aggregates of Aβ₄₂. These findings provide further evidence that hIAPP analogues with aromatic residues substituted with alanine have reduced ability to cross-seed with Aβ₄₂ compared to hIAPP. To obtain further insight into the structural nature of their interaction, we next assessed secondary structure changes in the Aβ₄₂-hIAPP mixtures using CD spectro-scopy.

Fig. 5

Size distribution of cross-linked Aβ₄₂-IAPP and Aβ₄₂-IAPP analogue mixtures. Densiometric analysis of CL co-oligomerized (A) Aβ₄₂ versus Aβ₄₂-hIAPP, (B) Aβ₄₂ versus Aβ₄₂-A1, (C) Aβ₄₂ versus Aβ₄₂-A2, and (D) Aβ₄₂ versus Aβ₄₂-A3 mixtures at 22 h period. E) Significance of co-oligomerized peptides large aggregates. Band intensities represented as fold-change relative to Aβ₄₂ (mean±SEM, n = 3, ^*p < 0.05, ^**p < 0.01).

Secondary structural analysis of Aβ₄₂-hIAPP analogues

CD spectroscopy is commonly used to determine the amount of secondary structure content of a protein by analyzing the differences in absorption of left- and right-circularly polarized light by an aqueous protein sample [54 , 58]. This is measured in millidegrees, which is then converted to delta epsilon (Δ ε, ellipticity) to create the CD spectra. Each protein secondary structure (α-helix, β-sheet, and random coil) exhibits a CD spectrum with a characteristic shape. Aggregated hIAPP has been reported to have a CD spectrum consistent with predominantly β-sheet structure [29]. Conversely, the co-oligomerized Aβ₄₂-hIAPP mixture as well as aggregated Aβ₄₂ had CD spectra characteristic of random coil structures and were thus considered to be amorphous [29]. Here, CD spectroscopy was utilized to determine the secondary structure content in aggregated independent hIAPP analogues and as co-oligomerized mixtures with Aβ₄₂.

Compared to the spectra of independent and co-oligomerized hIAPP, a slight negative change in the spectra from 200–230 nm was observed in all three hIAPP analogues and their combinations (Fig. 6A–D). From 205 nm the magnitude gradually increased to 0 Δ ε for all peptides (Fig. 6A–C, D). Standard random coil spectra usually display a more negative curve at the lowest wavelength compared to β-sheet and α-helix [57, 58]. Between 210–230 nm, a standard curve of β-sheet and random coil is usually more positive than α-helix [57, 58]. Based on this, all three analogues were qualitatively deemed to have a spectral shape of mixtures of α-helix and random coil in comparison to independent and co-oligomerized Aβ₄₂, hIAPP and, Aβ₄₂: hIAPP (Fig. 6A–D).

Fig. 6

CD spectra of independent and co-oligomerized peptides. A) CD spectra of Aβ₄₂, hIAPP, and Aβ₄₂-hIAPP mixtures [1:1], (B) CD spectra of Aβ₄₂, A1, and Aβ₄₂-A1 mixtures [1:1], (C) CD spectra of Aβ₄₂, A2, and Aβ₄₂-A2 mixtures [1:1], and (D) CD spectra Aβ₄₂, A3, and Aβ₄₂-A3 mixtures [1:1] at 24 h period. Each spectrum represents three replicates per peptide sample. Spectra represented as a mean±SEM.

The CD spectra were further analyzed using the JFIT analysis program to estimate the secondary structure content in each of the independent and combination peptides (Table 1A, B). Aggregated Aβ₄₂ was estimated to have 14% α-helix, 19% β-sheet, and 67% random coil (Table 1A). By contrast, aggregated hIAPP was estimated to have 16% α-helix, 43% β-sheet, and 40% random coil (Table 1A). Consistent with previous findings, self-aggregated hIAPP predominantly exhibited β-sheet structure, whereas the Aβ₄₂:hIAPP combination predominantly exhibited random coil content (Table 1A). In contrast, all three independent and co-oligomerized analogues predominantly exhibited random coil content (A3 > A1 > A2) (Table 1A), followed by α-helix and then β-sheet contents. All three independent and co-oligomerized analogues were estimated to have 20–23% α-helix, 12–20% β-sheet, and 57–67% random coil content (Table 1A, B). In particular, the random coil content of A1, A2, and A3 significantly increased by 23, 17, and 26% points, respectively, compared to hIAPP (p < 0.001) (Table 1A). In comparison to Aβ₄₂:hIAPP, the random coil content in Aβ₄₂:A1, Aβ₄₂:A2, and Aβ₄₂:A3 significantly increased by 17, 13, and 18% points, respectively (p < 0.001) (Table 1B). In line with the increase in random coil content in the three analogues, a significant concomitant decrease in β-sheet content in both independent and co-oligomerized analogues was also observed when compared to hIAPP and Aβ₄₂:hIAPP, respectively (p < 0.001) (Table 1A, B). Overall, A1, A2, and A3 peptides had significantly reduced β-sheet content and increased random coil content compared to hIAPP. The same effect was also reflected in the co-oligomerized mixtures with Aβ₄₂. Collectively, our Th-T fluorescence and CD spectroscopy data reveal that the hIAPP analogues attenuate the formation of large aggregates and induce a distinct change in secondary structure content to a more amorphous nature.

