Small molecule targeting amyloid fibrils inhibits Streptococcus mutans biofilm formation

Amyloid fibrils are important scaffold in bacterial biofilms. Streptococcus mutans is an established cariogenic bacteria dwelling within biofilms, and C123 segment of P1 protein is known to form amyloid fibrils in S. mutans biofilms, among which C3 segment could serve as a promising anti-amyloid target due to its critical role in C123-P1 interactions. Recently, small molecules have been found to successfully inhibit biofilms by targeting amyloid fibrils. Thus, our study aimed to screen small molecules targeting C3 segment with the capacity to influence amyloid fibrils and S. mutans biofilms. In silico screening was utilized to discover promising small molecules, which were evaluated for their effects on bacterial cells and amyloid fibrils. We selected 99 small molecules and enrolled 55 small molecules named D1–D55 for crystal violet staining. Notably, D25 selectively inhibit S. mutans biofilms but had no significant influence on biofilms formed by Streptococcus gordonii and Streptococcus sanguinis, and D25 showed no bactericidal effects and low cytotoxicity. In addition, amyloid fibrils in free-floating bacteria, biofilms and purified C123 were quantified with ThT assays, and the differences were not statistically significant in the presence or absence of D25. Morphological changes of amyloid fibrils were visualized with TEM images, where amorphous aggregates were obvious coupled with long and atypical amyloid fibrils. Moreover, amyloid-related genes were upregulated in response to D25. In conclusion, D25 is a promising antimicrobial agent with the capacity to influence amyloid fibrils and inhibit S. mutans biofilms. Supplementary Information The online version contains supplementary material available at 10.1186/s13568-021-01333-2.


Virtual screening
We used MOE v2015.1001 for structure-based virtual screening. The X-ray structure of C123 segment was obtained from RCSB Protein Data Bank (PDB ID: 3QE5), where C3 was extracted. Specs library were used as a screening library. Firstly, all compounds were dealt with the Wash module in MOE, then all compounds were ranked according to London dG scoring with high throughput rigid docking, and top 10 K compounds were enrolled in the flexible docking with the aid of "induced fit" method.
Next, the force field in AMBER12: EHT as well as implicit solvation model in Reaction Field (R-field) were utilized before docking. We optimized the protonation and hydrogen state of the C3 segment with LigX module and set the situation at the pH = 7 and temperature = 300 K. At last, flexible docking were conducted where the docked poses were firstly ranked based on London dG scoring, and a force field was refined on the top 10 poses with a rescoring of GBVI/WSA dG. Finally the top 99 hits of Specs were identified.

qRT-PCR For gene expression
qRT-PCR was utilized to quantify the gene expression level according to manufactures' instruction.
Reaction mixture at the volume of 20 μL were depicted below: 10 μL SYBR green PCR master mix (Hieff® qPCR SYBR® Green Master Mix, Yeasen Biotech CO., Ltd, China), 0.8 μL for forward and reverse primers, 2 μL template DNA and 6.4 μL DEPC water. Real-time PCR were designated as follows: 95 ℃ for 300s, 40 repeats of 95 ℃ for 10 s, 60 ℃ for 20 s and 72 ℃ for 20 s. Sequences of primes were presented in Table S1.

Biofilm removal assays
The effects of D25 on S. mutans biofilms at distinct conditions were assessed. In brief, S. mutans were incubated anaerobically to form biofilms. After 4, 8, 16 and 24 h, the medium were aspirated and the biofilms were treated with D25 at the concentration of 6.25 μg/mL and incubated for another 24 h. Thereafter, the biofilms were fixed and measured biomass using crystal violet staining as described above.
Meanwhile, mature biofilms formed after 24 h were treated with D25 at the concentration of 1.56 to 25 μg/mL as detailed in the text. The biomass was also evaluated by crystal violet staining as mentioned before. Statistical analysis of all data were performed with GraphPad Prism 7 (GraphPad Software, San Diego, USA) using Student's two-sample t-test. Significant differences were set at the 95% confident level.

Figure S1
The flowchart of structure-based virtual screening.

Figure S2
Effects of compound D25 on the expression of vicR, brpA, comDE, atpD and relA genes. The results from qRT-PCR were obtained with the 2 -ΔΔCT method. All results were shown as mean ± standard deviations from three independent experiments. *, differences were statistically significant when compared with control group (P < 0.05).

Figure S3
Effects of D25 on biofilms at different concentrations and time intervals. a S. mutans biofilms formed after 24 h were treated with different concentrations of D25. NS, differences were not statistically significant when compared with control group. b Effects of D25 on S. mutans biofilms formed after 4, 8, 16 and 24 h. *, differences were statistically significant when compared with control group (P < 0.05). NS, differences were not statistically significant when compared with control group.

Table S1
Primers used in this study for qRT-PCR to quantify gene expression level.

Primes
Nucleotide Sequence (  c: Small molecules were presented as specific codes in this article.