A molecular mechanism for a general amyloid re-modelling activity

Lead Research Organisation: University of Sheffield
Department Name: Molecular Biology and Biotechnology


Brain degeneration is one of the principal reasons why our quality of life is eroded as we age. The appearance of diseases like Alzheimer's and Parkinson's is strongly linked to the formation of amyloid aggregates in the brain. Amyloids are pathogenic fibrous assemblies of otherwise normal protein molecules. The removal of these assemblies has been shown to prevent disease progression in model animals and even lead to cognitive improvements. Strong contenders for the design of disease-modifying agents are molecules that are able to modify existing amyloid fibrils as well as prevent their formation. However, our current understanding of how these molecules work is insufficient to direct a powerful knowledge-based drug discovery programme. Our proposal will address this deficiency by using state-of-the-art structural biology techniques to unravel the mechanism behind the activity of these molecules. We have chosen as our model agent the protein G3P because of its unusually efficacious effect in animal studies and its ability to tackle amyloids of many different types. The latter argument is particularly important as more than one kind of amyloid is often found in people suffering with these degenerative conditions.

Of key importance will be the definition of the atomic surfaces involved as this information will directly feed into future drug design processes. Using nuclear magnetic resonance spectroscopy, we can map changes in the surface of the molecule using a novel hydrogen atom labelling assay. To complement this work, a molecular footprint for the amyloids on the surface of the G3P will also be measured using "FPOP", which will decorate the surface of the complex with highly reactive hydroxyl radicals, readily detected by mass spectrometry. Finally, we will also produce an array of different atomic surfaces on the G3P using genetic engineering and, by a process of elimination, work out the key components of its activity. Not only is G3P able to stick to amyloids but we suspect it uses its ability to act as a "molecular packman" to then open its mouth and expose a more powerful interface responsible for the dismantling activity. We will investigate whether this is true by exploiting the available extremely detailed knowledge of how this protein uses this ability to operate in bacterial infection.

Compared with small molecule effectors, the larger size of the G3P molecule will also allow us to visualise how it interacts with the amyloid fibrils directly using the imaging technology that the University of Sheffield has recently invested in. The relative orientations of molecules within the amyloid assemblies have the potential to reveal with near atomic detail how the G3P grabs the amyloid for dismantling. We will explore what happens to these structures as they become re-modelled into less toxic molecular assemblies. Our single particle analysis techniques are ideal for observing these mixed populations.

Finally, we will investigate the key steps G3P is affecting within the molecular processes defining amyloid assembly and disassembly. We will do this by carefully measuring how this amyloid modifying agent affects the different populations of the precursor molecule, then the partially assembled forms referred to as "oligomers" and also the mature amyloidogenic protein structures as we incubate them together. By feeding the data into existing mathematical models, we will start visualising what its activity consists of: for example, does it promote the breakage of fibrils by straining the structure or does it favour dissociation of single protein molecules one at a time? By the end of the project, we will have details of the molecular events involved and be able to apply this knowledge not only to improving the design of future therapeutics but also to an improved regulation of fibril formation by novel amyloid-based biomaterials.

Technical Summary

Protein misfolding and aggregation is responsible for some of the most prevalent aging diseases in humans. Fibrillar "amyloid" species accumulating as extracellular plaques or intracellular tangles act as sinks for the more neurotoxic soluble species and at later stages of the disease dominate the conditions. Drugs able to target only soluble species have a limited use and so we have chosen to study an amyloid binding and modifying agent, G3P. Despite the existence of many candidate drugs, our understanding of how these exert their activity at the molecular level is poor and we propose here to visualise this process directly using our state-of-the-art "Imagine" centre. We will study the molecular interfaces of G3P responsible for the activity using a combination of site-directed mutagenesis and labelling approaches including hydrogen-exchange (NMR) and Fast Photochemical Oxidation of Proteins (MS). G3P is a 25KDa two domain protein which exhibits archetypal domain movements on substrate binding. We will define the relevance of key residues involved in domain dynamics to amyloid binding activity. We will then interrogate the changes in the structure of the amyloids (Abeta(1-42) and alpha-synuclein) directly using state-of-the-art Cryo-EM and time-resolved AFM. Finally, we will investigate the kinetics and thermodynamics of the re-modelling reaction of beta-amyloid as well as exploring the impact of G3P on the assembly process. As well as standard methods, we propose to detect changes in the population of different intermediate species directly using asymmetric flow fractionation methods uniquely suited to the properties and molecular weight range of these reactions. Using current mathematical models, we will deduce the key stages of its inhibitory and remodelling activities. We expect our results to contribute to a step change in the design of crucial disease modifying drugs for degenerative diseases.