Table 1

Secondary structure content of independent and co-oligomerized peptides. A) Secondary structure content table representing the α-helix, β-sheets, and random coils of 20μM independent Aβ₄₂, hIAPP, A1, A2, and A3 peptide as a percentage. B) Secondary structure content table of co-oligomerized Aβ₄₂-hIAPP, Aβ₄₂-A1, Aβ₄₂-A2 and Aβ₄₂-A3 peptide mixture (1:1 ratio, 20μM)

Sample	% alpha-helix	% beta-sheet	% random coil
A)
Aβ₄₂	14.16±1.45	19.29±1.49	66.55±1.17
hIAPP	16.14±0.74	43.42±2.86	40.44±2.16
A1	20.01±0.42	16.38±2.46	63.61±2.05
A2	23.06±0.94	19.66±2.14	57.29±2.80
A3	20.22±0.14	12.96±0.46	66.79±0.32
B)
Aβ₄₂:hIAPP	19.08±1.10	32.21±3.13	48.71±2.04
Aβ₄₂:A1	20.42±0.48	13.83±0.62	65.74±1.10
Aβ₄₂:A2	20.62±0.06	17.58±1.29	61.80±1.35
Aβ₄₂:A3	21.07±0.64	12.45±0.95	66.48±1.19

The error bars represent±SEM. Each value represents 9 technical and 3 biological replicates. All three independent and co-oligomerized analogues β-sheet and random coils content had a p-value < 0.0001 when compared to hIAPP and Aβ₄₂-hIAPP, respectively.

Neuroprotective effect of hIAPP analogues against Aβ₄₂-induced toxicity

Multiple lines of evidence from clinical studies, animal and cell models have shown that extremely high hIAPP concentrations promote Aβ₄₂ aggregation and neuronal cell death, whereas at lower, non-aggregating concentrations hIAPP alleviates it [29]. The novel hIAPP analogues with aromatic residues substituted by alanine characterized in this study show reduced self-aggregation and cross-seeding interaction with Aβ₄₂. Extending these findings, we investigated if these analogues on their own and/or in combination with Aβ₄₂ induced neuronal cell death using the SH-SY5Y human neuroblastoma cell model [29]. For this purpose, SH-SY5Y cells were treated with increasing doses of Aβ₄₂, hIAPP, and hIAPP analogues A1, A2, and A3 (10, 20, 30μM) for 72 h followed by an MTS assay to measure mitochondrial dehydrogenase activity as a measure of cell viability (Fig. 7A). The viability of cells treated with hIAPP analogues in combination with Aβ₄₂ (1:1 ratio) at 20μM total peptide concentration was also determined and compared to that of the independent peptides at 20μM (Fig. 7B).

Fig. 7

SH-SY5Y neuronal cell toxicity of independent and co-oligomerized peptides. A) Cell toxicity assay of Aβ₄₂, hIAPP, A1, A2, and A3 at increasing concentrations (10, 20, 30μM). B) Cell toxicity assay of independent Aβ₄₂, hIAPP, A1, A2, A3 peptide and co-oligomerized Aβ₄₂-hIAPP, Aβ₄₂-A1, Aβ₄₂-A2, and Aβ₄₂-A3 peptides at 20μM concentration. Untreated cells contain DMSO + F-12. Data represented as a mean±SEM (n = 3, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001).

At higher concentrations of 30μM, Aβ₄₂, hIAPP, and A1 induced cell death up to 60–70%, whereas A2 and A3 induced cell death up to 30–40%, respectively (Fig. 7A). At a lower concentration of 10μM, hIAPP and A1, and Aβ₄₂ induced cell death up to 20% and 40%, respectively. On the other hand, A2 and A3 showed no significant cell death (Fig. 7A). Overall, A2 and A3 were significantly less toxic compared to Aβ₄₂ and hIAPP (Fig. 7A).

Co-oligomerized Aβ₄₂:hIAPP displayed higher toxicity than hIAPP (Fig. 7B). Notably, Aβ₄₂:A1, Aβ₄₂:A2, and Aβ₄₂:A3 induced cell death by only 10%, 2%, and 1%, respectively (Fig. 7B). Moreover, all three co-oligomerized hIAPP analogues significantly increased cell viability when compared to hIAPP and Aβ₄₂:hIAPP (Fig. 7B). In particular, cell viability of all three co-oligomerized analogues increased by 1.2–1.5-fold, 2-fold, and 3.2–3.5-fold, when compared to their own individual peptides, Aβ₄₂ and Aβ₄₂:hIAPP, respectively (Fig. 7B). Interestingly, unlike the other two analogues, A1 was toxic at 20μM individually, but in combination with Aβ₄₂ no significant cell death was observed (Fig. 7B). Overall, all three co-oligomerized hIAPP analogues were significantly less toxic than Aβ₄₂, hIAPP, A1, A2, A3, and Aβ₄₂:hIAPP. These observations are in agreement with the above aggregation and cross-seeding studies, indicating that a reduction in aggregation is associated with reduced cell death in Aβ₄₂:A1-3 combinations.

DISCUSSION

With aging there is a decline in the ability of cells to maintain protein homeostasis or proteostasis, which is a characteristic feature of chronic degenerative diseases like AD and T2D [59]. Diseases involving loss of function are often the result of disruption of normal protein homeostasis, typically caused by mutation in a related protein, thereby compromising its cellular folding [59]. On the other hand, diseases involving gain of function often occur because of a chronic disorder in proteostasis leading to decreased ability to degrade misfolded proteins and thus accumulation of toxic protein aggregates, such as that involving amyloidogenic proteins like Aβ and hIAPP [59]. Human diseases involving loss of function and gain of toxic function are generally viewed as exclusive pathological events, but the dichotomy of hIAPP has revealed that they can occur simultaneously and even as a feedback mechanism that contribute to AD [39 , 59]. Excess hIAPP accumulation can lead to toxic protein aggregation (gain of function) and at the same time cause its deficiency (loss of function) while both events may contribute to disease pathogenesis [39, 40]. Therefore, hIAPP replacement therapy has been proposed as an intervention for AD. Due to its high propensity for aggregation [29], hIAPP analogues with reduced self-aggregation are an important consideration for IAPP-based treatments [49].

