Discovery of structurally diverse polyprenylated acylphloroglucinols with quorum sensing inhibitory activity from Hypericum seniawinii Maxim.

1 Introduction

The discovery of antibiotics has provided an effective way to control life-threatening infections [1]. However, the long-term excessive usage of antibiotics has led to antimicrobial resistance [2–4]. With at least 700,000 people dying from drug-resistant infectious diseases every year, and the figure is predicted to increase to 10 million deaths per year by 2050 without effective measures taken, there is an urgent required for novel treatment strategies to combat antimicrobial resistance limitation [3–6]. Biofilms mediated by quorum sensing (QS) are known to be one of the vital factors for drug resistance [7]. The QS system is the cell-to-cell communication that controls various collective behaviors, including virulence factor production, biofilm formation, and bioluminescence, resulting in forming a barrier to escape from the harsh environment to assist microbial pathogenesis [8–10]. Inhibiting this system will reduce bacterial virulence, prevent biofilm formation without affecting bacterial growth, and potentially slow the development of resistance [8, 11, 12]. QS inhibition was an alternative therapeutic strategy for antimicrobial therapy [13, 14].

The secondary metabolites from plants, as a chemical defense, can resist the invasion of pathogenic bacteria in the environment and play a role in disease resistance, which are important sources for discovering novel antibiotic drugs [15–20]. Some have been identified as potential QS inhibitors, such as caffeine, hordenine, and iberin [21–25]. To discover more abundant scaffolds of QS inhibitors, we have focused on the plant Hypericum seniawinii, a perennial herb widely distributed in the southern region of China, which was a folk medicine used for detoxification, regulating menstruation, and activating blood [26, 27]. The chemical constituents of H. seniawinii lead to the isolation of four novel polyprenylated acylphloroglucinols, hyperisenins A–D (1–4), along with two known analogs (5 and 6) (Fig. 1). Compounds 1 and 2 were two highly degraded cyclohexanone-monocyclic polyprenylated acylphloroglucinols that might originate from bicyclic polyprenylated acylphloroglucinols (BPAPs) via a series of complex retro-Claisen, keto − enol tautomerism, and intramolecular cyclization, and compound 3 was a unique O-prenylated acylphloroglucinol with a 6/6/6 ring system. All isolates were assayed for QS inhibitory activity against Pseudomonas aeruginosa. The results showed that compound 4 could inhibit the QS system by decreasing the activation of the rhl system with no effects on the growth of P. aeruginosa, and reducing rhamnolipid levels by activating the las and pqs systems with molecular docking. Herein, the isolation, structural identification, plausible biogenetic pathways, and biological assay of the isolates were reported.

Open image in new window

2 Results and discussion

Hyperisenin A (1) was obtained as a colorless oil, and the molecular formula was defined as C₃₀H₄₀O₄ based on its HRESIMS data (m/z [M + Na]⁺: 487.2802, calcd 487.2819), which possessed eleven degrees of unsaturation. The ¹H NMR data showed five aromatic protons [δ_H 8.01 (2H, dd, J = 8.5, 1.4 Hz), 7.47 (1H, tt, J = 7.3, 1.4 Hz), 7.41 (2H, tt, J = 7.4, 1.4 Hz)], four olefinic protons [δ_H 6.15 (1H, d, J = 15.5 Hz), 5.98 (1H, dd, J = 15.5, 7.7 Hz), 5.15 (1H, m), 5.13 (1H, m)], and seven methyls [δ_H 1.74 (3H, s), 1.71 (3H, s), 1.66 (3H, s), 1.60 (3H, s), 1.38 (3H, s), 1.37 (3H, s), 1.07 (3H, s)] (Table 1). The ¹³C NMR spectrum showed 30 carbon signals (Table 1), which could be assigned to one ketone carbonyl (δ_C 196.0), one monosubstituted phenyl ring [δ_C 132.0, 130.5 (× 2), 129.4, 127.8 (× 2)], eight olefinic carbons (δ_C 166.8, 144.8, 135.6, 133.4, 122.3, 121.9, 119.0, 117.6), and three quaternary carbons including two oxygenated carbons (δ_C 76.3, 71.0), two methines, three methylenes and seven methyls. By analyzing the 1D NMR data and unsaturation, combined with the data reported in the literature [28–30], it is inferred that 1 was a polyprenylated acylphloroglucinol with a bicyclic system.

