A single nosocomial specimen of AbS was collected from the wound surface exudates of a patient admitted to the Department of Burns and Plastic Surgery at Ruijin Hospital (Shanghai Jiaotong University School of Medicine, Shanghai, China) in 2008. Antibiotic sensitivity was assessed according to the guidelines provided by the Clinical and Laboratory Standards Institute (CLSI) (36), which included using ATB test strips (BioMérieux, Marcy l'Etoile, France) and the Kirby-Bauer disk diffusion method with antibiotic discs from Oxoid, Ltd. (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Agrobacterium tumefaciens strain KYC55 was used as a biosensor, and was kindly provided by Professor Jun Zhu (College of Life Sciences, Nanjing Agricultural University, Nanjing, China). A. baumannii was cultured statically in Luria-Bertani (LB) medium or Mueller-Hinton (MH) medium (Oxoid, Basingstoke, UK) at 37°C; A. tumefaciens KYC55 was cultured statically in LB medium at 28°C.

A. baumannii and A. tumefaciens were stored at −80°C in bacteria stock solution (Beyotime Biotechnology, Shanghai, China). A. baumannii were inoculated on LB agarose plates and incubated overnight at 37°C. Individual colonies (1×10⁸ colony-forming units (CFU)/ml) were selected and cultured in 15 ml LB medium at 37°C.

For HPLC-MS, bacteria were cultured in 500 ml LB medium from the overnight LB agarose plates, and 500 ml bacterial liquid was collected at 8 h, as determined by the AHL activity curve. Bacterial samples were centrifuged (4,500 × g for 20 min) and supernatants were passed through a 0.22 µm filter. An equal volume of 100% ethyl acetate was added to the filtrate, and the ethyl acetate phase was collected for AHL extraction and dried in a vacuum centrifuge. The residue was the AHL extract and was then re-dissolved in 50 µl ethyl acetate.

AHL activity was measured at 4, 8, 16, 24, 32, 40 and 48 h after seeding. At each time point, 3 bacterial liquid were collected and the optical density (OD) 600 was measured. AHL extracts in 50 µl ethyl acetate from 4, 8, 16, 24, 32, 40 and 48 h were added to cultures of A. tumefaciens KYC55 cultures, and β-galactosidase activity was measured to indirectly indicate AHL activity, as described previously (15,37). Following overnight incubation, the OD₆₀₀ was measured and 0.8 ml Z buffer (in each litre containing: 16.1 g (Na₂HPO₄) 7H₂O, 5.5 g (NaH₂PO₄) H₂O, 0.75 g KCl, 0.245 g (MgSO₄) 7H₂O, 2.7 ml 2-mercaptoethanol, adjusted to pH 7.0 with HCl), 10 µl 0.05% sodium dodecyl sulphate, 15 µl chloroform and 0.1 ml ortho-nitrophenyl-β-galactoside (4 mg/ml) were added, with a final sample volume of 0.2 ml. The time (T) taken for the solution to turn yellow was recorded, and 0.6 ml 1 M Na₂CO₃ was added to terminate the reaction. The OD₄₂₀ of supernatants was determined, and relative AHL activity was calculated as follows: Activity in Miller units=(1,000xOD₄₂₀)/(OD₆₀₀xTx0.2).

AHLs of different structures contain the same homoserine lactone (HSL) ring, and this moiety generates characteristic fragment ions at m/z 102 (25). Based on this principle, AbS AHLs were identified by HPLC combined with either TQ or Q-TOF HRMS using a 1200 HPLC-6140 TQ MS or a 1260 HPLC-6538 Q-TOF HRMS (Agilent Technologies, Inc., Santa Clara, CA, USA), respectively. Resultant chromatograms were compared with those of standard substances to elucidate the structure of AbS AHLs. The test conditions were as follows: An Agilent Poroshell 120 SB-C18 chromatographic column (2.7 µm, 2.1×100 mm; Agilent Technologies, Inc.) was used with acetonitrile and water as the mobile phase. The initial acetonitrile concentration was 40%, which was increased to 100% after 30 min, with a z-flow rate of 0.3 ml/min and sample injection volume of 10 µl. Positive-ion ESI was conducted with the ion source at 350°C. The dry N₂ flow rate was 8 l/min, and the air pressure of atomizing N₂ was 40 psi. The capillary voltage was 4,000 V. HPLC-TQ HRMS was performed with a precursor ion scan and daughter ion scan (collision energy, 15–30 units), while HPLC-Q-TOF HRMS involved MS1 and MS2 full scans (collision energy, 15–30 units).

