Abstract
In a recent study, Hervas et al. extracted Orb2 fibrils, that are involved in long-term memory formation, from Drosophila brains, characterised their function, and determined their structure using cryo-electron microscopy (cryo-EM). The fibrils show a remarkable resemblance to amyloid β (Aβ) fibrils associated with Alzheimer’s disease, highlighting the subtle difference between functional and dysfunctional amyloid.
Amyloid, first observed in 1639, typically shows a cross β-sheet fibrillar structure and was initially associated with tissue damage and human disease [1]. The best known examples of amyloid diseases are neurodegenerative disorders such as Alzheimer’s disease (Aβ/Tau), Parkinson’s disease (α-synuclein), and Huntington’s disease (huntingtin with polyQ expansion) [1]. However, some amyloid fibrils have biological functions [2–4]. In 2000, Wösten and de Vocht pioneered the term 'functional amyloid' based on their work on hydrophobins, a family of small proteins in fungi that self-assemble at hydrophobic/hydrophilic surfaces and that are important for growth and development [3]. Since then, functional amyloids have been identified in bacteria (e.g., curli in biofilm formation), fungi (e.g., HET-s, programmed cell death), yeast (e.g., Ure2p, nitrogen catabolism), plants (e.g., luminidependens, regulating flowering), arachnids (e.g., spidroin, spider silk), and mammals (e.g., pmel17, melanin formation), highlighting their wide range of functional roles [2–4]. Pmel17, the first functional amyloid protein identified in humans, was described only 15 years ago. Today, functionally important amyloids in humans are known to be involved in diverse functions including hormone storage (peptide hormones), regulating necrosis (receptor interacting protein1/RRIP3), and removing damaged sperm (semenogelin proteins 1/SEM2) (reviewed in [4]).
One family of functional amyloids includes RNA-binding proteins of the cytoplasmic polyadenylation element binding protein (CPEB) family [5]. In several organisms, including humans, CPEB is located in germ cells and the central nervous system, and is involved in controlling polyadenylation and translation, crucial for cell division during development, but also in the formation and storage of long-term memories [5]. For its function in memory, CPEB has been shown to be required to self-assemble into prion-like self-propagating fibrillar aggregates [6]. These fibrils bind to mRNAs and control the translation of synaptic proteins, changing the protein composition within the neuron and, by stabilising enhanced synaptic activity, allowing long-term memory formation and storage. Focusing on the activity and structure of fibrils of the CPEB orthologue, Orb2, from the brain of the fruit fly Drosophila melanogaster, Hervas et al. [7] started to answer the fascinating, and key, question of how ‘molecular memory’ is made.
Although Orb2 is monomeric in Drosophila early embryos, Hervas et al. [7] extracted Orb2 fibrils (as well as oligomeric and monomeric Orb2) from adult fly brain. The authors found that, although extracted monomeric Orb2 supresses translation of a specific mRNA by binding to a protein GC13928 in vitro, binding of oligomeric or fibrillar Orb2 to a different protein, CG4612, activates translation. For example, Orb2 oligomers and fibrils positively regulate the translation of mRNA encoding tequila, a protein required for long-term memory formation, whereas the monomer supresses tequila production, explaining how translation regulation is brought about by this CPEB protein. Hence, the function of Orb2 depends crucially on its state of assembly and the repertoire of mRNAs being translated.
Hervas et al. [7] also showed that Orb2 fibrils have the biophysical properties of amyloid – the ability to seed (catalyse) fibril growth of monomers, to bind the amyloid-specific fluorescent dye thioflavin-T and amyloid-specific antibody OC, and to be resistant to proteases, heat, and SDS treatment. The authors then exploited the power of cryo-EM to solve a high-resolution structure (2.6 Å) of Orb2 fibrils isolated from fly brain. Their beautiful atomic structure revealed that Orb2 fibrils comprise three identical protofilaments related by C3 symmetry that form a triangular cross-section (Figure 1). In the fibril core, each protofilament adopts a hoop-like conformation with a wide turn, linking two β-strands that stack into a cross-β structure typical of amyloid. Only 31 residues (176–206) of the 704 residue protein form the amyloid core, whereas the less well defined protein density in the structure probably corresponds to the protein interaction domain and the RNA-recognition motif, which would allow these functional regions to bind to other molecules to fulfil their physiological function. Interestingly, the fibril core is hydrophilic owing to the 20 glutamines and seven histidines within the structure. The inter-protofilament polar interface is built of three glutamines and two histidines that form multiple hydrogen bonds, thereby stabilising the fibril structure.
Figure 1. Comparison between the Functional Amyloid Fold of Orb2 (PDB: 6VPS) (Left) versus the Toxic Fibril Architecture of Aβ40 (PDB: 2M4J) (Right).
