Inhibition of Staphylococcus aureus biofilm-forming functional amyloid by molecular tweezers

Abstract

Biofilms are rigid and largely impenetrable three-dimensional matrices constituting virulence determinants of various pathogenic bacteria. Here, we demonstrate that molecular tweezers, unique supramolecular artificial receptors, modulate biofilm formation of Staphylococcus aureus. In particular, the tweezers affect the structural and assembly properties of phenol-soluble modulin α1 (PSMα1), a biofilm-scaffolding functional amyloid peptide secreted by S. aureus. The data reveal that CLR01, a diphosphate tweezer, exhibits significant S. aureus biofilm inhibition and disrupts PSMα1 self-assembly and fibrillation, likely through inclusion of lysine side chains of the peptide. In comparison, different peptide binding occurs in the case of CLR05, a tweezer containing methylenecarboxylate units, which exhibits lower affinity for the lysine residues yet disrupts S. aureus biofilm more strongly than CLR01. Our study points to a possible role for molecular tweezers as potent biofilm inhibitors and antibacterial agents, particularly against untreatable biofilm-forming and PSM-producing bacteria, such as methicillin-resistant S. aureus.

Keywords: MRSA; PSMa1; Staphylococcus aureus; amyloid peptides; antibacterial; biofilm; functional amyloid; molecular tweezer; phenol-soluble modulins.

Figure 1.. Effects of the molecular tweezers on bacterial viability

(A) Molecular structures of the tweezers. The compounds are partially protonated at physiologic pH. (B-E) Dose-dependent effects of CLR01, CLR03, and CLR05 on the viability of (B) S. aureus, (C) B. subtilis, (D) E. coli and (E) P. aeruginosa. Bacteria were grown for ~6 h in the absence or presence of increasing tweezer concentrations, and cell concentrations were determined by measuring absorbance at 600 nm. Data are presented as the mean ± SD of three replicates.

Figure 2.. Inhibition of S. aureus biofilms formation by the molecular tweezers

(A) 50 μM CLR01, CLR03, or CLR05 were added to S. aureus cultures at t= 0. Biofilm images showing viable cells stained in green using the LIVE/DEAD BacLight Bacterial Viability Kit. The images were recorded after 24-h incubation using confocal fluorescence microscopy (excitation/emission: 485 nm/498 nm). Scale bars, 50 μM. (B) Scanning electron microscopy (SEM) micrographs of S. aureus biofilms grown in 10% mouse serum for 24 h. Scale bars in the magnified images represent 1 μM. (C) Photographs of the wells in which S. aureus cells were cultured overnight in the absence or presence of tweezers and stained with crystal violet (blue) indicating biofilm formation. The white areas account for well areas not coated with biofilm. (D) Optical density corresponding to biofilm stained with crystal violet. Tweezer concentrations were: 12.5,25, and 50 μM. Data are presented as the mean ± SD of three replicates. *p < 0.05 compared with control (black bar).

Figure 3.. Molecular tweezers inhibit PSMα1 fibrillation and disaggregate preformed fibrils

(A) PSMα1 β-sheet formation kinetics in the absence or presence of molecular tweezers at 1:1 and 1:5 molar ratio monitored by thioflavin T fluorescence. Data are presented as the mean ± SD of three replicates. (B) Transmission electron microscopy (TEM) micrographs of PSMα1 at the end of the aggregation reactions in the absence or presence of equimolar molecular tweezer concentrations results in short scarce fibrils in the presence of CLR01 or multiple fibril bundles in the case of CLR03 and CLR05. (C) TEM micrographs testing potential disaggregation of preformed PSMα1 fibrils by CLR01, CLR03, or CLR05. The molecular tweezers were added at a 1:1 concentration ratio and further incubated for 24 h at 37°C. Scale bars, 300 nm.

Figure 4.. Fluorometric titration of PSMα1 into solutions of CLR01 or CLR05

Fluder6orescence dependence of the emission band at λ_em = 336 nm of CLR01 (A) and λ_em = 351 nm of CLR05 (B) on the PSMα1 concentration in 10 mM sodium phosphate (pH 7.4) (λ_ex = 285 nm). The data were used to calculate the dissociation constant for each molecular tweezer. K_D values were calculated from the fluorometric titration experiments using a standard nonlinear regression by fitting host and guest concentrations.

Figure 5.. Overlay of ¹H-¹³C HSQC spectra of apo PSMαl (red) and mixtures of PSMαl (blue) with each molecular tweezer at a 1:1 concentration ratio for selected residues

(A) CLR01; (B) CLR03; (C) CLR05. CLR01 shows strong interaction with the peptide, whereas CLR05 exhibits a much weaker binding and CLR03 shows no interaction.

Figure 6.. CLR01 and CLR05 form inclusion complexes with the PSM_αl peptide while CLR05 also forms non-inclusion structures

(A) Inclusion complex of CLR01 with K9 (population of the main cluster of structures: 50.1%). (B) Inclusion complex of CLR01 with K12 (population of the main cluster of structures: 51.7%). (C) Inclusion complex of CLR05 with K9 (population of the main cluster of structures: 52.1%). (D) Inclusion complex of CLR05 with K12 (population of the main cluster of structures: 32%). (E and F) Non-inclusion interaction mode of CLR05 with K12 (populations of the main clusters of structures 22.4% and 5.3%, respectively). (G) Binding free energies of inclusion complexes formed byCLR01 and CLR05 with PSMα1 in a fibril-like conformation.

Figure 7.. Complexes between the tweezers and the PSMα1 fibril model

(A) The initial configuration of the system, in which the tweezers do not encapsulate any lysine residue of the fibril model (peptide cluster), is shown with CLR01 as representative case. (B) CLR05-fibril model interactions. CLR05 molecules are indicated in violet. In all panels, the fibril model is shown as transparent surface and the lysine residues interacting with the tweezers are highlighted in pink. The type of complex formed is described as “in” for inclusion complexes and “out” for non-inclusion complexes. Ions and water molecules are omitted for clarity. (C) CLR01-fibril model interactions. Two views of the same structure are shown, in which CLR01 is depicted in yellow when interacting with the studied lysine residues of the fibril and in gray when it remains in the solvent, close to the fibril but not interacting with the target residues.