THP-1 (human monocytes, ATCC TIB-202) and THP-1 AhR reporter (Moura-Alves et al., 2019) cells were grown in RPMI 1,640 (GIBCO), supplemented with 10% (v/v) heat-inactivated fetal calf serum (FCS; GIBCO), and 1% (v/v) penicillin-streptomycin (GIBCO). All AhR reporter cell lines were maintained with an additional 5 mg/mL of Puromycin (Calbiochem). Cells were kept at 37°C in 5% CO₂. Lentiviral infection was performed as described previously (Moura-Alves et al., 2019) and according to the protocols available on the RNAi Consortium website (https://portals.broadinstitute.org/gpp/public/). A similar protocol was used to generate the THP-1 Control and THP-1 AhR-Knockdown (KD) cell lines.

To quantify the expression of genes of interest, the THP-1 cells were grown in 24-well plates for 24 h with or without PQS/Pht/vitamin K addition. Total RNA was extracted using the TRIzol LS reagent (Invitrogen, California, USA). The qRT-PCR assay was performed using the ABI 7500 sequence detection system (Applied Biosystems, California, USA). Three replicates were performed with the glyceraldehyde-3-phosphate dehydrogenase gene (GAPDH) as the internal control gene.

The AhR and PqsR PDB structure files were downloaded from the Uniprot database and modified in PyMOL v. 1.4 (Schrodinger, LLC) (Seeliger and De Groot, 2010). The PQS, Pht, and vitamin K molecule files were downloaded from the Pubchem database. The docking process between proteins and potential ligands was performed in the Autodock Vina plug-in inside Chimera (Pettersen et al., 2004) (UCSF). The software program LIGPLOT v.4.5.3 (Wallace et al., 1995) was used to map the interactions between proteins and potential ligands. Open babel software (O'boyle et al., 2011) was used to transfer proper file formats.

The human recombinant AhR protein with Val128-Asn399 and N-terminal His-Tag was purchased from ImmunoClone (New York, USA). The binding of ligands to the AhR protein was assessed by microscale thermophoresis (MST) experiments using the Monolith^® NT.LabelFree (NanoTemper Technologies GmbH). MST measurements were performed according to the manufacturer's instructions previously described (Moura-Alves et al., 2019). The interaction affinity of the dissociation constant Kd for each ligand was analyzed using the NanoTemper Affinity software by the changes in the normalized fluorescence (fraction bound) vs. the ligand concentration.

Bacterial strains and plasmids used in this study are described in Supplementary Table 1. Bacteria were cultured in Luria–Bertani (LB) broth (1% tryptone, 0.5% yeast extract, 0.5% NaCl) and ABTGC (B-medium (0.1% MgCl₂, 0.1% CaCl₂, 0.1% FeCl₃) supplemented with 10% A10, 0.2% glucose, and 0.2% casamino acids). The media were solidified with 1.5% Bacto Agar (Difco). Kings' medium [Milli-Q water supplemented with proteose peptone (20 g/L), potassium sulfate (10 g/L), magnesium chloride, anhydrous (1.640 g/L), and glycerol (10% v/v)] were used for the pyocyanin quantification assay.

Pseudomonas aeruginosa mutant strains ΔpqsC and ΔpqsR were generated by homologous recombination using the previously described protocol (Choi and Schweizer, 2005). The mutants were constructed by overlapping PCR to contain a gentamicin-resistance cassette. The mutant fragments were inserted into pK18, a suicide vector, to produce gene knockout plasmids. Each knockout plasmid was transformed into E. coli strain RK600 and conjugally transferred from RK600 to PAO1. The resultant integrants were selected on agar containing 30 μg/mL gentamicin (Gm30). To resolve merodiploids, we streaked the Gm-resistant colonies for single colonies on LB+Gm30 plates containing 5% sucrose. Potential mutants were screened by PCR using corresponding flanking primers and were confirmed by Sanger sequencing.