Its gross structure was confirmed by analyzing its 1D and 2D NMR data (Fig. 2). The HMBC correlations from H₂-4 to C-2, C-3, and C-6, from H₃-25 to C-1, C-5, and C-6, from H-13 to C-7, the ¹H-¹H COSY cross-peaks of H₂-4/H-5 and H-9/H-10/H-11/H-12/H-13, combined with chemical shifts of C-1 (δc 117.6) and C-7 (δc 166.8), indicated that the cyclohexanone system existed, and a benzoyl group converted into enol form was attached at C-1. Then, the HMBC correlations from H₂-15 to C-2, C-3, and C-4, from H₃-18 to C-16, C-17, and C-19, from H₃-24 to C-21, C-22, and C-23, from H₃-25 to C-6, from H₃-30 to C-28 and C-29, combined with the ¹H-¹H COSY cross-peaks of H₂-15/H-16, H-5/H₂-20/H-21, and H-26/H-27/H-28, revealed the existence of three side chains at C-3, C-5, and C-6. Further analysis of the downfield chemical shifts of C-3 (δc 76.3) and C-26 (δc 94.0) indicated that hydroxyl groups were attached to C-3 and C-26, respectively. The above-mentioned groups accounted for ten degrees of unsaturation. The remaining degree of unsaturation was attributed to a furan ring formed between C-7 and C-26, which was consistent with both the downfield chemical shift of C-26 (δc 94.0) and the HRESIMS data. Accordingly, the planar structure, featuring a cyclohexanone-monocyclic skeleton, was confirmed.

Open image in new window

The relative configurations of C-5, C-6, and C-26 in 1 were assigned by the NOESY spectrum, in which the cross-peaks of H-5/H-26 and H₃-25/H-27 indicated that H₃-25 was α-oriented, while H-5 and H-26 were on the same side with the β-orientations. The large coupling constant between H-27 and H-28 (J = 15.5 Hz) confirmed the E configuration. In order to determine the relative configuration of C-3, the calculated ¹³C NMR data with DP4+ analysis of two configurations (3R^*,5R^*,6S^*,26R^*)-1 and (3S^*,5R^*,6S^*,26R^*)-1 were applied at the mPW1PW91/6–311 + G^** level. The results showed that (3R^*,5R^*,6S^*,26R^*)-1 had a better linear correlation with 100% DP4+ probability (R² = 0.9990) (Fig. 3A), suggesting that the relative configuration of 1 was defined as 3R^*,5R^*,6S^*,26R^*. The absolute configuration of 1 was identified by ECD calculation at the PBE0/def2-TZVP level, the result showed that the calculated ECD (3R,5R,6S,26R)-1 curve matched well with the experimental one, allowing to assign its absolute configuration as 3R,5R,6S,26R (Fig. 4). Thus, the structure of 1 was elucidated.

Open image in new window

Open image in new window

Hyperisenin B (2), a colorless oil, possesses the molecular formula C₃₁H₄₀O₅ on the basis of the HRESIMS data (m/z [M + Na]⁺: 515.2753, calcd 515.2768). Its ¹H and ¹³C NMR spectra showed the presence of a monosubstituted benzene ring, two ketone carbonyls, one ester carbonyl, six olefin carbons, and seven methyl groups, which were similar to those of spirohypolactone B [31]. Further comparison of NMR data between 2 and spirohypolactone B revealed that they possessed the same skeleton, except for the acyl substituent and the number of the double bond [31]. The presence of a phenyl group at C-7 was verified by the ¹H-¹H COSY cross-peaks of H-9/H-10/H-11/H-12/H-13 and the HMBC cross-peak from H-13 to C-7, which was consistent with its 1D NMR data (Fig. 2). Furthermore, the 4-methylpenta-1,3-diene group was attached at C-6, deduced from the HMBC correlations from H₃-25 to C-6 and C-26, from H₃-30 to C-31, from H₃-31 to C-28, and the ¹H-¹H COSY cross-peaks of H-26/H-27/H-28. The relative configurations was the same as that of spirohypolactone B and norhyperpalum H assigned as 1R^*,3R^*,5S^*,6S^*,16R^* via the similar the NOESY cross-peaks of H-1/H-5, H₂-20/H₃-25, H₂-4a/H₂-15b, and H₂-4b/H₃-25, and along with the crucial absence of the NOESY cross-peaks of H₂-4/H-16 and H-5/H₂-15 (Fig. 2). To further confirm the relative configurations of C-3 and C-16, a DP4+ probability analysis of four isomers [A: (3R^*,16R^*), B: (3R^*,16S^*), C: (3S^*,16R^*), D: (3S^*,16S^*)] was conducted, and the calculated data of (1R^*,3R^*,5S^*,6S^*,16R^*)-2 has good linear correlations with the experimental data with 100% DP4+ probability (all data) (Fig. 3B and S40). This deduction was further supported by a critical 1D NMR comparison in the same solvent of 2 with similar compounds, spirohypolactones A and B, and norhyperpalum H [31, 32] (Figure S41). The Δ²⁶⁽²⁷⁾ double bond was assigned as E configuration based on the coupling constant (J_{H-26, H-27} = 15.4 Hz). Finally, the ECD calculation of 2 was conducted at the PBE0/def2-TZVP level, and its absolute configuration was determined to be 1R,3R,5S,6S,16R (Fig. 4).

