This story was originally published in Plos One Journal:
URL: plosone.org. The content has not been altered
Licensed under Creative Commons Attribution (CC BY) license .
url:https://journals.plos.org/plosone/s/licenses-and-copyright
(C) Plos One [1]
--------------------



42 chains in the presence of cyclic neuropeptides investigated by molecular dynamics simulations

['Min Wu', 'Department Of Electrical', 'Computer Engineering', 'University Of Alberta', 'Edmonton', 'Lyudmyla Dorosh', 'Gerold Schmitt-Ulms', 'Laboratory Medicine', 'Pathobiology', 'University Of Toronto']

Date: None

Alzheimer’s disease is associated with the formation of toxic aggregates of amyloid beta (Aβ) peptides. Despite tremendous efforts, our understanding of the molecular mechanisms of aggregation, as well as cofactors that might influence it, remains incomplete. The small cyclic neuropeptide somatostatin-14 (SST 14 ) was recently found to be the most selectively enriched protein in human frontal lobe extracts that binds Aβ 42 aggregates. Furthermore, SST 14 ’s presence was also found to promote the formation of toxic Aβ 42 oligomers in vitro. In order to elucidate how SST 14 influences the onset of Aβ oligomerization, we performed all-atom molecular dynamics simulations of model mixtures of Aβ 42 or Aβ 40 peptides with SST 14 molecules and analyzed the structure and dynamics of early-stage aggregates. For comparison we also analyzed the aggregation of Aβ 42 in the presence of arginine vasopressin (AVP), a different cyclic neuropeptide. We observed the formation of self-assembled aggregates containing the Aβ chains and small cyclic peptides in all mixtures of Aβ 42 –SST 14 , Aβ 42 –AVP, and Aβ 40 –SST 14 . The Aβ 42 –SST 14 mixtures were found to develop compact, dynamically stable, but small aggregates with the highest exposure of hydrophobic residues to the solvent. Differences in the morphology and dynamics of aggregates that comprise SST 14 or AVP appear to reflect distinct (1) regions of the Aβ chains they interact with; (2) propensities to engage in hydrogen bonds with Aβ peptides; and (3) solvent exposures of hydrophilic and hydrophobic groups. The presence of SST 14 was found to impede aggregation in the Aβ 42 –SST 14 system despite a high hydrophobicity, producing a stronger “sticky surface” effect in the aggregates at the onset of Aβ 42 –SST 14 oligomerization.

Improper folding of proteins causes disorders known as protein misfolding diseases. Under normal conditions most proteins adopt particular folds, which allow them functioning properly. However, for reasons that are not yet fully understood, proteins may misfold and aggregate, forming deposits known as amyloid fibrils, which accumulate in the brain or other tissues. This process affects functioning of the nervous system, gradually causing loss of cognitive abilities. Alzheimer’s disease is one of the most common diseases from this group. A better understanding of the aggregation of peptides implicated in Alzheimer’s disease, known as amyloid beta (Aβ) peptides, may facilitate the development of treatments that ameliorate or prevent the disease. We use detailed molecular dynamics simulations to investigate the influence of somatostatin-14 (SST 14 ), a cyclic neuropeptide that might be involved in the amyloidogenic aggregation of Aβ, on molecular processes occurring at the onset of Aβ aggregation. Results of these simulations explain how the presence of SST 14 might alter pathways of aggregation of Aβ, shedding light upon the possible role of extrinsic factors in the aggregation at a molecular level.

Funding: The work was supported by the Alberta Prion Research Institute (APRI), Projects 201600028 (to HW and GSU) and 201700016 (MS). Philanthropic financial support from the Borden Rosiak family is gratefully acknowledged (to GSU). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