The cross-seeding interaction of hIAPP with Aβ₄₂ is central to the increased risk of neurodegeneration and is, therefore, an important target for treatments [29 , 38]. Recent findings demonstrate that co-oligomerized Aβ₄₂-hIAPP form large amorphous aggregates and increase neuronal cell death by up to 3-fold compared to Aβ₄₂ or hIAPP alone [29]. In this study, novel IAPP analogues with alanine mutations in the amyloidogenic region (F23 and I26) were characterized as potential candidates for AD [27 , 50–52]. This study revealed that self-aggregation of all hIAPP analogues is significantly reduced compared to hIAPP and Aβ₄₂. In particular, self-aggregation of A1 with single point mutation at F23 is greatly reduced, followed by A3 with double mutation at F23 and I26, and then A2 with single mutation at I26. Th-T fluorescence aggregation observations (22 h, 20μM) were consistent with a recent Th-T fluorescence and transmission electron microscopy study (7-day, 12μM) that showed that the amyloidogenicity of all three analogues was reduced by 20-fold compared to hIAPP [51]. This suggests that the mutations within this region may reduce its hydrophobicity, self-recognition pattern, and self-binding affinity, thus decreasing its ability to aggregate [51].

Cross-seeding has been established as a key mechanism through which amyloidogenic hIAPP acts as a seed to promote the aggregation of Aβ₄₂ in AD [27 –31]. Consistent with this view, this study showed that hIAPP promotes co-oligomerization of Aβ₄₂. Studies suggest that this could potentially be due their 50% sequence similarity, the presence of opposite electric charges, increased hydrophobic interactions and high cross-binding affinities, particularly between the hIAPP 20–28 and Aβ₄₂ 25–33 regions (which form U-shaped, β-strand–turn–β-strand motifs in their fibril forms) [27 , 60]. Consistent with the Th-T fluorescence observations and previous studies [29], our immunoblotting analysis found increased amounts of large aggregate formation in co-oligomerized Aβ₄₂-hIAPP compared to Aβ₄₂ alone. This suggests that hIAPP amplifies Aβ₄₂ oligomerization [29].

By contrast, Th-T fluorescence analysis showed that all three hIAPP analogues exhibited reduced co-oligomerization with Aβ₄₂ compared to hIAPP. This could potentially be due to reduced hydrophobic interactions and cross-binding affinities of the alanine mutations [51, 60]. Moreover, previous studies have shown that A1 and A2 have slightly reduced binding affinity towards Aβ₄₂ in comparison to A3, which is consistent with our Th-T fluorescence findings [51]. In addition to this, co-oligomerized Aβ₄₂:A1, Aβ₄₂:A2, and Aβ₄₂:A3 reduced the observed levels of small oligomers and large aggregates in comparison to Aβ₄₂:hIAPP, although only the reduction of large aggregates of Aβ₄₂:A2 and Aβ₄₂:A3 were statistically significant. This reduction in the formation of potential Aβ₄₂:analogue heterocomplexes could be due to reduced oligomer-monomer/oligomer interactions [29, 38]. In particular, the Aβ₄₂:A3 combination showed a distinct reduction in Aβ₄₂ co-oligomerization compared to Aβ₄₂ and Aβ₄₂:hIAPP, which indicates that the double point alanine mutation is more effective than single point mutations. This suggests that the reduced co-oligomerization with Aβ₄₂ may have been driven by reduced hydro-phobic interactions rather than by reduced cross-binding affinities [27–29 , 60].

The IAPP analogues and its co-oligomerized mixtures with Aβ₄₂ were also assessed using IAPP antibody to determine heterocomplex formation. Like the non-aggregating rat IAPP (rIAPP) [29], A1, A2, and A3 were detected as monomeric isoforms (Supplementary Figure 1A). Similar to our previous study [29], the IAPP antibody detected only the monomeric isoforms and small oligomers and large aggregate species of IAPP were undetectable, even in the co-oligomerized Aβ₄₂: hIAPP mixtures (Supplementary Figure 1B). Consequently, to compare Aβ₄₂-heterocomplex formation of hIAPP:Aβ₄₂ and hIAPP analogues:Aβ₄₂, future studies should consider testing these peptides with other existing or tailored IAPP antibodies with high binding affinities towards IAPP, or the specific IAPP peptide being tested, respectively.

CD spectroscopy analysis was performed to determine the structural nature of the aggregates formed by the analogues with themselves and with Aβ₄₂. Consistent with previous studies, aggregated Aβ₄₂ predominantly has random coil content, hIAPP mainly has β-sheet content and Aβ₄₂:hIAPP has mostly random coil content, closely followed by β-sheet content [29, 51]. The distinct differences in structural content between aggregated hIAPP and Aβ₄₂:hIAPP are likely related to the ability of these peptides to form heterocomplexes [29]. In contrast to hIAPP and Aβ₄₂:hIAPP, this study demonstrated that the independent and co-oligomerized hIAPP analogues predominantly exhibit random coil conformation with a marked decrease in β-sheet content. This confirms the potential reduction in their ability to form large amorphous heterocomplexes structures.