PQS and Pht compounds were prepared in a 96-well microtiter plate (Nunc, Denmark) at a concentration of 100 mM in DMSO. PQS and Pht were then mixed with ABTGC medium and serial dilutions to give a concentration of 125, 250, 500, 750, 1,000, 1,500, and 2,000 nM, respectively. An overnight culture of the PAO1-pqsA-gfp strain (grown in LB medium at 37°C, 200 rpm) was diluted in ABTGC medium to a final optical density at 600 nm (OD_600nm) of 0.02 (~2.5 × 10⁸ CFU/mL). A DMSO control (0.1% final concentration) and blank control (ABTGC medium) were used. The microtiter plate was incubated at 37°C in a Tecan Infinite 200 Pro plate reader (Tecan Group Ltd., Mannedorf, Switzerland) to measure the cell density (OD_600nm) and GFP fluorescence (excitation at 485 nm, emission at 535 nm) at 30 min intervals for at least 12 h. The inhibition assay for PQS and Pht was done in triplicate. Growth curve assays were performed only by measuring the cell density (OD_600nm) using the same method above.

Overnight cultures of P. aeruginosa WT and ΔlasIΔrhlI mutant strains were standardized to an OD_600nm of 0.1 and diluted 100 times into 25 mL of fresh King's Medium in a 250 mL flask. The PQS and Pht compounds were added to give a final concentration of 750 nM. The cultures were grown for 48 h at 37°C in shaking conditions (200 rpm). The cultures were monitored every 24 h to observe any color change from light-yellow to greenish-yellow in the untreated flask. The mutant strain ΔlasIΔrhlI was used as the negative control. The final cell density was measured at 600 nm (OD_600nm) using a Tecan Infinate 200 Pro plate reader (Tecan Group Ltd., Mannedorf, Switzerland). The cultures were then centrifuged for 10 min at 10,000 rpm, and 7.5 mL of the supernatants were transferred into new Falcon tubes. Pyocyanin extraction was conducted using chloroform (3 mL) and 0.2 M HCl (1.5 mL). The top aqueous layer of HCl containing pyocyanin was pipetted into a microtiter plate, and absorbance was measured at 520 nm. The data was normalized by dividing the OD_520nm reading by the final OD_600nm values.

Biofilm formation was quantified by crystal violet staining. Overnight LB-grown cultures were diluted in fresh medium (1:100) and incubated in 2 mL of LB in 15 mL tubes at 37°C for 24 h, statically allowing biofilm formation. After removing the spent media, we removed the loosely associated bacteria by washing them with water two times, and the remaining bacteria were stained with 0.1% crystal violet for 20 min. Then, the crystal violet stain was discarded while the stained biofilms were washed twice thoroughly with ddH₂O and air-dried. The stained biofilm was then dissolved into 200 μL of 95% ethanol and quantified at 630 nm using a Tecan Infinate 200 Pro plate reader (Tecan Group Ltd., Mannedorf, Switzerland).

EMSA assays were performed using cell lysates of PAO1 and pqsR mutants as previously described (Gruber et al., 2016). For each sample, cell-lysate was incubated with biotinylated 248-bp pqsA promoter DNA (pqsA′), which was generated by PCR using biotinylated forward primer (TTCTTGCTTGGTTGCCG) and reverse primer (GACAGAACGTTCCCTCTT). The reaction mixture contained 10 mM Tris–HCl (pH 8.0), 1 mM EDTA, 50 mM NaCl, 1 mM DTT, 1 μg/μL Poly (dI-dC) and was incubated at room temperature (24°C) for 20 min. Samples were separated on a 5% polyacrylamide gel in 0.5 × Tris Borate EDTA (TBE) run at 100 V for 120 min and transferred onto a positively-charged nylon membrane at 40 V for 1 h. Biotinylated DNA on the nylon membrane was probed by streptavidin-HRP conjugate, detected by a chemiluminescent substrate (Beyotime EMSA kit; Shanghai, China), and visualized by exposing it to a GE ImageQuant RT-ECL instrument.

Overnight PAO1 cultures were diluted 100 times into 25 mL of fresh LB medium. A Pht compound was added to one sample to give a final concentration of 750 nM. Then, total proteins were extracted from bacterial samples and trypsin digested. Then, the digested sample was labeled with a TMT reagent. Next, TMT labeled peptide was separated by a RIGOL L-3000 dual gradient High-Performance Liquid Chromatography (HPLC) using an Agela Durashell-C18 column. An HPLC-separated peptide was identified by nano UPLC-MS/MS consisting of a Nanoflow HPLC system (EASY-nLC 1000 system from Thermo Scientific) with an EASY-Spray C18 column and an Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific). Then, the protein was identified using the Uniprot_HUMAN database. The resulting MS/MS data were processed using Proteome Discoverer 2.4. The final results were functionally annotated by Gene Ontology (GO) (Zhang et al., 2019), KEGG (https://www.genome.jp/kegg/) (Jones et al., 2014), and STRING (https://www.string-db.org) (Franceschini et al., 2013).