Hyperisenin C (3), obtained as a yellow oil, had the molecular formula C₃₈H₅₀O₅ based on its HRESIMS data (m/z [M + Na]⁺: 609.3572, calcd 609.3550), suggesting fourteen indices of hydrogen deficiency. Its ¹H NMR spectrum showed the existence of five characteristic protons of the benzene ring [δ_H 7.57 (2H, dd, J = 8.2, 1.1 Hz), 7.52 (1H, tt, J = 7.5, 1.3 Hz), 7.37 (2H, t, J = 7.8 Hz)], three olefinic protons [δ_H 5.56, (1H, m), 5.22 (1H, m), 4.52 (1H, m)], and nine methyl groups [δ_H 1.81 (3H, s), 1.76 (3H, s), 1.70 (3H, s), 1.68 (3H, s), 1.59 (3H, s), 1.48 (3H, s), 0.96 (3H, s), 0.81 (3H, s), 0.76 (3H, s)] (Table 1). Its ¹³C NMR and DEPT data indicated 38 carbons (Table 1). The 1D NMR spectra characteristics of the above analysis indicated that 3 belonged to a polyprenylated acylphloroglucinol derivative.

Detailed analysis of its HSQC, HMBC, and ¹H–¹H COSY spectra indicated that 3 had the same skeleton as that of madeleinol A [33], and the main differences were acyl side chain and isopentenyl side chain (Fig. 2). The HMBC correlations from H₂-14 to C-2, C-3, and C-4, from H₃-17 to C-15 and C-18, from H₃-32 to C-30, C-31, and C-33, from H₃-33 to C-27, from H₃-37 to C-35 and C-38, combined with the ¹H–¹H COSY correlations of H-27/H₂-34/H-35 and H₂-14/H-15, suggested the location of the gem-dimethyl group, and two isoprenyl groups located at C-3 and C-27, respectively. Furthermore, the O-isoprenyl side chain was attached at C-4, which was deduced from the HMBC correlations from H₂-19 to C-4, from H₃-23 to C-20 and C-22, along with the ¹H–¹H COSY cross-peaks of H₂-19/H-20, and the downfield chemical shift of C-19 (δ_C 71.1). In addition, the benzene ring was attached at C-7 by analysis of its 1D NMR data and the HMBC correlation from H-9 to C-7, and the ¹H–¹H COSY cross-peaks of H-9/10/H-11/H-12/H-13. In the NOESY spectrum (Fig. 2), the cross-peaks of H-25/H-30, H-27/H-30, and H-30/H₃-33 indicated that these groups were cofacial, assigned as α-orientations, while H-24β/H₃-28, H₃-28/H-29β, and H-29β/H₃-32 clarified that they were β-oriented. The calculated ECD method was applied to determine its absolute configuration as 25R,26S,27S,30S (Fig. 4). Thus, a unique O-prenylated acylphloroglucinol with a 6/6/6 ring system was established.

