A New PqsR Inverse Agonist Potentiates Tobramycin Efficacy to Eradicate Pseudomonas aeruginosa Biofilms

Abstract Pseudomonas aeruginosa (PA) infections can be notoriously difficult to treat and are often accompanied by the development of antimicrobial resistance (AMR). Quorum sensing inhibitors (QSI) acting on PqsR (MvfR) – a crucial transcriptional regulator serving major functions in PA virulence – can enhance antibiotic efficacy and eventually prevent the AMR. An integrated drug discovery campaign including design, medicinal chemistry‐driven hit‐to‐lead optimization and in‐depth biological profiling of a new QSI generation is reported. The QSI possess excellent activity in inhibiting pyocyanin production and PqsR reporter‐gene with IC50 values as low as 200 and 11 × 10−9 m, respectively. Drug metabolism and pharmacokinetics (DMPK) as well as safety pharmacology studies especially highlight the promising translational properties of the lead QSI for pulmonary applications. Moreover, target engagement of the lead QSI is shown in a PA mucoid lung infection mouse model. Beyond that, a significant synergistic effect of a QSI‐tobramycin (Tob) combination against PA biofilms using a tailor‐made squalene‐derived nanoparticle (NP) formulation, which enhance the minimum biofilm eradicating concentration (MBEC) of Tob more than 32‐fold is demonstrated. The novel lead QSI and the accompanying NP formulation highlight the potential of adjunctive pathoblocker‐mediated therapy against PA infections opening up avenues for preclinical development.

Synthesis of IIa, IIb described in Scheme S1 (see below).

S3
To a solution of the corresponding 2-trifluoromethyl-4-aminopyridine (20.0 mmol) in DCM (60 mL) were added Boc 2 O (9.82 g, 45.0 mmol), DIPEA (7.66 mL, 44.0 mmol) and DMAP (806 mg, 6.60 mmol). The reaction mixture was stirred for 18 h at r.t., and saturated NH 4 Cl solution was added followed by extraction with EtOAc. The combined organic layers were washed with Brine, dried over Na 2 SO 4 , filtrated and evaporated under reduced pressure. The crude product was used without any further purification.
The corresponding di-tert-butyl(2-(trifluoromethyl)pyridin-4-yl)dicarbamate (6.00 mmol) was dissolved in DCM/TFA 9:1 (12 mL) and stirred for 16 h at r.t. until the starting material was fully consumed (TLC control). Saturated Na 2 CO 3 solution was added followed by extraction with DCM. The combined organic layers were dried over Na 2 SO 4 filtrated and concentrated in vacuo. The desired compound was used without further purification.
Addition of saturated NH 4 Cl solution and extraction with Et 2 O followed by washing the combined organic layers with saturated LiCl solution yielded the crude product after drying over Na 2 SO 4 , filtration and concentration in vacuo, which was purified by automated column chromatography using a gradient of PE/EtOAc.

General procedure 4: Boc-Deprotection
Synthesis of Va, Vb described in Scheme S1 (see below).
The corresponding tert-butyl prop-2-yn-1-yl(2-(trifluoromethyl)pyridin-4-yl)carbamate (4.03 mmol) was dissolved in DCM (1 mL) and TFA (1 mL) was added. After stirring at r.t. for 18 h, the mixture was cooled to 0 °C, and saturated Na 2 CO 3 solution was added. The reaction mixture was extracted with DCM, and the combined organic layers were washed with Brine, dried over Na 2 SO 4 , filtered and concentrated under reduced pressure. Automated flash chromatography using a gradient of PE/EtOAc afforded the title compound.