All statistical analyses were performed using GraphPad Prism software with a two-sided Mann-Whitney U-test and Benjamini-Hochberg-corrected P-values. Data represent the mean ± standard deviation (SD) of three independent experiments unless otherwise indicated. P-values < 0.05 were considered statistically significant. All false discovery rate controls were performed with the Benjamini-Hochberg procedure, and a false-discovery rate of 10% (q < 0.1) was selected as the significant threshold. Statistical details for all tests performed can be found in the figure legends.

Pseudomonas aeruginosa QS signal molecule PQS has been reported to downregulate AhR-induced cytochrome P4501 (CYP1) enzyme and CYP1A1 enzymatic activities in mouse liver cells (Hepa-1c1c7) (Moura-Alves et al., 2019). We demonstrated by qRT-PCR that the transcription of two AhR downstream genes, CYP1B1 and AhR repressor (AHRR), was repressed 2.3- and 2.5-fold with the addition of 10 μM PQS in THP-1 human monocytes compared to control (additional of DMSO alone) (Figures 1A,B). In addition, there were no significant expression changes of these genes in the THP-1 KD cells upon PQS addition.

The MTb natural metabolite Pht was found to activate CYP1B1 and AHRR transcription in THP-1 cells in an AhR-dependent manner (Moura-Alves et al., 2014). Our qRT-PCR results also confirmed that CYP1B1 and AHRR transcription were promoted by 2.2- and 1.9-fold with the addition of 10 μM Pht in THP-1 human monocytes compared to control (DMSO only), respectively (Figures 1C,D). These results demonstrated that PQS and Pht adversely affected AhR regulatory pathways.

The chemical structures of the P. aeruginosa auto-inducer PQS and MTb Pht are shown in Figures 2A,D. Molecular docking was performed using the ligands PQS and Pht against the ligand-binding domain of the human AhR protein. These compounds and their structures are shown in Figures 2B,E. To map the interactions between PQS and the residues within the ligand-binding site on AhR, we used the software LIGPLOT v. 4.5.3 (Wallace et al., 1995). This program provides a 2-dimensional map showing the hydrogen-bond and hydrophobic interactions between atoms in the ligand and those of the binding partner.

The results showed that PQS binds to the AhR protein through hydrophobic forces, and they appear to interact at residues Gln 277, Tyr 221, Pro 200, Ile 281, Ser 220, Phe 228, Phe 201, Ala 224, Lys 225, and Tyr 264 (Figure 2C). However, the docking analysis suggested that Pht and AhR appeared to interact at residues Tyr 84, Ala 46, Ile 101, Pro 22, Phe 23, Tyr 43, Ser 42, and Leu 39 (Figure 2F).

Next, we used an MST assay to quantify the AhR binding affinity with different ligands, including PQS and Pht. The MST results indicated that the dissociation constant (Kd) values of AhR protein with PQS and Pht were 3.75 μM and 474.41 nM, respectively (Figure 2G). These results confirmed that PQS and Pht directly bind to AhR protein as ligands.

To engineer a PQS auto-inducer reporter system, we first made a PQS deficient ΔpqsC mutant by deleting the pqsC gene. The pqsA-gfp reporter plasmid (Yang et al., 2007) was next transferred into the ΔpqsC mutant to make the PQS biosensor strain ΔpqsC P_pqsA−gfp. In addition, the pqsA-gfp reporter plasmid was transferred into the PAO1 WT to make the PQS bioreporter strain PAO1 P_pqsA−gfp. Induction of the pqsA gene is mainly controlled by PqsR (the signal receptor of the pqs QS system), which is activated by the presence of PQS (Soheili et al., 2019). Thus, in the PAO1 P_pqsA−gfp strain, a decrease in GFP fluorescence would indicate the presence of an antagonist of PQS, while exogenous PQS was needed to induce the GFP fluorescence of the ΔpqsC P_pqsA−gfp strain.

