Terrein is an inhibitor of quorum sensing and c-di-GMP in Pseudomonas aeruginosa: a connection between quorum sensing and c-di-GMP

To address the drug-resistance of bacterial pathogens without imposing a selective survival pressure, virulence and biofilms are highly attractive targets. Here, we show that terrein, which was isolated from Aspergillus terreus, reduced virulence factors (elastase, pyocyanin, and rhamnolipid) and biofilm formation via antagonizing quorum sensing (QS) receptors without affecting Pseudomonas aeruginosa cell growth. Additionally, the effects of terrein on the production of QS signaling molecules and expression of QS-related genes were verified. Interestingly, terrein also reduced intracellular 3,5-cyclic diguanylic acid (c-di-GMP) levels by decreasing the activity of a diguanylate cyclase (DGC). Importantly, the inhibition of c-di-GMP levels by terrein was reversed by exogenous QS ligands, suggesting a regulation of c-di-GMP levels by QS; this regulation was confirmed using P. aeruginosa QS mutants. This is the first report to demonstrate a connection between QS signaling and c-di-GMP metabolism in P. aeruginosa, and terrein was identified as the first dual inhibitor of QS and c-di-GMP signaling.


NMR, MS, and optical rotation data of the isolated terrein
Acetylation of terrein. Acetic anhydride (1 mL) was added to a solution of terrein (10 mg) in pyridine (1 mL), and the mixture was stirred at 25°C for 3 h, after which time the pyridine and excess acetic anhydride were evaporated under a vacuum. The residue was partitioned with EtOAC, and the resultant EtOAc extract was purified by SiO 2 TLC developed with Hexane-EtOAc (2:1) to give diacetyl terrein (8.5 mg) at an Rf of 0.16. The structure of diacetyl terrein was confirmed by NMR and MS data. Pyocyanin and rhamnolipid assay. For both pyocyanin and rhamnolipid production, after overnight cultures of P. aeruginosa PAO1 were diluted 100-fold in LB medium, 5 mL of culture was dispensed into 50 mL conical tubes, treated with test compounds dissolved in DMSO, and incubated at 220 rpm at 37°C for 24 h. The cultures were then centrifuged at 12,000 rpm at 4°C for 10 min. For pyocyanin assay, 5 mL of supernatant was mixed with 3 mL of chloroform and 2 mL of 0.2 N HCl was added to the chloroform fraction. The resultant aqueous fraction was measured at 520 nm by a microplate reader. For rhamonolipid assay, 500 L supernatant was mixed with 500 L diethyl ether. The ether fraction was evaporated to dryness and dissolved in 500 L deionized water. 100 L of the water extract was mixed with 900 L of Orcinol solution (0.19% Orcinol (Sigma) in 53% H2SO4). The mixture was boiled for 30 min, cooled at room temperature for 15 min, and then measured at 421 nm by a microplate reader.
Biofilm assay. Overnight cultures of P. aeruginosa PAO1 were diluted 100-fold in M63 medium and the dilutions were dispensed at 0.1 mL/well in a 96-well polystyrene microplate.
Test compounds or DMSO as a negative control were added to the wells. After incubation at 37°C for 9 h without agitation, the unattached cells and media were removed and the cells forming the biofilm, which remained attached to the well's surface, were stained using 120 L of 0.1% crystal violet for 10 min. The bound crystal violet was solubilized with 150 L of 30% acetic acid in water for 15 min. The OD of the eluted crystal violet was measured at 550 nm by a microplate reader.

Quantification of QS signaling molecules by LC-MS/MS
The samples were subjected to an HPLC system (Luna C18(2), 100 × 2.0 mm, 3 μm, Phenomenex, Torrance, CA, USA) connected to a QTrap 3200 with a Turbolon Spray source (AB SCIEX, Singapore). The column was maintained at 20°C with a flow rate of 0.4 mL/min and a gradient of acetonitrile in 0.1% (v/v) aqueous formic acid; 0-10 min from 70% to 100%.
MRM was performed by selecting the two mass ions set specifically for the selected analytes to detect the transition from parent ion to product ion, i.e., m/z 298. 211 > 197.200 for OdDHL, m/z 172.181 > 71.000 for BHL, and m/z 260.244 > 188.100 for PQS (Sigma). For analysis of OdDHL, BHL, and PQS, the Turbolon Spray source-dependent parameters were optimized to the following values: 10 psi curtain gas, high collision gas, 5500 V ion spray voltage, 400°C temperature, and 12 psi ion source gas. The compound-dependent parameters for OdDHL, BHL, and PQS were optimized to the following values: 19, 17, and 41 eV collision energy; 5, 8, and 12 V entrance potential; 276, 31, and 71 V declustering potential; and 18, 10, and 16 V collision cell exit potential, respectively.