Hyperisenin D (4) was also obtained as a yellow oil and had the molecular formula C₃₃H₄₂O₆ based on its HRESIMS at m/z 557.2876 [M + Na]⁺ (calcd 557.2874), which possessed thirteen degrees of unsaturation. The 1D NMR and HSQC spectra revealed 33 carbons (Table 1), including a benzoyl group [δc 193.9, 138.0, 133.5, 129.6 (× 2), 128.5 (× 2)], a geranyl group (δc 141.4, 132.1, 123.8, 116.7, 40.1, 36.3, 26.8, 25.9, 17.9, 16.4), two oxidated quaternary carbon (δc 72.2, 71.2), two methines (δc 93.0, 90.8), two methylenes (δc 31.6, 28.0), and four methyl groups (δc 27.6, 25.5, 23.8, 23.4), and the remaining six carbons were characteristic of a dearomatized phloroglucinol core including an enolic 1,3-diketone moiety (δc 182.3, 175.7, 114.7), one oxygen-bearing ene (δc 170.0, 113.3), and one quaternary carbon (δc 50.5). The mentioned groups occupied eleven degrees of unsaturation. Two additional rings should be formed in the structure of 4. Thus, the aforementioned evidence suggested that 4 should be a tricyclic dearomatized prenylated acylphloroglucinol derivative.

The planar structure was confirmed by analyzing its HSQC, HMBC, and ¹H-¹H COSY spectra (Fig. 2), similar to that of hypermonin C (5) [34], and the main difference was the isoprenyl side chain at C-5. The HMBC correlations from H₂-19 to C-5, C-6, and C-24, from H₃-22 to C-20, from H₃-23 to C-20 and C-21, combined with the ¹H-¹H COSY cross-peaks of H₂-19/H-20, indicated that the oxidized isopentenyl side chain was located at C-5. Furthermore, the downfield chemical shift of C-20 (δ_C 90.8), combined with degrees of unsaturation, indicated that C-6 and C-20 were connected via an oxygen atom to form a furan ring. In the NOESY spectrum (Fig. 2), the cross-peaks of H₃-18/H₃-27 and H-20/H₂-24 indicated that H-20 and the geranyl group at C-5 were on the same side, assigned as β-orientations, while H-15 was on the opposite side with α-orientation. Then, the NOESY cross-peaks of H-25/H₂-28 revealed the E-configuration of C-25/C-26 double bond. Thus, the relative configuration of 4 was determined to be 5S^*,15S^*,20R^*. Finally, the absolute configuration of 4 was defined as 5S,15S,20R by ECD calculation (Fig. 4).

Two known compounds, hypermonin C (5) [34] and vismiaguianone B (6) [35], were obtained from this plant. Their structures were confirmed by comparing the 1D NMR data with those of the literature.

Hyperisenins A (1) and B (2), possessing a unique cyclohexanone-monocyclic system, were proposed to biogenetically originate from BPAPs (Fig. 5) [28, 36, 37]. It underwent a retro-Claisen reaction to yield the crucial intermediate ⅰ, followed by two distinct pathways to form intermediates ⅱ and ⅲ. Subsequently, compound 1 was constructed from ⅱ via oxidation, keto − enol tautomerism, and intramolecular cyclization. On the other hand, ⅲ underwent oxidation and intramolecular cyclization to obtain compound 2.

Open image in new window

Considering QS is the vital target for antimicrobial therapy [13, 14] and Pseudomonas aeruginosa is an opportunistic pathogen that is typically resistant to multiple clinically available antibiotics [38], compounds 1–6 were evaluated for the QS inhibitory activity against P. aeruginosa (Figure S46). The results showed that compound 4 was a potential QS inhibitor that decreased the activation of the rhl system, as evidenced by the reduced fluorescence density of the reporter strain PAO1-rhlA-gfp (Fig. 6A). As expected, compound 4 did not affect the growth of P. aeruginosa, consistent with its role as a QS inhibitor. We further examined the production of rhamnolipids, a virulence factor regulated by the rhl system, in a clinically isolated carbapenem-resistant P. aeruginosa (CRPA) strain. Compound 4 significantly reduced rhamnolipid levels at a concentration of 100 µM (Fig. 6B). P. aeruginosa’s QS system includes two other well-defined pathways apart from the rhl system: the las and pqs systems [8]. Due to the limited yield of compound 4, its potential mechanism was explored through molecular docking with three QS receptors (lasR, rhlR, and pqsR).