The total mass of Aβ in the AD brain is estimated to be 6.5 mg compared to 1.7 mg in the control. Due to its heterogenous distribution and high propensity for aggregation, free amyloid concentration is lower compared to local environments within and surrounding the senile insoluble plaques in AD brain, which contain high micromolar levels of Aβ, comparable to the concentrations required to induce neurotoxicity in cell culture models [89, 90]. Consistent with previous studies, the findings of this work demonstrate that Aβ₄₂:hIAPP hetero-aggregates increase cell death in comparison to aggregates of Aβ₄₂ and hIAPP by themselves [29]. The cell toxicity of the novel IAPP analogues was analyzed for the first time in this study, showing that both the independent and co-oligomerized analogues reduce neurotoxicity. Although dose-dependent cell toxicity of the independent analogues was evident, no significant loss of cell viability was observed compared to hIAPP. However, even though the self-aggregation of A1 was significantly reduced, it exhibited significant neurotoxicity as the concentration increased. This suggests that A1 could form distinct hetero-oligomers that may interfere with the cell-membrane and promote toxicity. However, the toxicity of A1 could be caused by the presence of the I26 residue. Even though the aggregation of A3 was slightly higher than that of the other analogues, its cell toxicity was significantly reduced in comparison to that of Aβ₄₂ and hIAPP alone. This reduction in cell toxicity could mean that the F23 and I26 mutations in these analogues reduce the ability of IAPP aggregates or oligomers to interact with cell membranes and stimulate cell death pathways [60].

Furthermore, all three analogues showed significantly reduced toxicity when co-aggregated with Aβ₄₂. This finding reveals that hIAPP analogues can reduce cross-seeding aggregation with Aβ₄₂ but, at the same time, Aβ₄₂ can also reduce some of the toxic effects of A1. The reduced amounts of cross-seeding aggregates in these analogues may prevent Aβ-induced neurotoxicity [29 , 61]. There are multiple pathways by which amyloid proteins mediate cell toxicity. They may include protein aggregation causing oxidative stress, vesicle trafficking dysfunction, membrane damage, mitochondrial damage, receptor activity modulation, etc. [62–65 , 69]. Some pathways are common amongst amyloid proteins, but some also have distinct toxic properties. For example, different isoforms of Aβ (i.e., 40 versus 42) show different mechanisms of toxicity [29]. All peptides tested here were toxic at high concentrations, but at 30μM, A2 and A3 were significantly less toxic (i.e., 60–70% viability) compared to Aβ₄₂ (30%), hIAPP (30%), and A1 (40%). This suggests that the alanine substitutions of residues 23 and 26 did not completely abrogate all toxicity in the A1-3 analogues and that these peptides exhibit toxic properties due to the presence of other residues. Therefore, one main reason for the low toxicity in combinations may be due to cross-seeding of these peptides, thus masking epitopes in both peptides that are responsible for mediating toxicity via receptor binding. Previous studies have suggested that hIAPP or Aβ₄₂:hIAPP heterocomplex aggregates could bind to amylin receptors on the cell membrane and promote cell death by altering cell signaling [62]. However, other studies have suggested that IAPP peptides with reduced amyloidogenicity (like the truncated AC253 and AC187 IAPP peptides) could act as antagonists for amylin receptors and prevent Aβ-induced neurotoxicity [63 –65]. Consequently, co-oligomerized analogue aggregates or their potential heterocomplexes with Aβ₄₂ might exert their neuroprotective effects by binding to amylin receptors [63 –65]. Further research is therefore necessary to determine if these analogues and their potential heterocomplexes with Aβ₄₂ interact with the cell membrane and modulate binding affinity and amylin receptor activity to maintain their neuroprotective effects [63 –65]. Overall, A3 and A2 exhibited reduced self-aggregation and reduced Aβ₄₂ cross-seeding aggregation and neurotoxicity, suggesting that the I26 residue may play a crucial role in Aβ₄₂ aggregation and Aβ₄₂-induced neurotoxicity.

In conclusion, residues F23 and I26 within the amyloidogenic core of hIAPP appear to play an essential role in its self-aggregation as well as its hetero-aggregation with Aβ₄₂. It is notable that all hIAPP analogues showed reduced self-aggregation but had distinct effects on cross-seeding aggregation, large aggregate formation, and cell death with Aβ₄₂. Moreover, analogues A2 and A3 showed reduced toxicity compared to A1. Overall, these findings indicate that all three analogues are beneficial and isoleucine (I26) in hIAPP is more critical for Aβ₄₂ aggregation and cell toxicity.

Footnotes

ACKNOWLEDGMENTS

The authors acknowledge the facilities and the scientific and technical assistance of the Curtin Health Innovation Research Institute at Curtin University. We acknowledge the support of NH&MRC-ARC dementia research development fellowship (APP1107109) to PB.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

World Health Organization (2017) Global action plan on the public health response to dementia 2017-2025, https://apps.who.int/iris/bitstream/handle/10665/259615/9789241513487-eng.pdf?sequence=1 , Accessed on November 20, 2020

Verdile

, Fuller

, Martins

(2015) The role of type 2 diabetes in neurodegeneration. Neurobiol Dis 84, 22–38.

Querfurth

, LaFerla

(2010) Alzheimer’s disease. N Engl J Med 362, 329–344.

World Health Organization (2012) Dementia, a public health priority, https://apps.who.int/iris/bitstream/handle/10665/259615/9789241513487-eng.pdf?sequence=1 , Accessed on November 20, 2020

World Health Organization (2016) Global report on diabetes, https://apps.who.int/iris/bitstream/handle/10665/204871/9789241565257_eng.pdf?sequence=1&isAllowed=y, Accessed on November 20, 2020

Ninomiya

(2014) Diabetes mellitus and dementia. Curr Diab Rep 14, 487.

Ott

, Stolk

, Hofman

, van Harskamp

, Grobbee

, Breteler

(1996) Association of diabetes mellitus and dementia: The Rotterdam Study. Diabetologia 39, 1392–1397.