In order to elucidate how SST 14 might influence the onset of Aβ oligomerization, we have performed all-atom molecular dynamics (MD) simulations of model systems containing mixtures of Aβ 42 or Aβ 40 monomeric peptides and SST 14 molecules in explicit water. We also performed similar simulations for mixtures of Aβ 42 peptides and AVP molecules as negative controls. We investigated early stages of aggregation in each system, and analyzed the structure and dynamics of the early self-assembled aggregates. For this purpose, we combined well-established structural analysis tools with a novel essential collective dynamics (ECD) method developed in our group [ 14 , 16 ], which allows to accurately identify persistent correlations of motion (dynamics) in a molecule or supramolecular system without exhaustive conformational sampling, based on an original fundamental concept [ 31 , 32 ]. The method allows analysing dynamics correlations between selected pairs of atoms; characterising main-chain flexibilities; and identifying domains of correlated motion within the same framework. The presence of SST 14 was found to impede the aggregation in the Aβ 42 –SST 14 mixtures despite having a high hydrophobicity. We attribute differences in the structures and dynamics of the Aβ 42 –SST 14 , Aβ 42 –AVP and Aβ 40 –SST 14 systems to distinct tendencies of SST 14 and AVP to (1) interact with specific regions of the Aβ chains; (2) develop hydrogen bonds with Aβ peptides; and (3) expose hydrophilic or hydrophobic groups to the solvent.

The potential role of extrinsic factors in the aggregation and amyloidogenic conversion is a relatively under-explored aspect of immense importance for both the basic understanding of the aggregation process and the rational design of therapeutic strategies [ 23 , 24 ]. In particular, toxic Aβ*56 complexes have been hypothesized to require unknown cofactors for their assembly [ 25 ]. The cyclic neuropeptide somatostatin-14 (SST 14 ) was reported to promote the proteolytic degradation of Aβ through neprilysin induction [ 26 ]. The level of SST 14 in the brain decreases with ageing [ 26 ] and an accelerated decline is found in AD patients [ 27 ], potentially leading to an increase in steady-state Aβ levels. Recent experiments indicate that the cyclic neuropeptide somatostatin-14 (SST 14 ) is the most selectively enriched peptide in human frontal lobe extracts that binds oligomeric Aβ 42 aggregates [ 28 ]. Moreover, SST 14 ’s presence was found to inhibit fibrillization of Aβ 42 in vitro, while promoting formation of smaller oligomers, reminiscent of toxic Aβ*56 complexes [ 28 , 29 ]. Interestingly, Aβ 42 but not Aβ 40 peptides were prone to delayed fibrillization under the influence of SST 14 . This effect was also not observed upon replacement of SST 14 with other cyclic neuropeptides, such as arginine vasopressin (AVP) [ 29 ]. Another recent biochemical study [ 30 ] investigated binding of SST 14 to a specific membrane-associated Aβ 42 tetramer, termed βPFO Aβ(1–42) . This report validated the ability of SST 14 to selectively interact only with oligomeric assemblies of Aβ 42 . Consistent with previous data gathered with soluble Aβ 42 oligomers [ 29 ], the authors reported that the binding interface between SST 14 with βPFO Aβ(1–42) may involve a central tryptophan (W8) within SST 14 . Whereas prior experiments [ 29 ] with soluble Aβ 42 oligomers had pointed toward a contribution of the N-terminal half to the SST 14 binding, the interactions with the membrane associated βPFO Aβ(1–42) oligomer seems to primarily rely on residues 18–20 within the C-terminal half of Aβ 42 . Although at first glance these results may seem contradictory, the N-terminal half of Aβ 42 is required for the formation of βPFO Aβ(1–42) , an observation that would have precluded the ability to detect binding of SST 14 in assays that relied on truncated Aβ 18–42 in prior binding studies. Both studies agreed that binding occurs predominantly to Aβ 42 oligomers and is not observed with oligomers formed from Aβ 40 . Because analyses of [ 30 ] were restricted to a highly purified membrane-associated βPFO Aβ(1–42) , it remains unclear whether an alternative binding pose or binding stoichiometry would be available in soluble Aβ 42 oligomers. Consistent with such a scenario, the binding constant for binding of SST 14 to βPFO Aβ(1–42) was approximately threefold higher than the previously observed Kd for binding of SST 14 to soluble Aβ 42 oligomers.