Nanoparticle Preparation and Characterization
Squalenyl hydrogen sulfate was synthesized and characterized before use as previously described. [2] The drug-free and drug-loaded squalenyl hydrogen sulfate nanoparticles (SqNPs) were prepared using a single-step nanoprecipitation method as described in our previous study. [2] Briefly, the drug-free SqNPs were prepared as follows: a solution of squalenyl hydrogen sulfate in tetrahydrofuran (THF) was prepared at a concentration of 10 mg/mL from which 0.1 mL was dropped slowly into 1 mL of MilliQ water under constant stirring allowing nanoprecipitation. Following the 30 min stabilization, organic solvent was evaporated, and a dispersion of SqNPs 1 mg/mL in aqueous medium was obtained. The concentration of SqNPs in aqueous solution can be varied by changing the initial concentration of squalenyl hydrogen sulfate in THF.
The preparation of drug-loaded SqNPs is as below.
Single drug-loaded SqNPs preparation: the hydrophobic compound (4) loaded SqNPs were prepared using nanoprecipitation method. The optimal compound (4):squalenyl hydrogen sulfate ratio was 0.8:9 (w/w) was solubilized in THF allowing the maximum encapsulation S14 efficiency (EE%) and loading capacity (LC%) of compound (4) in SqNPs. The positively charged tobramycin (Tob) was prepared in the aqueous solution, which was loaded on SqNPs surface via electrostatic interaction during the precipitating procedure. The optimal Tob/ squalenyl hydrogen sulfate ratio was 3:7 (w/w), providing the maximum EE% and LC% of Tob as well as allowing the long-term drug-loaded SqNPs stability.
The Tob and compound (4) co-loaded SqNPs were produced by simultaneously coassembling squalenyl hydrogen sulfate with Tob and QSI 4, using the above protocol. The Tob concentration (from 6.25 -200 µg/mL) loaded in SqNPs was tuned by varying the Tob initial concentration in the aqueous phase.

Reporter Gene Assay
The inverse agonistic/antagonistic activity of potential PqsR ligands was evaluated in a βgalactosidase reporter gene assay in E. coli as previously described. [3] In short, test compounds were co-incubated with E. coli DH5α cells containing the plasmid pEAL08-2, [4] which expresses PqsR and the β-galactosidase LacZ under the control of the pqsA promoter.
The inverse agonistic/antagonistic activity of compounds was measured in the presence of 50 nM PQS. After incubation, galactosidase activity was quantified photometrically and expressed as ratio of controls. The given IC50 values represent the mean of at least two independent experiments using at least eight different concentrations of test compound with n = 4. Non-linear regression analysis was performed using OriginPro 2020.

Pyocyanin assay
The pyocyanin inhibitory effect of the synthesized compounds and the compound (4) loaded SqNPs on PA14 wt were investigated by pyocyanin assay as described in previous work. [2,5] Briefly, PA14 wt was grown in PPGAS medium, overnight, at 37 °C, shaking at 200 rpm. The OriginPro 2020. Figure S2 shows the dose-response pyocyanin inhibitory assay of compound (4) loaded SqNPs on PA14 wt. The pyocyanin inhibitory efficacy was improved by a factor of two when using compound (4) loaded SqNPs (IC 50 is 100 ± 0.018 nM) compared to free compound (4) prepared in DMSO (IC 50 is 199 nM). It is noted that as reported in our prior study, the drug-free SqNPs did not show any effects on PA14 pyocyanin production. [2] Figure S2. The production of pyocyanin levels compared to control PA14 wide type (wt) of the samples treated with 200 μg/mL drug free SqNPs. Three independent experiments were conducted in triplicate.

Alkylquinolone Inhibition Assay
The inhibitory effect of the synthesized compounds on alkylquinolone production in PA was measured as previously described with the modification that wt PA14 was used instead of PA14pqsH. [ Table S1. Data acquisition and quantification was performed using Xcalibur software with the use of a calibration curve relative to the area of the IS. IC 50 values were calculated according to the procedure of the pyocyanin assay (see above).   Table S2.

