For more than three decades protein aggregates have been known to be implicated in neurodegenerative diseases, like Parkinson’s and Alzheimer’s diseases¹. α-Synuclein, amyoid-β (Aβ) and tau are examples of proteins that self-aggregate in cross β-sheet motifs and have been found in the brains of people with these diseases². Protein aggregation has been extensively studied in vitro using recombinant proteins and so we understand the aggregation process fairly well and have characterised many of the different species formed. On the other hand, the precise mechanism by which protein aggregates lead to progressive loss of neuronal cells and result in subsequent pathophysiologic effects like dementia and movement disorders remains poorly understood. In previous studies, using cell models, it was shown that the small early stage species (<50 nm) are cytotoxic, but so far no one has been able to capture them in the brain of animal models or human biofluids. Their low abundance, small size and heterogeneity in conformation and composition^3-5 represent major technological challenges. The development of new strategies capable of selectively isolate the small toxic protein aggregates implicated in neurodegenerative diseases and identification of their exact molecular composition is fundamental for the development of early diagnostic tools. In turn, offering early diagnosis to patients will help implementation of disease-modifying therapies and improve patient selection criteria for drug trials.

Our aim was to design an unbiased method to capture and characterize all the cross β-sheet aggregates present in human biofluids and determine their composition (see Fig. A). To this end we developed a new dimeric bio-inspired molecule that binds to small misfolded protein structures and used cerebrospinal fluid (CSF) as the model biofluid and α-synuclein as the main model for protein aggregates.

 A new class of molecule to ‘find a needle in a haystack’

Antibodies are a class of molecules widely used for immunoprecipitation due to their high affinity towards complementary proteins, but since they target a particular linear sequence, they are not suitable for capture and isolation of complex aggregates potentially composed of different proteins. Our new molecule, CAP-1 was inspired by previously known β-sheet binding molecules, thioflavin-T (ThT) and Pittsburgh compound B (PiB) and was synthesised as a dimer (to increase avidity⁶) coupled to a biotin moiety for surface immobilization (see Fig. A). We used different optical techniques to characterize CAP-1 and concluded that CAP-1 is highly selective to the cross β-sheet structures present in early stage species, henceforth referred as oligomers (t=8h) and fibrils (t=24h), but importantly, not to monomers. Also, CAP-1 affinity towards amyloids (α-synuclein) is 100-fold higher than for ThT supporting its potential as an effective pulldown molecule and finally, the compatibility with single molecule microscopy widens the range of tools that can be used to characterise protein aggregates. The latter was not anticipated, given CAP-1 similarity with PiB (non-fluorescent) but revealed to be particularly useful.

Amyloid precipitation (AP) – a new method to isolate amyloids

Following initial characterization of CAP-1 we incubated magnetic streptavidin-beads with CAP-1 and used them to pulldown amyloids from a solution containing α-synuclein aggregates, i.e amyloid precipitation (see Fig. A). Using techniques such as fluorescence and electron microscopy and AFM we demonstrated that AP captures virtually all protein aggregates in solution, both oligomers and fibrils, but not monomers. Also, we showed that it is possible to recover aggregates from the beads after AP which can then be studied using other methods. Using mass spectrometry to quantify the number α-synuclein peptides pulldown after AP with or without CAP-1, we found 5 to 13-fold more peptides, depending on the peptide, in the presence of CAP-1. This encouraging set of results using recombinant protein pointed to the likely success of AP in capturing toxic species in human biofluids.

Proof-of-concept - AP can pulldown oligomers from biofluids

To determine whether AP could detect and even enrich aggregates from cerebrospinal fluid (CSF), we started by spiking known amounts of α-synuclein in CSF. Working with CSF was quite challenging, the sample volumes available are very limited and being a complex biofluid made of more than two thousand different proteins⁷, increases the chances of unspecific binding. Results were exciting and showed that AP captured almost no monomers (0.6% from initial concentration) and approximately 50% of small aggregates confirming the specificity of CAP-1 in capturing protein aggregates, but not monomers, within complex biofluids. Finally, to assess the efficacy of AP in removing toxic species we used a sensitive membrane permeability assay previously developed and validated in the lab⁸. If the CSF (or supernatant after AP) contains amyloids that cause membrane permeability, Ca²⁺ ions enter the liposome, resulting in increased fluorescence signal.  The decrease in Ca²⁺ influx after AP demonstrated successful removal of most toxic species, in both control CSF and PD CSF.

In a nutshell, AP is a structure-based approach that captures small and toxic protein aggregates present in biofluids. To the best of our knowledge this is the first unbiased method to pulldown naturally occurring complex aggregates made of different proteins. The compatibility with mass spectrometry and on-bead digestion paves the way to unravelling the exact molecular species responsible for neurodegeneration in humans, which will hasten development of direct and robust early diagnosis methods.

For more details and further discussions please see our article: https://www.nature.com/articles/s41557-022-00976-3

References

1 Selkoe, D. J. The molecular pathology of Alzheimer’s disease. Neuron 6, 487–498 (1991).

 2 Goedert, M. Alzheimer’s and Parkinson’s diseases: the prion concept in relation to assembled Aβ, tau, and α-synuclein. Science 349, 61–69 (2015). 

3 Bruggink, K. A., Müller, M., Kuiperij, H. B. & Verbeek, M. M. Methods for analysis of amyloid-β aggregates. J. Alzheimers Dis. 28, 735–758 (2012).

4 Schuster, J. & Funke, S. A. Methods for the specific detection and quantitation of amyloid-β oligomers in cerebrospinal fluid. J. Alzheimers Dis. 53, 53–67 (2016).

5 Soto, C. & Pritzkow, S. Protein misfolding, aggregation, and conformational strains in neurodegenerative diseases. Nat. Neurosci. 21, 1332–1340 (2018).

6 Qin, L., Vastl, J. & Gao, J. Highly sensitive amyloid detection enabled by thioflavin T dimers. Mol. Biosyst. 6, 1791–1795 (2010).

7 Guldbrandsen, A. et al. In-depth characterization of the cerebrospinal fluid (CSF) proteome displayed through the CSF proteome resource (CSF-PR). Mol. Cell. Proteomics 13, 3152–3163 (2014).

 8 Flagmeier, P. et al. Ultrasensitive measurement of Ca2+ influx into lipid vesicles induced by protein aggregates. Angew. Chem. Int. Ed. 56, 7750–7754 (2017).