Pht was exogenously added to both the above reporter strains to evaluate its impact on the pqs QS system. As shown in Figure 3A, both PQS and Pht could not affect the growth of the P. aeruginosa strain. PQS—not Pht—induced the pqsA-gfp expression in a dose-dependent manner in the PAO1 ΔpqsC strain (Figure 3B). In contrast, Pht was able to inhibit pqsA-gfp expression in a dose-dependent manner in the PAO1 P_pqsA−gfp strain, with the optimal effect at a concentration of 750 nM (Figure 3C). These results showed that PQS activated while Pht inhibited the P. aeruginosa pqs QS system.

The production of pyocyanin and the formation of biofilms are pqs QS system-regulated phenotypes related to virulence in P. aeruginosa (Pearson et al., 1997). Pyocyanin is a key virulence factor produced by P. aeruginosa that plays a major role in cystic fibrosis (CF) lung infections (Lau et al., 2004a). Pyocyanin was shown to cause tissue damage and necrosis in a P. aeruginosa murine model for a lung infection (Lau et al., 2004b). Biofilm formation by P. aeruginosa is well known as the cause of chronic infections (Costerton et al., 2005). We thus evaluated the effects of Pht on these two phenotypes.

As expected, the pyocyanin production by P. aeruginosa PAO1 strain was found to be suppressed by Pht to a level comparable to that of PAO1 ΔlasIΔrhlI strain and the ΔpqsC strain. It could be seen from the intensity of the green pigment in the supernatants of different samples that the PAO1 WT culture exhibited a dark green, ΔlasIΔrhlI and ΔpqsC cultures exhibited light green, and the WT+Pht culture did not have a visible green color. These results indicated that Pht inhibited pyocyanin production (Figure 3D). Crystal violet staining showed that the biofilm formation of P. aeruginosa PAO1 WT was suppressed by Pht to a level comparable to that of the PAO1 ΔlasIΔrhlI and the ΔpqsC strains (Figures 3E,F). Besides, we detected that Pht did not inhibit P. aeruginosa rhamnolipid productions (Supplementary Figure 1). These results indicated that Pht repressed P. aeruginosa pqs QS system, pyocyanin production, and biofilm formation.

The interactions between alkyl quinolones with PqsR are simulated by the crystal structure of PqsR complexed with its ligand, which suggests the influence of ligand binding on PqsR affinity to its DNA-binding site (Ilangovan et al., 2013). Molecular docking was first performed using the reference ligand PQS and potential ligand Pht against the ligand-binding domain of the P. aeruginosa PqsR protein. These compounds and their structures were shown (Figures 4A,C). To map the interactions between PQS and the residues within the ligand-binding site on PqsR, the software program LIGPLOT v. 4.5.3 was used.

The results showed that PQS binds to PqsR protein by hydrophobic forces, and they appeared to interact at residues Leu 189, Tyr 258, Val 170, Ile 236, Ile 149, Ala 168, Thr 265, and Leu 207 (Figure 4B). While the docking analysis suggested that Pht and PqsR appeared to interact at residues Ile 56, Leu 114, r103 (A), n101 (A), Met 131, Ile 143, Phe 128, and Leu 104 (Figure 4D). Except for hydrophobic forces, Pht and PqsR are also predicted to be bound by a hydrogen bond at Leu 115 (A). These molecular docking results revealed that the QS signaling molecule PQS and Pht competed for the same active pocket in the PqsR protein. The hydrogen bond between Pht and PqsR might offer Pht an advantage when competing with PQS. Thus, we suspected that Pht competitively disrupted the PQS-mediated PqsR activation and signaling in P. aeruginosa. These predicted binding modes of PqsR and Pht need further experimental verification.

The electrophoretic mobility shift assay (EMSA) is a rapid and sensitive experimental method to detect protein-nucleic acid interactions (Hellman and Fried, 2007). The protein-nucleic acid complexes migrate more slowly than the corresponding free nucleic acid in the gel. To test whether Pht will interact with PqsR and further affect its DNA binding capacity, we performed EMSA with the biotinylated-pqsA promoter (Bio-pqsA) using cell lysates of PAO1 WT and ΔpqsR strains. The results showed that only PAO1 cell lysates shifted Bio-pqsA migration, while ΔpqsR cell lysates did not show any effect. The addition of Pht attenuated the interaction between PAO1 cell lysates and Bio-pqsA, whereas there was no effect by adding the cell lysates of ΔpqsR (Figures 4E,F). These results indicated that Pht could indeed block the binding of PqsR protein to the pqsA promoter, which further leads to the inhibitory effect against P. aeruginosa pqs QS systems.