Whereas Orb2 is involved in long-term memory formation, Aβ40 is involved in the neurodegenerative disorder Alzheimer’s disease. Both structures (one from Drosophila brain and one from human brain) have C3 symmetry with three protofilaments (highlighted in yellow, green, and pink) that form the mature fibril. Whereas the core of Orb2 is mainly hydrophilic (blue), the Aβ40 core is hydrophobic (red). In addition, for Aβ40 nearly the whole protein is structured in the fibril core (red bar), whereas for Orb2 the majority of the protein sequence is dynamically disordered (grey bar) allowing functionally important interactions with other molecules.
Comparing the Orb2 fibril architecture determined by Hervas et al. with previously published structures of functional and disease-related amyloid fibrils demonstrates the peculiarity of this three-protofilament amyloid structure. The literature is dominated by amyloid structures built of two protofilaments twisted around each other [8]. Two other examples of a C3 symmetric fibril with three protofilaments have been reported to date, namely formed from presynaptic Aβ40 in vitro [9] and by elongation of Aβ40 seeds from a patient with Alzheimer’s disease [10] (Figure 1). Interestingly, whereas Orb2 aggregates help to form memory, Aβ is involved in Alzheimer’s disease, a type of dementia that destroys memory and thinking processes [11]. Can one explain these contrary activities for such similar fibril architectures based on their molecular structure? Focusing on the differences reveals characteristics that might possibly explain their opposing effects. (i) Although Orb2 fibrils have a hydrophilic core (formed from glutamine and histidine residues), the core of Aβ40 fibrils is hydrophobic (formed mainly of valine, methionine, glycine, and isoleucine). Changes in pH would destabilise the functional amyloid core, allowing fibril depolymerisation and reformation, thus enabling the dynamics required for memory formation. By contrast, the uncharged core of toxic amyloid would be less affected by pH differences. (ii) Both fibril structures show ~30 residues within the inflexible, structured fibril core, hence in Aβ40 only the first ~4–10 residues are unstructured, whereas in Orb2B (the longest Orb2 isoform, primarily involved in fibril formation and memory), N650 residues are dynamically disordered, enabling interactions with mRNAs and proteins and controlling the translation of synaptic proteins. (iii) Orb2 fibril extraction from ~3 million flies resulted in one structure. By contrast, Aβ amyloid isolated from Alzheimer’s brains revealed variations between patients that were dependent on disease phenotype [11]. Hence, to fulfil a physiological function, fibrils might need to adopt a uniquely defined structure across individuals. It should also be remembered that function and toxicity are not black and white: for example, Aβ40 fibrils have been reported to protect against microbial infection, suggesting that they also have a functional role [4].
In summary, the report by Hervas et al. highlights the commonality in architecture of both functional and pathogenic amyloid, and reveals how the details of the fibril structure are crucial in spotting the differences between fibrils that have a physiological function and pathogenic fibrils.
Acknowledgements
We acknowledge, with thanks, funding from the Wellcome Trust (215062/Z/18/Z to S.M.U., and 204963 to S.E.R.).
References
- 1.Iadanza MG, et al. A new era for understanding amyloid structures and disease. Nat Rev Mol Cell Biol. 2018;19:755–773. doi: 10.1038/s41580-018-0060-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Perrett S, et al. Functional amyloid: widespread in nature, diverse in purpose. Essays Biochem. 2014;56:207–219. doi: 10.1042/bse0560207. [DOI] [PubMed] [Google Scholar]
- 3.Shanmugam N, et al. Microbial functional amyloids serve diverse purposes for structure, adhesion and defence. Biophys Rev. 2019;11:287–302. doi: 10.1007/s12551-019-00526-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jackson MP, Hewitt EW. Why are functional amyloids non-toxic in humans? Biomolecules. 2017;7:71. doi: 10.3390/biom7040071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Richter JD. CPEB: a life in translation. Trends Biochem Sci. 2007;32:279–285. doi: 10.1016/j.tibs.2007.04.004. [DOI] [PubMed] [Google Scholar]
- 6.Si K, Randall ER. The role of functional prion-like proteins in the persistence of memory. Cold Spring Harb Perspect Biol. 2016;8:a021774. doi: 10.1101/cshperspect.a021774. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hervas R, et al. Cryo-EM structure of a neuronal functional amyloid implicated in memory persistence in Drosophila . Science. 2020;367:1230–1234. doi: 10.1126/science.aba3526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gallardo R, et al. Amyloid structures: much more than just a cross-β fold. Curr Opin Struct Biol. 2020;60:7–16. doi: 10.1016/j.sbi.2019.09.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Paravastu AK, et al. Molecular structural basis for polymorphism in Alzheimer’s β-amyloid fibrils. Proc Natl Acad Sci U S A. 2008;105:18349–18354. doi: 10.1073/pnas.0806270105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lu J-X, et al. Molecular structure of β-amyloid fibrils in Alzheimer’s disease brain tissue. Cell. 2013;154:1257–1268. doi: 10.1016/j.cell.2013.08.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Qiang W, et al. Structural variation in amyloid-β fibrils from Alzheimer’s disease clinical subtypes. Nature. 2017;541:217–221. doi: 10.1038/nature20814. [DOI] [PMC free article] [PubMed] [Google Scholar]