Contributing to this status quo is a high level of heterogeneity in regards to both the building blocks and the architecture of early oligomeric assemblies. First, the Aβ peptide itself exists in multiple alloforms. Aβ 42 comprises two C-terminal residues, I41 and A42, not present in Aβ 40 . Although the Aβ 40 variant is more abundant than Aβ 42 , the latter is believed to be more pathogenic, primarily on the grounds of a relative increase in Aβ 42 in the brain of AD patients, and higher rates of fibrillization in-vitro [ 4 – 6 ]. Second, Aβ oligomers are an ensemble of highly dynamic assemblies [ 7 , 8 ]. Experiments indicate that the initial small aggregates tend to adopt a largely unstructured morphology [ 9 – 11 ]. Quaternary structures without pronounced alignments of peptide chains were also predicted by molecular dynamics (MD) simulations for dimers of Aβ 40 and Aβ 42 [ 12 – 14 ] and larger multimeric Aβ aggregates [ 15 , 16 ]. Characterizations by a variety of biophysical methods in vitro suggest that initially unstructured oligomers may either undergo a transition into more ordered β-sheet-rich conformations, or mediate formation of β-sheet-rich assemblies [ 9 – 12 ]. However, the majority of early non-fibrillary aggregates dissociate into monomers rather than convert into fibrils [ 7 , 8 ], albeit they are potentially toxic [ 8 , 11 ]. Neither the detailed molecular mechanisms of aggregation, nor the accompanying misfolding or the specific structures and morphologies of the various non-fibrillary and pre-fibrillary assemblies are sufficiently understood. On theoretical grounds, it has been inferred that the aggregation process is driven by a competition of hydrophobic collapse and hydrogen bonding, with the former favoring unstructured conformations, and the latter giving rise to β-sheets [ 17 – 19 ]. The toxicity of non-fibrillary and pre-fibrillary oligomeric assemblies has been hypothetically linked to the exposure of hydrophobic groups and unpaired β-strands at their surfaces [ 9 , 20 ], often referred to as a “sticky surface” effect. Recent modeling studies [ 21 , 22 ] indicate that accumulation of subtle structural perturbations of early aggregates at the onset of the oligomerization process may result in distinct aggregation pathways at later, more advanced stages of Aβ oligomerization. Altogether, both experimental and theoretical evidence suggests that molecular events occurring early in the process of aggregation play a key role in determining both the structure and toxicity of Aβ oligomers [ 14 ].

Alzheimer’s disease (AD) is one of the most devastating neurodegenerative disorders of our time due to its high prevalence in the aging population, the challenges of early diagnosis and the lack of efficient therapeutics. AD is associated with the misfolding and formation of toxic aggregates of the amyloid β (Aβ) peptide [ 1 , 2 ]. It is believed that misfolding in spontaneously formed aggregates of Aβ chains eventually leads to a cascade of self-replicating misfolding events producing β-sheet rich amyloid deposits in the brain. However, mounting evidence suggests that relatively small prefibrillary oligomers rather than mature amyloid fibrils may be the primary toxic assemblies underlying the pathogenesis of the disease [ 2 , 3 ]. Insights into the molecular mechanisms of misfolding and aggregation, as well as co-factors that might influence these processes, remain incomplete.