Lutz

, Meyer

(2015) Amylin at the interface between metabolic and neurodegenerative disorders. Front Neurosci 9, 216.

Jackson

, Barisone

, Diaz

, Jin

, DeCarli

, Despa

(2013) Amylin deposition in the brain: A second amyloid in Alzheimer disease? Ann Neurol 74, 517–526.

10.

Biessels

, Strachan

, Visseren

, Kappelle

, Whitmer

(2014) Dementia and cognitive decline in type 2 diabetes and prediabetic stages: Towards targeted interventions. Lancet Diabetes Endocrinol 2, 246–255.

11.

Sipe

, Benson

, Buxbaum

, Ikeda

, Merlini

, Saraiva

, Westermark

(2010) Amyloid fibril protein nomenclature: 2010 recommendations from the nomenclature committee of the International Society of Amyloidosis. Amyloid 17, 101–104.

12.

Krishnaswamy

, Verdile

, Groth

, Kanyenda

, Martins

(2009) The structure and function of Alzheimer’s gamma secretase enzyme complex. Crit Rev Clin Lab Sci 46, 282–301.

13.

Mirzabekov

, Lin

, Kagan

(1996) Pore formation by the cytotoxic islet amyloid peptide amylin. J Biol Chem 271, 1988–1992.

14.

Newsholme

, Gaudel

, Krause

(2012) Mitochondria and diabetes. An intriguing pathogenetic role. Adv Exp Med Biol 942, 235–247.

15.

Atwood

, Obrenovich

, Liu

, Chan

, Perry

, Smith

, Martins

(2003) Amyloid-beta: A chameleon walking in two worlds: A review of the trophic and toxic properties of amyloid-beta. Brain Res Brain Res Rev 43, 1–16.

16.

Kamenetz

, Tomita

, Hsieh

, Seabrook

, Borchelt

, Iwatsubo

, Sisodia

, Malinow

(2003) APP processing and synaptic function. Neuron 37, 925–937.

17.

Chimon

, Shaibat

, Jones

, Calero

, Aizezi

, Ishii

(2007) Evidence of fibril-like β-sheet structures in a neurotoxic amyloid intermediate of Alzheimer’s β-amyloid. Nat Struct Mol Biol 14, 1157–1164.

18.

Bharadwaj

, Dubey

, Masters

, Martins

, Macreadie

(2009) Abeta aggregation and possible implications in Alzheimer’s disease pathogenesis. J Cell Mol Med 13, 412–421.

19.

Cras

, Kawai

, Lowery

, Gonzalez-DeWhitt

, Greenberg

, Perry

(1991) Senile plaque neurites in Alzheimer disease accumulate amyloid precursor protein. Proc Natl Acad Sci U S A 88, 7552–7556.

20.

Alberdi

, Sánchez-Gómez

, Cavaliere

, Pérez-Samartín

, Zugaza

, Trullas

, Domercq

, Matute

(2010) Amyloid beta oligomers induce Ca2+dysregulation and neuronal death through activation of ionotropic glutamate receptors. Cell Calcium 47, 264–272.

21.

Yan

, Chen

, Fu

, Chen

, Zhu

, Roher

, Slattery

, Zhao

, Nagashima

, Morser

, Migheli

, Nawroth

, Stern

, Schmidt

(1996) RAGE and amyloid-beta peptide neurotoxicity in Alzheimer’s disease. Nature 382, 685–691.

22.

Parri

, Dineley

(2010) Nicotinic acetylcholine receptor interaction with beta-amyloid: Molecular, cellular, and physiological consequences. Curr Alzheimer Res 7, 27–39.

23.

Verdier

, Penke

(2004) Binding sites of amyloid beta-peptide in cell plasma membrane and implications for Alzheimer’s disease. Curr Protein Pept Sci 5, 19–31.

24.

Lim

, Rhein

, Baysang

, Meier

, Poljak

, Raftery

, Guilhaus

, Ittner

, Eckert

, Götz

(2010) Abeta and human amylin share a common toxicity pathway via mitochondrial dysfunction. Proteomics 10, 1621–1633.

25.

Haass

, Selkoe

(2007) Soluble protein oligomers in neurodegeneration: Lessons from the Alzheimer’s amyloid beta-peptide. Nat Rev Mol Cell Biol 8, 101–112.

26.

Cooper

, Willis

, Clark

, Turner

, Sim

, Reid

(1987) Purification and characterization of a peptide from amyloid-rich pancreases of type 2 diabetic patients. Proc Natl Acad Sci U S A 84, 8628–8632.

27.

O’Nuallain

, Williams

, Westermark

, Wetzel

(2004) Seeding specificity in amyloid growth induced by heterologous fibrils. J Biol Chem 279, 17490–17499.

28.

Andreetto

, Yan

, Tatarek-Nossol

, Velkova

, Frank

, Kapurniotu

(2010) Identification of hot regions of the Abeta-IAPP interaction interface as high-affinity binding sites in both cross- and self-association. Angew Chem Int Ed Engl 49, 3081–3085.

29.

Bharadwaj

, Solomon

, Sahoo

, Ignasiak

, Gaskin

, Rowles

, Verdile

, Howard

, Bond

, Ramamoorthy

, Martins

, Newsholme

(2020) Amylin and beta amyloid proteins interact to form amorphous heterocomplexes with enhanced toxicity in neuronal cells. Sci Rep 10, 10356.

30.

Bharadwaj

, Wijesekara

, Liyanapathirana

, Newsholme

, Ittner

, Fraser

, Verdile

(2017) The link between type 2 diabetes and neurodegeneration: Roles for amyloid-β, amylin, and tau proteins. J Alzheimers Dis 59, 421–432.