To study the changes in the proteomic profile of P. aeruginosa PAO1 upon Pht addition, TMT was used as the labeling strategy for comparative quantitative proteomic analysis (performed with a false discovery rate below 1%). The following cutoffs were used for protein identification: an unused protein score of at least 2 (i.e., 99% confidence of identification) and having more than 1 peptide identified. Using these cutoffs, we identified 2,203 proteins. Using a p-value cutoff of 0.05, 37 proteins were found to be significantly affected by Pht. The expression of 10 proteins was upregulated, whereas the expression of 27 proteins was downregulated (Supplementary Table 2).

Table 1 shows the 27 proteins whose abundances were significantly decreased in the Pht-treated P. aeruginosa PAO1 strain compared to those in the control PAO1 strain without Pht addition. Of these 27 proteins, five have been previously found to be pqs QS regulated: phnA (Anthranilate synthase component 1), pqsB (Anthraniloyl-CoA anthraniloyltransferase), phzD2 (Phenazine biosynthesis protein PhzD2), pqsD (Anthraniloyl-CoA anthraniloyltransferase), and pqsA (Anthranilate–CoA ligase). These results further confirmed that Pht directly attenuates P. aeruginosa pqs QS regulated global proteome network.

By comparing the online PubChem compound library (https://pubchem.ncbi.nlm.nih.gov/), we found several structural analogs of the Pht molecule, including vitamins K₁, K₂, and K₃ (Figures 5A–C). Vitamin K is a fat-soluble vitamin that participates in blood clotting (Palermo et al., 2017). Vitamin K deficiency may lead to prolonged blood clotting time, bleeding, and even death in severe cases (Palermo et al., 2017). Vitamin K not only has protective effects on the heart and the cerebrovascular system but also has a maintenance effect on the normal elasticity of blood vessel walls and participates in the metabolism of bone for the prevention and cure of osteoporosis and fracture (Fusaro et al., 2020). However, no evidence has yet shown that vitamin K can directly affect bacterial pathogenesis.

To reveal whether Pht analogs vitamin K families influence the AhR pathway, we performed qRT-PCR experiments to measure the transcription of AhR downstream genes. The results showed that CYP1B1 and AHRR transcription was increased by 1.9- and 1.6-fold with the addition of 10 μM vitamin K₁ in THP-1 human monocytes compared to that with DMSO (Figures 5D,E). Besides, there is no change in AhR KD cells with vitamin K₁ stimulation. These results indicated that vitamin K₁ promoted AhR downstream genes CYP1B1 and AHRR through AhR in THP-1 human monocytes.

Molecular docking was performed using the ligands vitamins K₁, K₂, and K₃ against the ligand-binding domain of the human AhR protein. Figures 6A,C,E show these compounds and their structures. The LIGPLOT results showed that vitamin K₁ and AhR protein were bound by hydrophobic forces that appeared to interact at residues Tyr 221, Lys 225, Gln 277, Phe 201, Ala 224, Pro 200, Tyr 264, Phe 228, Val 253, Val 255, and Ile 281 (Figure 6B). Similarly, the docking analysis suggested that vitamins K₂ and K₃ with AhR appeared to interact at several vital residues (Figures 6D,F).

We further used the MST assay to detect AhR protein binding with vitamins K₁, K₂, and K₃. The MST results indicated that the dissociation constant (Kd) values of AhR protein with vitamins K₁, K₂, and K₃ were 1.99 μM, 586.14 nM, and 2.17 μM, respectively (Figure 6G). These findings showed that vitamins K₁, K₂, and K₃ were able to bind AhR and are likely to modulate its activity.

We then tested whether Pht analogs, vitamins K1, K2, and K3, had an inhibitory effect on the P. aeruginosa pqs QS system. The growth of the reporter strain (OD_600nm) was monitored to ensure that vitamins K₁, K₂, or K₃ would not affect the growth of P. aeruginosa strains (Figure 7A). GFP expression was normalized by dividing the GFP values by the OD values measured at respective time points. Vitamin K₁, K₂, and K₃ were found to inhibit pqsA-gfp expression dose-dependently without affecting growth (Figures 7B–D). In conclusion, Pht and its analogs, vitamins K1, K2, and K3, inhibited the P. aeruginosa pqs QS system. In addition, we performed molecular docking to predict the binding modes of vitamins K₁, K₂, and K₃ to PqsR protein (Supplementary Figure 2).