Open image in new window

Unexpectedly, compound 4 failed to dock with rhlR but successfully engaged with lasR and pqsR, demonstrating significant affinity within their ligand-binding pockets. For lasR, it is proposed that compound 4 forms two aromatic hydrogen bonds with Tyr47 and a combination of a hydrogen bond and an aromatic bond with Tyr56, yielding a docking score of -6.476 kcal/mol (Fig. 6C). Regarding pqsR, compound 4 appears to bind to the ligand-binding site, albeit slightly shifted towards the dimer interface, forming a hydrogen bond with Glu151 in chain A and both a hydrogen bond and an aromatic bond with Glu151 in chain B, resulting in a docking score of -6.217 kcal/mol (Fig. 6D). Given that both the las and pqs systems enhance the activation of the rhl system [8], we hypothesize that the binding of compound 4 to lasR and pqsR may competitively inhibit their contribution to the rhl system, thereby regulating rhamnolipid production (Fig. 6E).

In summary, the phytochemical investigation of the dried aerial parts of H. seniawinii Maxim. resulted in the isolation of four undescribed polyprenylated acylphloroglucinols (1–4), as well as two known analogs (5 and 6). All isolates were obtained from this plant for the first time. Compounds 1 and 2 were two degraded polyprenylated acylphloroglucinols bearing the unique cyclohexanone-monocyclic system, and their plausible biosynthetic pathway was proposed. Furthermore, compound 4 was a potential QS inhibitor that decreased the activation of the rhl system and reduced rhamnolipid levels. Its mechanism might be the ability to bind between 4 lasR and pqsR. Our findings might provide a potential candidate as QS inhibitors to treat infectious diseases for further research.

3 Experimental section

3.1 General experimental procedures

Optical rotations were recorded with a PerkinElmer 341 polarimeter (PerkinElmer Inc., Fremont, California, USA). UV spectra were obtained in CH₃OH using a Lambda 35 instrument (PerkinElmer Inc., Fremont, California, USA). ECD spectra in CH₃OH were detected on a JASCO-810 spectrometer (JASCO, Tokyo, Japan). A Bruker Vertex 70 FT-IR spectrophotometer (Bruker, Karlsruhe, Germany) was used to acquire IR spectra. 1D and 2D NMR spectra were collected using Bruker AM-400 and AV-600 spectrometers. HRESIMS data were measured with a Bruker micOTOF Ⅱ and SolariX 7.0 spectrometer (Bruker, Karlsruhe, Germany). Analytical HPLC was performed on a Dionex HPLC system with a DAD detector, and semipreparative HPLC was performed on an Agilent 1200 system equipped with a reversed-phase (RP) C₁₈ column (5 µm, 10 × 250 mm, Welch Ultimate XB-C₁₈). Column chromatography (CC) including Silica gel (100–200 and 200–300 mesh; Qingdao Marine Chemical Inc., China), ODS (50 μm, YMC Co. Ltd., Japan), Sephadex LH-20 (Pharmacia Biotech AB, Uppsala, Sweden), and MCI gel (75–150 μm, Merck, Germany) was used to separation and purification of the sample.

3.2 Plant material

The aerial parts of H. seniawinii Maxim. were collected from the Shennongjia area in Hubei Province, People’s Republic of China. The plants were identified by Prof. C. G. Zhang of Huazhong University of Science and Technology (HUST). A voucher specimen (no. HP20230824) was deposited in the herbarium of Tongji Medical College, HUST.