Minimum Biofilm Eradicating Concentration (MBEC) assay
Chemical and biological materials for in vitro assays. All chemicals were purchased from Sigma-Aldrich unless otherwise specified. Yeast extract was obtained from Fluka. Bacto™  Table S2.
The MBECs of free Tob, Tob-loaded SqNPs, free Tob plus free compound 4 prepared in DMSO, and Tob and compound (4) co-loaded SqNPs were determined by the MBEC assay as reported in previous study. [2,7] Briefly, PA14 wt was grown in PPGAS medium, overnight at  Figure S3 shows that the drug-free SqNPs and 20 μM of compound 4 either in free form or loaded in SqNPs did not show any PA14 wt biofilm inhibitory effects. Figure S4 shows the results of Figure

CEREP off-targets assay
Experiments towards possible off-target effects were conducted following Eurofins Discovery protocol. Detailed information about the assays can be found in: https://www.eurofinsdiscoveryservices.com/catalogmanagement/viewitem/SafetyScreen44-

Cytotoxicity Assays
Hep G2 or HEK293 cells (2 × 10 5 cells per well) were seeded in 24-well, flat-bottomed plates. Culturing of cells, incubations, and OD measurements were performed as described previously with slight modifications. [8] Twenty-four hours after seeding the cells, the incubation was started by the addition of test compound in a final DMSO concentration of 1%. The living cell mass was determined after 48 h. At least two independent measurements were performed for each compound.

Kinetic Turbidimetric Solubility Assay
The desired compounds were sequentially diluted in DMSO in a 96-well plate.  Table S4.   masses for caffeine, 4 and Tob can be found in Table S6; peak areas of each sample and of the corresponding internal standard were analyzed using MultiQuant 3.0 software (AB Sciex).
Peak areas of the respective sample of 4 and Tob were normalized to the internal standard peak area. Peaks of PK samples were quantified using the calibration curve. The accuracy of the calibration curve was determined using QCs independently prepared on different days (Table S5). PK parameters were determined using a non-compartmental analysis with PKSolver. [9] ELF concentrations were calculated using equations (S3) and (S4). [10]  Background studies were performed in order to generate typical mean and standard deviation values for biomarkers of the control group. Sample size was calculated using BiAS.
(Software Version 11.08) from these values with alpha = 0.05 and P = 0.8 for biomarker reduction of 80% resulting in a sample size of ten mice or higher for a two-sample t-test.
**Values in parentheses refer to the highest resolution shell.
a The statistics are for data that were corrected for anisotropy by STARANISO. [13] b The resolution limits for three directions in reciprocal space (a*, b*, c*) are indicated here. To accomplish this, STARANISO 1 computed an ellipsoid postfitted by least squares to the cutoff surface, removing points where the fit was poor. Note that the cutoff surface is unlikely to be perfectly ellipsoidal, so this is only an estimate. Chemical Computing Group), [19] while graphic processing for manuscript figures was done using YASARA structure (YASARA Biosciences GmBH) [20] and POV-Ray 3.7.0.
For Figure 1C, 6Q7W was loaded into MOE, and the amide motif of compound (2) was modelled into the ligand structure using the built-in "Builder" function of MOE followed by a refinement step using the built-in "QuickPrep" function with standard parameters and the Amber10:EHT force field.
For Figure 2A-B, newly solved complex structure (PDB ID: 6YIZ) was loaded into MOE followed by a refinement step using the built-in "QuickPrep" function with standard parameters and the Amber10:EHT force field. 2D interaction profile was generated using the built-in "Ligand Interactions" function of MOE.

Statistical Analysis
Results of biological experiments are reported as means of at least three independent experiments, unless otherwise stated.
For MBEC determination, One-way ANOVA was performed using GraphPad Prism 8.4.2.
For animal studies, unpaired t-test (one-sided, p-value) was performed with Microsoft Excel.
A summary of data processing and statistical analysis is shown in Table S13. Detailed information can be found in the respective experimental section.
The CEREP panel, determination of CYP and hERG inhibition as well as Mouse MTD were performed at CROs. One-way ANOVA, / AUC determination OriginPro 2020