Results

Main-chain flexibility profiles Within the same ECD framework, main-chain flexibility profiles [16,33–35] can be calculated for all the peptides in each system. High levels of the flexibility descriptor usually correspond to flexible loops and termini, whereas minima indicate more restrained regions. The main-chain flexibility profiles averaged over the first and the last 20 ns of the MD simulations for the Aβ 42 -SST 14 , Aβ 42 -AVP and Aβ 40 -SST 14 systems are shown in Fig 6. PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Fig 6. ECD main chain flexibility profiles of the Aβ 42 -SST 14 –I system (A), Aβ 42 –AVP-II system (B) and Aβ 40 -SST 14 -II system (C). Red and blue lines indicate, respectively, the average flexibilities for the first 20 ns and the last 20 ns of the production MD simulations. The labels on top of the horizontal axis denote the chains, and those at the bottom of the horizontal axis denote the residue numbers for each chain. https://doi.org/10.1371/journal.pcbi.1008771.g006 In the Aβ 42 -SST 14 system, most Aβ chains show a decreased flexibility at the last 20 ns of production MD simulation compared to the first 20 ns (Fig 6A), as one could expect since oligomerization limits mobility of the chains. The flexibility profiles of individual Aβ chains tend to adopt oscillating shapes with several maxima and minima. Although by the end of the simulations N- and C-terminal ends of the chains remain flexible, deep minima developed within 10–12 residues from one or both of the termini in most chains. This is especially the case for chains B, D, and H, which have formed a trimeric self-assembled aggregate (S6A Fig). In central regions of the Aβ chains the flexibility profiles are often M-shaped with two maxima separated by a minimum (chains B, C, and E). However in chains A, D, and G the central regions exhibit flexible loops located at the periphery of the respective aggregates. Minima of main-chain flexibility often coincide with locations of stable β-strands. In particular, this is the case for the flexibility minima in chains A, B, C, F, G and H, where stable β-strands are found (see S6A Fig). However, not all flexibility minima are necessarily associated with the presence of β-structure. For example, we do not observe stable β-sheet content in chains D and E, which exhibit pronounced minima of main-chain flexibility. A trend of adopting lower main-chain flexibility in central regions in comparison to the N- and C-termini is observed in most SST molecules in the Aβ 42 -SST 14 system. Moreover, at the end of the simulation the central regions of the SST molecules I, J, L, and M exhibit comparable flexibilities with parts of the Aβ chains B, D, H, A, and C, with which they directly interact. Interestingly, the SST molecules K, O, and P adopted a higher flexibility during the last 20 ns in comparison to the beginning of the simulation. An analysis of the trajectory indicated that initially these molecules were surrounded by neighbouring Aβ 42 chains (S1A Fig), which somewhat limited their motion. By the end of the simulations the Aβ 42 chains and other SST 14 molecules formed three aggregates, whereas the molecules K, O, and P remained free. Eventually, remaining in solution resulted in a less constrained motion for these SST molecules by the end of the simulation. Fig 6B depicts the main-chain flexibility profile for the Aβ 42 -AVP system. After the equilibration most chains exhibit relatively uniform flexibility patterns, whereas by the end of the simulation pronounced differences emerged across the various chains. Both termini of chains B and C, as well as the N-terminus of chain F, and the C-terminus of chain G adopted higher flexibility levels in the course of the simulation, which can be explained by their location at the periphery of the self-assembled aggregate. In contrast, extensive regions of chains A, E, F, and G, which are located at the core of the three domains of correlated motion (Figs 5 and S6B), exhibited low flexibility. Flexibilities of AVP molecules tended to follow those of the Aβ chains with which they interacted over the last 20 ns of the simulation. The main-chain flexibility profile of the Aβ 40 -SST 14 system is shown in Fig 6C. Extensive regions of Aβ chains A, B, D, G and H, which are located in the interior regions of the aggregate, developed a decrease in flexibility during MD simulation. In contrast, chain F and most of chain C adopted relatively high flexibilities due to their less constrained positions. Central regions of SST 14 molecules K, M, O, and P exhibited relatively low main-chain flexibility over the last 20 ns of the MD simulation. These chains were actively involved in the oligomerization process, which resulted in their insertion inside the self-assembled aggregate where they became a part of the largest domain of correlated motion (see also Figs 5C and S6C). SST 14 molecules I and M, which attached to the free Aβ chain F, exhibited high flexibility during last 20 ns of the simulations. Table 3 lists the main-chain flexibility of the Aβ 42 or Aβ 40 C-termini averaged over eight Aβ chains in each of the three MD trajectories in each system. The table also lists the results of averaging over the three MD trajectories for each system. Further details of C-terminal flexibilities in individual Aβ chains can be found in S4 Table in the Supporting Information. As Table 3 illustrates, the average flexibility of the C-termini is greater in the AVP-containing systems than in the SST-containing systems. AVP-containing systems also exhibit the greatest variation of C-terminal flexibility across individual trajectories. Interestingly, the average flexibility of the C-termini in the Aβ 40 -SST 14 system is very close to that in the Aβ 42 -SST 14 system (Table 3), despite the absence of two hydrophobic residues at the C-terminus of Aβ 40 and the fact that the Aβ 40 C-termini are often buried inside the Aβ 40 -SST 14 aggregates. PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Table 3. ECD main-chain flexibilities of the C-termini averaged over eight Aβ 42 or Aβ 40 chains in each trajectory over the last 20 ns of MD simulations, their averages over three trajectories for each system, and standard deviations of the data. https://doi.org/10.1371/journal.pcbi.1008771.t003

[1] Url: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1008771

(C) GlobalVoices
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.