31.

Oskarsson

, Paulsson

, Schultz

, Ingelsson

, Westermark

(2015) In vivo seeding and cross-seeding of localized amyloidosis: A molecular link between type 2 diabetes and Alzheimer disease. Am J Pathol 185, 834–846.

32.

Fawver

, Ghiwot

, Koola

, Carrera

, Rodriguez-Rivera

, Hernandez

, Dineley

, Kong

, Li

, Jhamandas

, Perry

, Murray

(2014) Islet amyloid polypeptide (IAPP): A second amyloid in Alzheimer’s disease. Curr Alzheimer Res 11, 928–940.

33.

Ono

, Takahashi

, Ikeda

, Mizuguchi

, Hamaguchi

, Yamada

(2014) Exogenous amyloidogenic proteins function as seeds in amyloid β-protein aggregation. Biochim Biophys Acta 1842, 646–653.

34.

Wijesekara

, Ahrens

, Sabale

, Wu

, Ha

, Verdile

, Fraser

(2017) Amyloid-β and islet amyloid pathologies link Alzheimer’s disease and type 2 diabetes in a transgenic model. FASEB J 31, 5409–5418.

35.

Srodulski

, Sharma

, Bachstetter

, Brelsfoard

, Pascual

, Xie

, Saatman

, Van Eldik

, Despa

(2014) Neuroinflammation and neurologic deficits in diabetes linked to brain accumulation of amylin. Mol Neurodegener 9, 30.

36.

Bharadwaj

, Solomon

, Malajczuk

, Mancera

, Howard

, Arrigan

DWM

, Newsholme

, Martins

(2018) Role of the cell membrane interface in modulating production and uptake of Alzheimer’s beta amyloid protein. Biochim Biophys Acta Biomembr 1860, 1639–1651.

37.

Young

, Mahood

, Saunders

, Tu

, Raleigh

, Radford

, Ashcroft

(2015) Insights into the consequences of co-polymerisation in the early stages of IAPP and Aβ peptide assembly from mass spectrometry. Analyst 140, 6990–6999.

38.

Berhanu

, Yaşar

, Hansmann

(2013) In silico cross seeding of Aβ and amylin fibril-like oligomers. ACS Chem Neurosci 4, 1488–1500.

39.

Roostaei

, Nazeri

, Felsky

, De Jager

, Schneider

, Pollock

, Bennett

, Voineskos

(2017) Genome-wide interaction study of brain beta-amyloid burden and cognitive impairment in Alzheimer’s disease. Mol Psychiatry 22, 287–295.

40.

Zhu

, Tao

, Ang

TFA

, Massaro

, Gan

, Salim

, Zhu

, Kolachalama

, Zhang

, Devine

, Auerbach

, DeCarli

, Au

, Qiu

(2019) Association of plasma amylin concentration with Alzheimer disease and brain structure in older adults. JAMA Netw Open 2, e199826.

41.

Zhu

, Wang

, Wallack

, Li

, Carreras

, Dedeoglu

, Hur

, Zheng

, Li

, Fine

, Mwamburi

, Sun

, Kowall

, Stern

, Qiu

(2015) Intraperitoneal injection of the pancreatic peptide amylin potently reduces behavioral impairment and brain amyloid pathology in murine models of Alzheimer’s disease. Mol Psychiatry 20, 252–262.

42.

Zhu

, Xue

, Wang

, Wallack

, Na

, Hooker

, Kowall

, Tao

, Stein

, Wolozin

, Qiu

(2017) Amylin receptor ligands reduce the pathological cascade of Alzheimer’s disease. Neuropharmacology 119, 170–181.

43.

Zhu

, Stern

, Tao

, Bourlas

, Essis

, Chivukula

, Rosenzweig

, Steenkamp

, Xia

, Mercier

, Tripodis

, Farlow

, Kowall

, Qiu

(2017) An amylin analog used as a challenge test for Alzheimer’s disease. Alzheimers Dement 3, 33–43.

44.

Mohamed

, Zhu

, Mousa

, Wang

, Qiu

, Kaddoumi

(2017) Amylin enhances amyloid-β peptide brain to blood efflux across the blood-brain barrier. J Alzheimers Dis 56, 1087–1099.

45.

Ridgway

, Eldrid

, Zhyvoloup

, Ben-Younis

, Noh

, Thalassinos

, Raleigh

(2020) Analysis of proline substitutions reveals the plasticity and sequence sensitivity of human IAPP amyloidogenicity and toxicity. Biochemistry 59, 742–754.

46.

Pullman

, Darsow

, Frias

(2006) Pramlintide in the management of insulin-using patients with type 2 and type 1 diabetes. Vasc Health Risk Manag 2, 203–212.

47.

Ravussin

, Smith

, Mitchell

, Shringarpure

, Shan

, Maier

, Koda

, Weyer

(2009) Enhanced weight loss with pramlintide/metreleptin: An integrated neurohormonal approach to obesity pharmacotherapy. Obesity (Silver Spring) 17, 1736–1743.

48.

da Silva

, Fontes

, Erthal

, Lima

(2016) Amyloidogenesis of the amylin analogue pramlintide. Biophys Chem 219, 1–8.

49.

Qiu

, Zhu

(2014) Amylin and its analogs: A friend or foe for the treatment of Alzheimer’s disease? Front Aging Neurosci 6, 186.

50.

Moore

, Sonar

, Bharadwaj

, Deplazes

, Mancera

(2018) Characterisation of the structure and oligomerisation of islet amyloid polypeptides (IAPP): A review of molecular dynamics simulation studies. Molecules 23, 2142.