An AbS mutant that is unable to produce AHLs was established using pKNG101.abaI::Km, as previously described (38); the pKNG101.abaI::Km plasmid was provided by Professor Philip N. Rather (Department of Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA, USA). This vector was transformed into Escherichia coli strain SM10, and the resultant SM10/pKNG101.abaI::Km was cultured with AbS in a filter-mating system in LB medium at 37°C without antibiotics for 24 h, after which it was cultured and screened on LB agarose plates containing 10% sucrose without NaCl. Sucrose resistance indicates that the bacteria have lost the integrated pKNG101 plasmid and therefore streptomycin sensitivity. AbS-M was screened for kanamycin resistance, and Southern blotting was used to confirm that colonies with this phenotype had abaI::Km disruption, as previously described (38).

The minimum inhibitory concentration (MIC) of common antibacterial drugs (including, meropenem, piperacillin, ceftazidime, ciprofloxacin, sulfamethoxazole/trimethoprim and minocycline) against AbS, AbS-M and AbS-M supplemented with 10 µmol N-3-OH-C₁₂-HSL (AbS-M+HSL; #53727; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) was assessed using the broth-micro-dilution method, according to the CLSI protocol. Briefly, overnight bacterial cultures were inoculated at 5×10⁵ CFU/ml in 1 ml MH medium containing a range of concentrations (128, 64, 32, 16, 8, 4, 2, 1, 0.5, 0.25, 0.125, 0.0625, 0.03125 and 0 µg/ml) of the antibacterial drugs. Following 24 h incubation at 37°C, the MIC against bacterial growth was assessed by visual examination. Each antibiotic concentration was tested three times.

A total of 45 µl 0.5 McFarland bacterial liquid (AbS, AbS-M and AbS-M+AHL) was added to 3 ml LB medium containing 0.125 µg/ml meropenem, with a final concentration of 10 µM AHL (N-3-OH-C₁₂-HSL). Alternatively, 45 µl 0.5 McFarland bacterial liquid was added into 3 ml LB medium without meropenem (AbS-U). The cultures were incubated at 37°C. After 24 h, 1 ml bacterial liquid was centrifuged (10,621 × g for 1 min) and supernatants were discarded. Total RNA was extracted using TRIzol Reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA), according to the manufacturer's protocol; concentration and purity were determined using an ultraviolet spectrophotometer. RNA was reverse transcribed into cDNA using the AMV First Strand cDNA Synthesis kit (New England Biolabs, Inc., Ipswich, MA, USA), according to the manufacturer's protocol. AbS-U, AbS, AbS-M and AbS-M+HSL cultures were incubated for 24 h and the expression levels of 16S rRNA, Oxacillinase (OXA)-51, AmpC type β-lactamase (AmpC), oxacillinase (OXA)-23, IMP type metallo-β-lactamase (IMP)-4, verona integron-mediated metallo-β-lactamase (VIM)-2, Acinetobacter drug efflux (Ade) A, AdeB and AdeC were assessed by quantitative polymerase chain reaction (qPCR) using a StepOnePlus Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.) and SYBR-Green Master Mix (Thermo Fisher Scientific, Inc.); primers used are listed in Table I. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression was measured as a reference, and gene expressions were calculated in terms of fold change using the comparative Cq method; relative mRNA expression was calculated using the 2^−ΔΔCq method (39). The experiments were repeated 3 times.

Data were presented as the mean ± standard deviation and analysed using Student's t-test, analysis of variance and least significant difference pot hoc test with SPSS version 19.0 (IBM SPSS, Armonk, NY, USA). P<0.05 was considered to indicate a statistically significant difference.

AbS growth rate and AHL activity were measured periodically between 4 and 48 h incubation (Fig. 1). AHL activity increased from 5.00±1.00 Miller units at 4 h culture to a maximum of 279.33±27.59 Miller units at 8 h. Subsequently, the activity decreased to 28.67±4.16 Miller units at 16 h and plateaued. The AbS growth curve did not correlate with AHL activity after bacterial growth reached the log phase; the OD₆₀₀ (bacterial growth) peaked at 0.90±0.01 after 24 h of culture and then plateaued.

