Self-assembly of proteins to β-sheet rich amyloid fibrils is commonly observed in various neurodegenerative diseases. However, amyloid also occurs in the extracellular matrix of bacterial biofilm, which protects bacteria from environmental stress and antibiotics. Many Pseudomonas strains produce functional amyloid where the main component is the highly fibrillation-prone protein FapC. FapC fibrillation may be inhibited by small molecules such as plant polyphenols, which are already known to inhibit formation of pathogenic amyloid, but the mechanism and biological impact of inhibition is unclear. Here, we elucidate how polyphenols modify the self-assembly of functional amyloid, with particular focus on epigallocatechin gallate (EGCG), penta-O-galloyl-β-d-glucose (PGG), baicalein, oleuropein, and procyanidin B2. We find EGCG and PGG to be the best inhibitors. These compounds inhibit amyloid formation by redirecting the aggregation of FapC monomers into oligomeric species, which according to small-angle X-ray scattering (SAXS) measurements organize into core-shell complexes of short axis diameters 25–26 nm consisting of ~7 monomers. Using peptide arrays, we identify EGCG-binding sites in FapC’s linker regions, C and N-terminal parts, and high amyloidogenic sequences located in the R2 and R3 repeats. We correlate our biophysical observations to biological impact by demonstrating that the extent of amyloid inhibition by the different inhibitors correlated with their ability to reduce biofilm, highlighting the potential of anti-amyloid polyphenols as therapeutic agents against biofilm infections.

Bacterial functional amyloids are evolutionarily optimized to aggregate, so much so that the extreme robustness of functional amyloid makes it very difficult to examine their structure-function relationships in a detailed manner. Previous work has shown that functional amyloids are resistant to conventional chemical denaturants, but they dissolve in formic acid (FA) at high concentrations. However, systematic investigation requires a quantitative analysis of FA’s ability to denature proteins. Amyloid formed by Pseudomonas sp. protein FapC provides an excellent model to investigate FA denaturation. It contains three imperfect repeats, and stepwise removal of these repeats slows fibrillation and increases fragmentation during aggregation, but the link to stability is unclear. We first calibrated FA denaturation using three small, globular and acid-resistant proteins. This revealed a linear relationship between the concentration of FA and the free energy of unfolding with a slope of mFA+pH (the combined contribution of FA and FA-induced lowering of pH), as well as a robust correlation between protein size and mFA+pH. We then measured the solubilisation of fibrils formed from different FapC variants with varying number of repeats as a function of the concentration of FA. This revealed a decline in the number of residues driving amyloid formation upon deleting at least two repeats. The midpoint of denaturation declined with removal of repeats Complete removal of all repeats led to fibrils which were solubilized at FA concentrations 2-3 orders of magnitude lower than the repeat-containing variants, showing that at least one repeat is required for the stability of functional amyloid.

Functional amyloid is produced by many organisms but is particularly well understood in bacteria, where proteins such as CsgA (E. coli) and FapC (Pseudomonas) are assembled as functional bacterial amyloid (FuBA) on the cell surface in a carefully optimized process. Besides a host of helper proteins, FuBA formation is aided by multiple imperfect repeats which stabilize amyloid and streamline the aggregation mechanism to a fast-track assembly dominated by primary nucleation. These repeats, which are found in variable numbers in Pseudomonas, are most likely the structural core of the fibrils, though we still lack experimental data to determine whether the repeats give rise to β-helix structures via stacked β-hairpins (highly likely for CsgA) or more complicated arrangements (possibly the case for FapC). The response of FuBA fibrillation to denaturants suggests that nucleation and elongation involve equal amounts of folding, but protein chaperones preferentially target nucleation for effective inhibition. Smart peptides can be designed based on these imperfect repeats and modified with various flanking sequences to divert aggregation to less stable structures, leading to a reduction in biofilm formation. Small molecules such as EGCG can also divert FuBA to less organized structures, such as partially-folded oligomeric species, with the same detrimental effect on biofilm. Finally, the strong tendency of FuBA to self-assemble can lead to the formation of very regular two-dimensional amyloid films on structured surfaces such as graphite, which strongly implies future use in biosensors or other nanobiomaterials. In summary, the properties of functional amyloid are a much-needed corrective to the unfortunate association of amyloid with neurodegenerative disease and a testimony to nature’s ability to get the best out of a protein fold.

No abstract

Bacterial functional amyloids are evolutionarily optimized to aggregate to help them fulfil their biological functions, e.g. to provide mechanical stability to biofilm. Amyloid is formed in Pseudomonas sp. by the protein FapC which contains 3 imperfect repeats connected by long linkers. Stepwise removal of these repeats slows down aggregation and increases the propensity of amyloids to fragment during the fibrillation process, but how these mechanistic properties link to fibril stability is unclear. Here we address this question. The extreme robustness of functional amyloid makes them resistant to conventional chemical denaturants, but they dissolve in formic acid (FA) at high concentrations. To quantify this, we first measured the denaturing potency of FA using 3 small acid-resistant proteins (S6, lysozyme and ubiquitin). This revealed a linear relationship between [FA] and the free energy of unfolding with a slope of m FA , as well as a robust correlation between protein residue size and m FA . We then measured the solubilisation of fibrils formed from different FapC variants (with varying number of repeats) as a function of [FA]. The resulting m FA values revealed a decline in the number of residues driving amyloid formation when at least 2 repeats were deleted. The midpoint of denaturation declined monotonically with progressive removal of repeats and correlated with solubility in SDS. Complete removal of all repeats led to fibrils which were solubilized at FA concentrations 2-3 orders of magnitude lower than the repeat-containing variants, showing that at least one imperfect repeat is required for the stability of functional amyloid.

No abstract

No abstract