51.

Bakou

, Hille

, Kracklauer

, Spanopoulou

, Frost

, Malideli

, Yan

, Caporale

, Zacharias

, Kapurniotu

(2017) Key aromatic/hydrophobic amino acids controlling a cross-amyloid peptide interaction versus amyloid self-assembly. J Biol Chem 292, 14587–14602.

52.

Morrison

, Weiss

(2001) Combinatorial alanine-scanning. Curr Opin Chem Biol 5, 302–307.

53.

Porter

, Bharadwaj

, Groth

, Paxman

, Laws

, Martins

, Verdile

(2016) The effects of latrepirdine on amyloid-β aggregation and toxicity. J Alzheimers Dis 50, 895–905.

54.

Circular Dichroism Spectroscopy, (2021) http://www.ruppweb.org/cd/cdtutorial.htm, Last updated 2021, Accessed on November 20, 2020.

55.

Biancalana

, Koide

(2010) Molecular mechanism of Thioflavin-T binding to amyloid fibrils. Biochim Biophys Acta 1804, 1405–1412.

56.

Rahimi

, Maiti

, Bitan

(2009) Photo-induced cross-linking of unmodified proteins (PICUP) applied to amyloidogenic peptides. J Vis Exp, 1071.

57.

Wei

, Thyparambil

, Latour

(2014) Protein helical structure determination using CD spectroscopy for solutions with strong background absorbance from 190 to 230nm. Biochim Biophys Acta 1844, 2331–2337.

58.

Correcirc

DHA

, Ramos

(2009) The use of circular dichroism spectroscopy to study protein folding, form and function. Afr J Biochem Res 3, 164–173.

59.

Kurtishi

, Rosen

, Patil

, Alves

, Møller

(2019) Cellular proteostasis in neurodegeneration. Mol Neurobiol 56, 3676–3689.

60.

, Zhang

, Chen

, Jiang

, Zheng

(2015) Cross-seeding interaction between β-amyloid and human islet amyloid polypeptide. ACS Chem Neurosci 6, 1759–1768.

61.

Seeliger

, Weise

, Opitz

, Winter

(2012) The effect of Aβ on IAPP aggregation in the presence of an isolated β-cell membrane. J Mol Biol 421, 348–363.

62.

Zhang

, Hu

, Ren

, Chen

, Jiang

, Ma

, Zheng

(2017) Molecular understanding of Aβ-hIAPP cross-seeding assemblies on lipid membranes. ACS Chem Neurosci 8, 524–537.

63.

, Patel

, Jhamandas

(2013) Amylin receptor: A common pathophysiological target in Alzheimer’s disease and diabetes mellitus. Front Aging Neurosci 5, 42.

64.

Jhamandas

, Li

, Westaway

, Yang

, Jassar

, MacTavish

(2011) Actions of β-amyloid protein on human neurons are expressed through the amylin receptor. Am J Pathol 178, 140–149.

65.

Soudy

, Patel

, Fu

, Kaur

, MacTavish

, Westaway

, Davey

, Zajac

, Jhamandas

(2017) Cyclic AC253, a novel amylin receptor antagonist, improves cognitive deficits in a mouse model of Alzheimer’s disease. Alzheimers Dement 3, 44–56.

66.

Ivanova

, Lin

, Lee

, Zheng

, Ramamoorthy

(2021) Biophysical processes underlying cross-seeding in amyloid aggregation and implications in amyloid pathology. Biophys Chem 269, 106507.

67.

Milardi

, Gazit

, Radford

, Xu

, Gallardo

, Caflisch

, Westermark

, Rosa

, Ramamoorthy

(2021) Proteostasis of islet amyloid polypeptide: A molecular perspective of risk factors and protective strategies for type II diabetes. Chem Rev 121, 1845–1893.

68.

Sharma

, Modi

, Sharma

, Deep

(2021) Kinetics theories to understand the mechanism of aggregation of a protein and to design strategies for its inhibition. Biophys Chem 278, 106665.

69.

Nguyen

, Ramamoorthy

, Sahoo

, Zheng

, Faller

, Straub

, Dominguez

, Shea

, Dokholyan

, De Simone

, Ma

, Nussinov

, Najafi

, Ngo

, Loquet

, Chiricotto

, Ganguly

, McCarty

, Li

, Hall

, Wang

, Miller

, Melchionna

, Habenstein

, Timr

, Chen

, Hnath

, Strodel

, Kayed

, Lesné

, Wei

, Sterpone

, Doig

, Derreumaux

(2021) Amyloid oligomers: A joint experimental/computational perspective on Alzheimer’s disease, Parkinson’s disease, type II diabetes, and amyotrophic lateral sclerosis. Chem Rev 121, 2545–2647.

70.

Miller

, Ma

, Nussinov

(2011) Synergistic interactions between repeats in tau protein and Aβ amyloids may be responsible for accelerated aggregation via polymorphic states. Biochemistry 50, 5172–5181.

71.

Clinton

, Blurton-Jones

, Myczek

, Trojanowski

, LaFerla

(2010) Synergistic interactions between Aβ, tau, and β-synuclein: Acceleration of neuropathology and cognitive decline. J Neurosci 30, 7281–7289.

72.

Frieden

(2007) Protein aggregation processes: In search of the mechanism. Protein Sci 16, 2334–2344.

73.

Buell

, Dobson

, Knowles

(2014) The physical chemistry of the amyloid phenomenon: Thermodynamics and kinetics of filamentous protein aggregation. Essays Biochem 56, 11–39.

74.

Kodaka

(2004) Interpretation of concentration-dependence in aggregation kinetics. Biophys Chem 109, 325–332.