AHLs that were extracted from AbS culture supernatants using ethyl acetate were screened using HPLC-TQ MS. As presented in Fig. 2, the precursor ion scan at m/z 102 detected 30 precursor ions, including those at m/z 282, 284 and 300 (each precursor ion represents one compound). MS2 spectrum analysis of these 30 ions confirmed that they could generate fragment ions at m/z 102, suggesting that they may be AHLs.

In the positive ionization mode of ESI, extracted AHLs generated a quasi-molecular ion at m/z 300 and major fragment ions at m/z 102 and m/z 74 (Fig. 3A). The ion at m/z 102 was the most abundant, and in order to determine its structure, HPLC-Q-TOF HRMS was used to examine the elemental composition of this ion and related daughter fragment ions. The elemental composition at m/z 300 was C₁₆H₃₀NO₄, representing the [M+H] ⁺ ions of N-3-OH-C₁₂-HSL. The two major fragment ions were C₄H₈NO (m/z 102) and C₃H₈NO (m/z 74), both of which were derived from the HSL ring of N-3-OH-C₁₂-HSL, which was the only AHL molecule identified (Fig. 3B). The major fragmentation pathway is shown in Fig. 4. In addition, some low-abundance ions were detected in the MS2 spectrum (m/z 121, 97 and 83), and they contained only two elements, H and C. We hypothesized that these were derived from the fragmentation of carbon chains near the acyl group.

Since components from the culture media may interfere with AHL detection, HPLC-Q-TOF HRMS and tandem MS were used to scan for AHL molecules identified in previous screens. The present study also determined the elemental compositions of the 30 precursor ions and their corresponding daughter ions at m/z 102. The results revealed that only ions detected at m/z 300≥102 met the structural requirement for AHLs. The daughter ions at m/z 102 were derived from 29 candidate molecules that contained C₅H₁₂NO and therefore could not be AHLs.

According to the composition and degree of unsaturation, the signal molecule at m/z 300 was inferred to be N-3-OH-C₁₂-HSL. HPLC-MS was then used to examine a commercially available N-3-OH-C₁₂-HSL, and the results confirmed that it was structurally identical to the N-3-OH-C₁₂-HSL detected in the present study.

AHLs were isolated from the supernatant of AbS and AbS-M cultures incubated for 8 h at 37°C and the activity levels were analysed. AHL activity was significantly lower in AbS-M (12.67±1.53 Miller units) compared with wild-type AbS (255.67±16.01 Miller units) (P<0.01; Fig. 5).

AbS was sensitive to amikacin, cefuroxime, ceftazidime, imipenem, gentamicin, ciprofloxacin, sulfamethoxazole/trimethoprim, sulperazone, tazocin, cefepime, panipenem, meropenem, ampicillin, Unasyn (which is a combination of ampicillin and sulbactam) and piperacillin. Then the MIC of AbS, AbS-M and AbS-M + HSL cultures to meropenem, piperacillin, ceftazidime, ciprofloxacin, sulfamethoxazole/trimethoprim and minocycline was assessed (Table II). The MICs of meropenem and piperacillin against AbS-M (0.25 and 1 µg/ml, respectively) were lower than the MICs of these antibiotics against wild-type AbS (0.5 and 2 µg/ml, respectively). However, the addition of HSL to the AbS-M culture raised the MICs of meropenem and piperacillin to similar levels as wild-type AbS (0.5 and 2 µg/ml, respectively). By contrast, the MICs of ceftazidime, ciprofloxacin, sulfamethoxazole/trimethoprim and minocycline were similar for AbS, AbS-M and AbS-M + HSL.

AbS, AbS-M and AbS-M + HSL were cultured for 24 h in LB medium supplemented with meropenem (0.125 µg/ml), and AbS untreated with meropenem (AbS-U) was additionally cultured. The mRNA expression levels of drug-resistance genes were assessed by qPCR (Table III). Meropenem treatment increased the expression of OXA-51, AmpC, AdeA and AdeB in all three bacterial cultures. The mRNA expression levels of these four genes were significantly lower in AbS-M compared with wild-type AbS; however, supplementation of AbS-M cultures with N-3-OH-C₁₂-HSL increased the mRNA expression of these four drug-resistance genes to higher levels compared with wild-type AbS and untreated AbS-M. The expression of OXA-23, IMP-4, VIM-2 and AdeC could not be detected in any of the three strains.