Planned Impact

Our quality of life as we age is severely impaired by degenerative conditions which include Alzheimer's, Parkinson's and other notable diseases. The aggregation of different proteins into amyloids is responsible for this, but can also drive a number of other lesser known processes such as bacterial biofilm formation and HIV infectivity. Our proposal will address fundamental questions regarding how protein amyloid formation can be regulated, and hence make a step change in the ability of the scientific community to control this reaction and importantly, prevent toxic side reactions. Specifically, we will determine how an amyloid re-modelling agent binds to amyloids, what the molecular determinants of this apparent broad specificity are and how the re-modelling activity takes effect on the amyloid assemblies.

These studies will provide a firmer foundation upon which to
1. accelerate progress in knowledge-based anti-amyloid drug design for use in Alzheimer's disease, Parkinson's disease and many other conditions including bacterial infection and HIV
2. provide material scientists working with beta-sheet rich protein fibres with better tools to regulate their toxicity and control their formation.

In order to ensure successful knowledge transfer of the results of our study, we will engage regularly in discussions with individuals from outside of the academic sector. We will achieve this in the following ways:
- We will choose to publish in journals and attend conferences that are designed to attract a broad audience and have major readership/attendance from industry
- We will continue participation in joint academic-industrial forums including collaboration with our local SITraN Institute (Sheffield Institute for Translational Neuroscience) and our Science and Medicine gateways as well as BBSRC funded networks (such as Bioprocessing, Metals in Biology, Biocatalyst discovery, Natural Products Discovery and Bioengineering, which are supported by Merck, BASF, Unilever, Pfizer, Codexis, Syngenta and Bayer).
- We have planned regular meetings with researchers in industry, in particular with Astrazeneca (monthly), C4X Discovery (monthly), Eli Lily (annual) and GSK (six monthly)
- The work proposed will be a source of useful discussions with NeuroPhage Pharmaceuticals with whom the research originally started and from whom we will receive information regarding clinical trials
- All investigators engage regularly with members of the public through public lectures as well as café scientifique, national science week and further outreach activities
- Last but not least, we will continue our support for the University of Sheffield's local "Dementia Interest Group" who organises regular events and publications aimed at local patient groups and their families who live with dementia.

The strength of these meetings will arm us with the correct information to bridge any gaps between translation to industry and our work. This will be done by engaging with current collaborators and new contacts which we have now identified within the international research community.


10 25 50
publication icon
Taylor AIP (2020) A two-step biopolymer nucleation model shows a nonequilibrium critical point. in The Journal of chemical physics

Description The aim of this work was to examine the molecular mechanism of an amyloid re-modelling agent, G3P. This protein is found on the surface of bacteriophage and was discovered to have broad anti-amyloid activity, potentially as part of a role in enabling the phage to traverse bacterial biofilms. The principal components of these secreted bacterial matrices are often amyloid-like in origin. G3P is currently in clinical trials as a therapeutic for neurodegeneration. It has demonstrable efficacy in animal models of Alzheimer's disease, but has broader activity on a range of amyloid conditions.

1. We have identified using biophysical methods how the amyloid re-modelling agent, G3P, binds amyloid fibrils. This objective has been met by using a novel competition-based hydrogen-exchange assay in combination with NMR. This challenging experiment has shown that G3P binding is governed by a proline molecular switch leading to domain opening. The pattern of protection in the amyloid-bound state is unlike that of G3P in any other previously characterised form, suggesting binding occurs in a conformation distinct from those populated by this protein during phage infection.

2. We have also identified how G3P binds amyloid fibrils using structural methods. We followed the time course of Abeta (1-42)-G3P fibril disassembly by Atomic Force Microscopy (AFM) in aqueous conditions. By returning to the same fibrils over the time course, we observe breakage along the length of the fibrils. This validates the binding sites of G3P as lateral to the fibrils but also suggests a fragmentation mechanism for de-fibrillisation, followed by the release of material into solution in an alternative conformation, no longer able to adhere to the mica and with a modified force profile. Electron microscopy supports this observation, showing binding of G3P to non-fibrillar forms.