Quantification of c-di-GMP by LC-MS/MS
The samples were subjected to an HPLC system (Luna C18(2), 100 × 2.0 mm, 3 μm, Phenomenex, Torrance, CA, USA) connected to a QTrap 3200 with a Turbolon Spray source (AB SCIEX, Singapore). The column was maintained at 20°C with a flow rate of 0.4 mL/min and a gradient of acetonitrile in 0.1% (v/v) aqueous formic acid; 0-5.6 min from 2% to 30% and 5.6-7 min from 30% to 80%. MRM was performed by selecting the following mass ions: m/z 691.058 > 152.200 for c-di-GMP. For analysis of c-di-GMP, the optimized Turbolon Spray source-dependent parameters were as follows: 10 psi curtain gas, medium collision gas, 5500 V ion spray voltage, 400°C temperature, and 12 psi ion source gas. The compound-dependent parameters were optimized to the following values: 49 eV collision energy, 8.5 V entrance potential, 81 V declustering potential, and 4 V collision cell exit potential.

RT-qPCR of QS-regulated genes
cDNA was synthesized from 2 g of RNA mixed with 1 μg of random primers (Promega C1181) and RNase-free water (Sigma W4502) in a total of 13.37 μL and incubated for 5 min at 70°C.  Table S1. Amplification and expression were carried out in a total volume of 20 L containing 10 L SYBR Premix Ex Taq TM (Takara, Shiga, Japan), 1 L each of the forward and reverse primers (5 M) of target genes, 2 L template cDNA, and 6 L RNase-free water. The cycling parameters were as follows: initial activation at 95℃ for 30 s; 40 cycles at 95℃ for 5 s, 60℃ for 30 s, and melting curve analysis at 95℃ for 15 s, 60℃ for 5 s and 95℃ for 5 s. mRNA expression was normalized using the endogenous rpoD gene.  Supplementary Figure S2. Effects of terrein on cell viability, growth-dependent elastase production, and elastase activity. (a) Dose-dependent effect of terrein on bacterial growth. The cell density was measured at 600 nm and the number of viable cells was counted in each culture after 18 h of growth. Three independent experiments were performed, and the mean  SD values are displayed in each bar. (b) Growth-dependent effect of terrein on elastase production. PAO1 cells were grown in LB medium containing two different concentrations of terrein for different times, and culture supernatants were harvested at the given time points to measure elastase activity. Each experiment was repeated in triplicate, and the mean  SD values are displayed. *, P < 0.05; **, P < 0.01 versus elastase activity from untreated PAO1 cells. (c) Growth curves of PAO1 cells grown aerobically in shaking LB medium and LB medium with terrein. Each growth curve experiment was repeated in triplicate, and the mean  SD values are displayed. (d) Dose-dependent effect of terrein on elastase activity. The PAO1 cells was grown in LB medium containing various concentrations of terrein for 18 h, and culture supernatants were harvested to measure elastase activity. Aliquots of PAO1 culture supernatants derived from 18 h of culture in LB medium were treated with various concentrations of terrein for 7 h and assayed for elastase activity. Two independent experiments in triplicate were carried out, and the mean  SD values are displayed in each bar. *, P < 0.001; **, P < 0.0001 versus elastase activity from untreated PAO1 cells.
Supplementary Figure S3. Chemical structure of diacetyl terrein (a) and comparison of terrein and diacetyl terrein on elastase activity (b). The PAO1 cells was grown in LB medium containing various concentrations of terrein or diacetyl terrein for 18 h, and then elastase activity was assayed from the culture supernatants. . Two independent experiments in triplicate were carried out, and the mean  SD values are displayed in each bar. *, P < 0.001; **, P < 0.0001 versus elastase activity from untreated PAO1 cells.
Supplementary Figure  h. After the biofilms were dissociated from the wells by gentle sonication, cellular c-di-GMP was extracted from the biofilm cells, measured, and normalized by total proteins. The experiment shown is representative of three independent experiments in triplicate, and the