3.3 Extraction and isolation

The dried aerial parts of H. seniawinii Maxim. (20.0 kg) were powered and extracted with 95% EtOH (3 × 25 L) at room temperature, removing the solvents in vacuo to yield the crude extract (0.4 kg). The crude extract was then suspended in water and successively partitioned with CH₂Cl₂ and EtOAc. The CH₂Cl₂ extract was separated into seven fractions (A–G) by a silica gel CC (100–200 mesh), eluted with a gradient of petroleum ether–ethyl acetate (80:1–0:1). Fr. E was subsequently further chromatographed on an RP-C₁₈ column (CH₃OH–H₂O, 50:50 to 100:0) to yield eight subfractions, E3a–E3h. Fr. E3b (1.0 g) was applied to Sephadex LH-20 (CH₃OH), obtaining three subfractions, Fr. E3b1–E3b3. Fr. E3b2 (500 mg) was then purified on semipreparative HPLC to afford 2 (3.3 mg, t_R 19 min, CH₃OH-H₂O, 91:9, v/v, 2 mL/min), and 6 (3.6 mg, t_R 27 min, CH₃OH-H₂O, 80:20, v/v, 2 mL/min). Fr. E3c (1.2 g) was conducted on Sephadex LH-20 (CH₃OH) and further purified by semi-preparative HPLC to afford 1 (7.5 mg, t_R 15 min, CH₃OH-H₂O, 88:12, v/v, 2 mL/min), 4 (1.7 mg, t_R 45 min, CH₃OH-H₂O, 77:23, v/v, 2 mL/min), and 5 (6.0 mg, t_R 12 min, CH₃OH-H₂O, 97:3, v/v, 2 mL/min). Compound 3 was isolated from Fr. E3e by silica gel CC (100–200 mesh), Sephadex LH-20 (CH₃OH), and HPLC (2.4 mg, t_R 32 min, CH₃OH-H₂O, 94:6, v/v, 2 mL/min).

Hyperisenin A (1): Colorless oil; [α]_D²⁹ +42.0 (c 0.1, CH₃OH); UV (CH₃OH) λ_max (log ε) = 201 (4.40) nm, 231 (4.06), 321 (3.77); IR (KBr) v_max 3424, 2973, 2929, 2873, 1721, 1634, 1603, 1553, 1490, 1449, 1382, 1363, 1328, 1233, 1180, 1148, 1066, 1027, 974, 838, 761, 694 cm^–1; ECD (CH₃OH) λ_max (Δε) 215 (+ 1.95), 241 (− 4.60), 365 (+ 1.81) nm. ¹H and ¹³C NMR data see Table 1; positive HRESIMS: m/z 487.2802 [M + Na]⁺ (calcd. for C₃₀H₄₀O₄Na⁺, 487.2819).

Hyperisenin B (2): Colorless oil; [α]_D²⁹ +48.8 (c 0.3, CH₃OH); UV (CH₃OH) λ_max (log ε) = 201 (4.49), 240 (4.28) nm; IR (KBr) v_max 3458, 2972, 2928, 1756, 1715, 1597, 1447, 1383, 1226, 1180, 988, 690 cm^–1; ECD (CH₃OH) λ_max (Δε) 206 (+ 4.15), 231 (− 3.45), 249 (+ 3.83), 267 (− 1.15), 303 (+ 5.13) nm. ¹H and ¹³C NMR data, see Table 1; positive HRESIMS: m/z 515.2753 [M + Na]⁺ (calcd for C₃₁H₄₀O₅Na⁺, 515.2768).

Hyperisenin C (3): Yellow oil; [α]_D²⁹ +33.0 (c 0.4, CH₃OH); UV (CH₃OH) λ_max (log ε) = 205 (4.29) nm; IR (KBr) v_max 3409, 2970, 2927, 2874, 1722, 1666, 1603, 1449, 1418, 1382, 1326, 1292, 1219, 1123, 1106, 1074, 1019, 955 cm^–1; ECD (CH₃OH) λ_max (Δε) 213 (+ 5.45), 260 (+ 0.02), 296 (+ 1.48), 340 (− 1.39) nm. ¹H and ¹³C NMR data; positive HRESIMS: m/z 609.3572 [M + Na]⁺ (calcd. for C₃₈H₅₀O₅Na⁺, 609.3550).

Hyperisenin D (4): Yellow oil; [α]_D²⁹ +22.1 (c 0.1, CH₃OH); UV (CH₃OH) λ_max (log ε) = 202 (4.60) nm; IR (KBr) v_max 3429, 2973, 2927, 2854, 1715, 1668, 1618, 1449, 1412, 1384, 1237, 1175, 1027, 961 cm^–1; ECD (CH₃OH) λ_max (Δε) 206 (+ 1.58), 215 (− 0.31), 264 (+ 2.99), 349 (− 0.04) nm. ¹H and ¹³C NMR data; positive HRESIMS: m/z 557.2876 [M + Na]⁺ (calcd. for C₃₃H₄₂O₆Na⁺, 557.2874).

3.4 NMR and ECD calculations

The details of NMR and ECD calculations were put in the Supporting Information (SI).