3. We have identified the mechanism by which G3P remodels amyloid fibrils. Time courses of fibre re-modelling were recorded as a function of G3P and Abeta(1-42) amyloid concentration, temperature and buffer conditions. From fibril specific dye-binding assays, including ThioT and a newly developed X-34 assay, we have observed that the isolated domains of G3P have distinct activities: while the N1 domain inhibits primary nucleation, the N2 domain inhibits secondary nucleation. Together with data on the whole protein, we predict that fragmentation of the amyloid fibrils occurs through initial binding to the lateral edge of the fibrils, followed by insertion into the fibrils driven by the high affinity of the N1 domain for a newly formed fibril end.

Additional work leading to the above results:
In planning the grant we had proposed to use existing kinetic models of amyloid formation to extract mechanistic information on the re-modelling activity of G3P. We soon found these to be inadequate for application to G3P and have invested considerable time in introducing intermediate species explicitly into our kinetic modelling. This has yielded much needed insight into the importance of the dynamics of conversion in explaining the diversity of oligomer populations. This has important repercussions with respect to our ability to quantify toxic amyloid species. In addition, while developing the experimental side of these kinetic assays, we have been able to propose new protocols for the successful experimental manipulation of amyloid beta peptide, and document the impact of different experimental environments on this system. As well as existing dye-binding assays, we have developed the use of asymmetric flow field flow fractionation techniques (AF4) coupled with multi-angle light scattering (MALS) for the quantification of both soluble and insoluble protein assemblies. Armed with this improved understanding of amyloid growth reactions, we are now able to model data of the activity of G3P on Abeta assembly but have provided important new experimental and analytical tools for the analysis of other promising candidates.
Exploitation Route This study provides a deeper understanding of the molecular detail of an amyloid re-modelling activity. The dual action of the different binding sites of G3P and the molecular switch activity which allows G3P to efficiently insert itself irreversibly into the amyloid until it has been dismantled provides us with insight into how such a mechanism can occur safely without shedding potentially toxic "unguarded" amyloid fragments. This argues for a more in depth search for proteins with similar roles, with potentially more suitable therapeutic profiles. At the same time, our extensive understanding of the dynamic properties of G3P will provide us with the potential to engineer the protein further and improve its properties with respect to different therapeutic targets.
Our work has highlighted the need for better mathematical modelling of amyloid assembly and dis-assembly reactions. We have made considerable inroads into the development of a new model which will provide much needed insight into the nature of oligomeric intermediates (manuscript currently on arXiv). Additionally, we expect a timely publication of our work into the development of consistent, reproducible experimental systems in this area of science.This will considerably improve the quality and impact of published data in this area and make it more accessible to the uninitiated.
Sectors Healthcare,Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology

Description Our collaboration with the company Postnova on applying asymmetric flow field-flow fractionation (AF4) techniques to research into Alzheimer's disease has been publicised on several online media fora and is serving to apply this novel technology in a new area. Our preliminary results on the use of hydrogen exchange and NMR to understand the binding of G3P to amyloid beta fibrils have been presented at a research conference in the United States which is attended by the private sector as well as academic experts from around the world (FASEB summer research conference, June 2017).
First Year Of Impact 2017
Sector Pharmaceuticals and Medical Biotechnology
Impact Types Economic

Description Celltech Limited
Amount £199,681 (GBP)
Funding ID KTP012141 
Organisation Innovate UK 
Sector Public
Country United Kingdom
Start 04/2020 
End 03/2022
Description Application of asymmetric flow field-flow fractionation (AF4) technology to therapeutic development in neurodegenerative disease 
Organisation Postnova Analytics
Country Germany 
Sector Private 
PI Contribution The novelty of the application of AF4 technology to neurodegeneration was of interest to the manufacturers and they decided to loan a multi-angle light scattering (MALS) detector in order to allow us best to carry out the work in the grant proposal. This has led to some publicity on the work we do and resulted in a new KTP with pharma (see grant awards).
Collaborator Contribution Postnova have actively supported the technical developments we have made and have ensured that we are supported at each stage of the project. This has included in house seminars where we have been able to share good practice with other users of this delicate technology.
Impact A concrete output is the existence of a funded Knowledge Transfer Partnership and the ongoing agreement to loan part of the equipment.
Start Year 2017