75.

Luca

, Yau

, Leapman

, Tycko

(2007) Peptide conformation and supramolecular organization in amylin fibrils: Constraints from solid-state NMR. Biochemistry 46, 13505–13522.

76.

Wiltzius

, Sievers

, Sawaya

, Cascio

, Popov

, Riekel

, Eisenberg

(2008) Atomic structure of the cross-β spine of islet amyloid polypeptide (amylin). Protein Sci 17, 1467–1474.

77.

Baram

, Atsmon-Raz

, Ma

, Nussinov

, Miller

(2016) Amylin–Aβ oligomers at atomic resolution using molecular dynamics simulations: A link between Type 2 diabetes and Alzheimer’s disease. Phys Chem Chem Phys 18, 2330–2338.

78.

Sun

, Kakinen

, Xing

, Faridi

, Nandakumar

, Purcell

, Davis

, Ke

, Ding

(2019) Amyloid self-assembly of hIAPP8-20 via the accumulation of helical oligomers, α-helix to β-sheet transition, and formation of β-barrel intermediates. Small 15, 1805166.

79.

Atsmon-Raz

, Miller

(2015) Non-amyloid-β component of human α-synuclein oligomers induces formation of new Aβ oligomers: Insight into the mechanisms that link Parkinson’s and Alzheimer’s diseases. ACS Chem Neurosci 7, 46–55.

80.

Atsmon-Raz

, Miller

(2016) Molecular mechanisms of the bindings between non-amyloid β component oligomers and amylin oligomers. J Phys Chem B 120, 10649–10659.

81.

von Bergen

, Friedhoff

, Biernat

, Heberle

, Mandelkow

(2000) Assembly of tau protein into Alzheimer paired helical filaments depends on a local sequence motif (306VQIVYK311) forming beta structure. Proc Natl Acad Sci U S A 97, 5129–5134.

82.

Goux

, Kopplin

, Nguyen

, Leak

, Rutkofsky

, Shanmuganandam

, Sharma

, Inouye

, Kirschner

(2004) The formation of straight and twisted filaments from short tau peptides. J Biol Chem 279, 26868–26875.

83.

Barrantes

, Sotres

, Hernando-Pérez

, Benítez

, de Pablo

, Baró

, Ávila

, Jiménez

(2009) Tau aggregation followed by atomic force microscopy and surface plasmon resonance, and single molecule tau-tau interaction probed by atomic force spectroscopy. J Alzheimers Dis 18, 141–151.

84.

Peterson

, Zhou

, Dahlquist

, Lew

(2008) A soluble oligomer of tau associated with fiber formation analyzed by NMR. Biochemistry 47, 7393–7404.

85.

Margittai

, Langen

(2006) Side chain-dependent stacking modulates tau filament structure. J Biol Chem 281, 37820–37827.

86.

Wiltfang

, Esselmann

, Bibl

, Smirnov

, Otto

, Paul

, Schmidt

, Klafki

, Maler

, Dyrks

, Bienert

, Beyermann

, Rüther

, Kornhuber

(2002) Highly conserved and disease-specific patterns of carboxyterminally truncated Aβ peptides 1-37/38/39 in addition to 1-40/42 in Alzheimer’s disease and in patients with chronic neuroinflammation. J Neurochem 81, 481–496.

87.

Brender

, Krishnamoorthy

, Sciacca

, Vivekanandan

, D’Urso

, Chen

, La Rosa

, Ramamoorthy

(2015) Probing the sources of the apparent irreproducibility of amyloid formation: Drastic changes in kinetics and a switch in mechanism due to micellelike oligomer formation at critical concentrations of IAPP. J Phys Chem B 119, 2886–2896.

88.

Stine

, Jungbauer

, Yu

, LaDu

(2010) Preparing synthetic Aβ in different aggregation states. Methods Mol Biol 670, 13–32.

89.

Roberts

, Lind

, Wagen

, Rembach

, Frugier

, Li

, Ryan

, McLean

, Doecke

, Rowe

, Villemagne

, Masters

(2017) Biochemically-defined pools of amyloid-β in sporadic Alzheimer’s disease: Correlation with amyloid PET. Brain 140, 1486–1498.

90.

Esparza

, Wildburger

, Jiang

, Gangolli

, Cairns

, Bateman

, Brody

(2016) Soluble amyloid-beta aggregates from human Alzheimer’s disease brains. Sci Rep 6, 38187.

Novel Amylin Analogues Reduce Amyloid-β Cross-Seeding Aggregation and Neurotoxicity

Abstract

Background:

Objective:

Methods:

Results:

Conclusion:

Keywords

INTRODUCTION

MATERIALS AND METHODS

Peptide preparation

Thioflavin-T aggregation assay

SDS-PAGE and western immunoblotting analysis using photo-crosslinked peptides

Cell viability analysis in SH-SY5Y neuronal cells following peptide treatment

Circular dichroism spectroscopy

Data analysis

RESULTS

Aggregation of IAPP analogues and cross-seeding interaction with Aβ42

Co-oligomerization and size distribution of Aβ42-IAPP analogues mixtures

Secondary structural analysis of Aβ42-hIAPP analogues

Neuroprotective effect of hIAPP analogues against Aβ42-induced toxicity

DISCUSSION

Footnotes

ACKNOWLEDGMENTS

References

Aggregation of IAPP analogues and cross-seeding interaction with Aβ₄₂

Co-oligomerization and size distribution of Aβ₄₂-IAPP analogues mixtures

Secondary structural analysis of Aβ₄₂-hIAPP analogues

Neuroprotective effect of hIAPP analogues against Aβ₄₂-induced toxicity