3.5 Strains and culture conditions

The GFP reporter strains were provided by Prof. Pinghua Sun from Jinan University and were cultured in Luria–Bertani (LB) medium (1% w/v NaCl, 1% w/v tryptone, and 0.5% w/v yeast extract) supplemented with 100 μg/mL ampicillin and 20 μg/mL gentamicin at 37 ℃. The CRPA strain was provided by Prof. Yan He from Tongji Hospital at Huazhong University of Science and Technology and cultured in LB medium at 37 ℃.

3.6 Screening for QS inhibitors

GFP reporter strains were cultured overnight in LB medium supplemented with 100 μg/mL ampicillin and 20 μg/mL gentamicin at 37 ℃. The cultures were then diluted 1:5 in ABTGC medium (2 g/L (NH₄)₂SO₄, 6 g/L Na₂HPO₄, 3 g/L NaCl, 1 mM MgCl₂, 0.1 mM CaCl₂, 0.01 mM FeCl₃, 0.025% thiamine, 0.5% glucose, 0.5% casamino acids). For the assay, 100 μL of either a 200 μM solution of the compound for high-throughput screening or gradient-diluted concentrations for secondary confirmation, or an equivalent volume of DMSO, was mixed with the diluted GFP reporter strain (100 μL). This mixture was transferred to a 96-well plate (200 μL per well) and incubated at 37 ℃. A microplate reader continuously monitored bacterial growth at OD₆₀₀ and GFP fluorescence (excitation at 485 nm, emission at 535 nm) every 15 min for approximately 24 h.

3.7 Determination of virulence factor production

CRPA was cultured overnight in LB medium at 37 ℃ and then 1:10 in ABTGC medium successively. A 100 μL volume of either gradient-diluted compound or an equivalent volume of DMSO was combined with the diluted CRPA (100 μL). This mixture was incubated for 24 h at 37 ℃ with agitation at 200 rpm in a 96-well plate (200 μL per well). The OD₆₀₀ of the culture was measured to normalize the virulence factor content. Virulence factors were then extracted from the supernatant after centrifugation at 4000 rpm for 15 min. Rhamnolipids were extracted with a three-fold volume of ethyl acetate three times. The ethyl acetate fraction was evaporated and redissolved in a freshly prepared reagent (0.19% orcinol in 60% H₂SO₄, 900 μL), incubated at 80 ℃ for 30 min, and then the solution (200 μL) was used to quantify rhamnolipid levels at OD₄₂₁ after cooling to room temperature.

3.8 Molecular docking

Molecular docking was performed using Schrödinger Maestro (Schrödinger Release 2023–1: Maestro, Schrödinger, LLC, New York, NY, 2023). Protein models were sourced from the Protein Data Bank (PDB) and prepared via the Protein Preparation Workflow module. Compound structures were pre-processed using the LigPrep module. Interactions and alignments were visualized and analyzed using PyMOL software.

Notes

Author contributions

Yonghui Zhang and Changxing Qi designed the experiments and revised the manuscript. Weiguang Sun and Guosheng Cao guided the experiments and wrote the manuscript. Yulin Duan contributed to the extraction, isolation, identification, and manuscript preparation. Xiaoxia Gu and Xincai Hao contributed to the bioactivity evaluation. All authors read and approved the final submission.

Funding

This work was financially supported by the National Key Research and Development Program of China (2021YFA0910500); National Natural Science Foundation of China (Nos. 32470422 and 32300335); the National Natural Science Foundation of Hubei Province (2023AFB791 and 2023AFB530); Knowledge Innovation Project of Wuhan Science and Technology Bureau (2023020201020534); Natural Science Foundation of Wuhan (2024040801020205); the Science and Technology Major Project of Hubei Province (2021ACA012); the National Key Research and Development Program (No.2023YFC2307004); the open foundation of Key Laboratory of Ethnomedicine (Minzu University of China), Ministry of Education (No. KLEM-KF202402); the open foundation of Key Laboratory of Tropical Medicinal Resource Chemistry of Ministry of Education (RDZH2024001); the open foundation of Hubei Key Laboratory of Wudang Local Chinese Medicine Research (Hubei University of Medicine) (WDCM2024002); the Fundamental Research Funds for the Central Universities (HUST: 2023JYCXJJ058).

Data availability

The data that support the findings of this study are openly available in the Science Data Bank at.

Declarations

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.