Materials and solvents
The antibiotics used in this study include COL, aztreonam (ATM), ceftazidime (CAZ), cefepime (FEP), imipenem (IPM), ciprofloxacin (CIP), levofloxacin (LVX), gentamicin (GEN), and tobramycin (TOB) were purchased from Wenzhou Kangtai Biological Technology Co, Ltd. (Zhejiang, China). The solvents and diluents used to prepare antibiotics comply with the latest guidelines issued by the Institute of Clinical and Laboratory Standards Institute (CLSI 2020). CR (CAS NO:42553-65-1, purity ≥ 98%) was purchased from Yuan-ye Bio-Technology Co., Ltd (Shanghai, China) and dissolved in double-distilled H2O.
Bacterial isolates
A total of 8 non-duplicated strains of K. pneumoniae (Table 1) were isolated from the First Affiliated Hospital of Wenzhou Medical University in China [13]. Three strains of K. pneumoniae were from 2019, and the other five strains were selected from 2015 to 2017. All isolates were quick-frozen in Luria Bertani (LB) (Sigma-Aldrich, St. Louis, USA) broth medium supplemented with 30% (vol/vol) glycerol at − 80°C.
Antimicrobial and non-antibacterial drugs susceptibility test
The MIC values of antibiotics and CR were determined by the cation-regulated Mueller-Hinton broth (CAMHB) micro-dilution method [14]. COL or CR was diluted twice, COL(128 ~ 1)µg/ml and CR (512 ~ 0.5) µg/ml were added to the CAMHB 96-well plate for two times dilution. The final 7.5 × 105 CFU/ml bacterial solution was added to each well and cultured with COL or CR at 37℃ for 16 to 20 h. The cell growth was observed with the naked, and the minimum inhibitory concentration (MIC) was defined as the lowest concentration for inhibiting the visible growth of the tested strains. According to CLSI 2020, the breakpoint of COL was explained.
Checkerboard analysis
The synergistic effect between COL and CR was determined by checkerboard experiments described in the previous study [15]. Briefly, COL and CR combine at different concentrations to produce different concentration combinations. The remaining steps include adding bacteria and measuring growth. The checkerboard test was carried out on a 96-well plate by micro-dilution method. After 16–20 h, the compound with the lowest inhibitory concentration to visible bacteria growth was selected as MIC. The fractional inhibition concentration index (FICI) of two compounds, A and B, is defined by the following formula: FICI = FICA + FICB= (CA/MICA) + (CB/MICB). MICA and MICB are the minimum inhibitory concentrations of compounds A and B alone. Compounds A and B are CR and COL, respectively. The minimal combined inhibitory concentrations of compounds A and B are CA and CB. (FICI ≤ 0.5, indicated synergy; 0.5 < FICI ≤ 1, indicated an additive effect; 1 < FICI ≤ 2, indicated an irrelevant effect; FICI > 2, indicated an antagonistic effect) [16].
Time-kill assays
In order to determine the effect of this combination on the growth kinetics of Col-R K. pneumoniae, the time-kill assay was conducted to evaluate the synergistic effect between COL and CR, following the results obtained from the checkerboard assay [17]. Firstly, untreated bacteria served as a negative control; the bacteria were cultured alone or in combination with CR and COL at a concentration of 1.5 × 106 CFU/ml. The drug concentrations used for the time-kill curve were derived from checkerboard results, with FICI < 0.5. The bacterial suspension was cultured at 37°C in a 180 rpm shaker for 2, 4, 6, 12, and 24 h, and a suitable diluent was prepared according to the bacterial growth rate. At the selected time point, 100 µl of the culture medium was inoculated on the mueller-hinton agar (MHA) plates, and CFU were counted overnight at 37°C. The synergistic activity was defined as the decrease of ≥ 2log10 CFU/ml at 24 h when the two drugs were used together compared with either being used alone [18–19].The data from each group are represented by an average of ± SD (n = 3 for each group).
In vivo evaluation of synergy in the mice infection model
Meanwhile, the in vivo curative effect was verified by establishing a mouse thigh infection model with neutropenia [20]. Male BALB/c mice aged 5–6 weeks, weighing about 25 g (Charles River, Hangzhou, China) were used in the experiment. The mice were raised according to the Chinese National Standard for Experimental Animals (GB14925-2010). All animal studies were approved by SYXK of the Zhejiang Association of Science and Technology [ID:SYXK (Zhejiang) 2018-0017] and carried out by the Animal Welfare and Ethics guidelines of Wenzhou Laboratory.
The strain K. pneumoniae FK 6663 was selected based on the pre-experimental of bacteria virulence. The mice were randomly divided into 4 groups, with 3 mice in each group. Cyclophosphamide was injected intraperitoneally for 3 days before thigh infection in mice (150 mg/kg). Then 100 µl of K. pneumoniae suspension (1.0 × 106 CFU per mouse) was injected into each mouse. The lower concentration of CR (20 mg/kg) was chosen following the pharmacological trial of saffron in vivo [21]. 2 h after infection, CR (20 mg/kg), COL (7.5 mg/kg) or both (20 + 7.5 mg/kg) were injected intraperitoneally. 24 h after infection, the mice were killed by cervical dislocation. The thighs were aseptically removed, homogenized, continuously diluted, and inoculated on trypsin soybean Agar to count the bacteria. The data from each group are represented by an average of ± SD (n = 3 for each group).
Biofilm formation inhibition assays
Based on the checkerboard test results and the results of bacterial biofilm formation ability in the pre-experiment, biofilm inhibition and eradication tests were carried out on 4 strains of bacteria [22]. The bacteria were inoculated on the plate and cultured overnight. Adjust the bacterial suspension to 0.5 McFarland, dilute 1:100 in fresh LB broth and add 100 µl to a 96-well plate. Then, add 100 µl drug solution separately or in combination to a 96-well plate. After incubation at 37℃ for 24 h, they were washed twice with 200 µl × phosphate balanced solution (PBS) (Sigma-Aldrich, Milan, Italy) to remove planktonic bacteria. The quality of stable biofilm was evaluated by crystal violet staining. Then, 1% crystal violet (CV) solution 150 µl was added to each well, and 15 min were stained. After staining, the CV was removed and washed with 1 × PBS solution three times. The combined CV was then solubilized with 150 µl ethanol-acetone (96:5 vol/vol). Read the absorbance at 595 nm on the mini tablet reader (Multiskan FC). P < 0.05 was significant. The data from each group are represented by an average of ± SD (n = 3 for each group).
Removal of mature biofilm assays
First, adjust the bacterial samples to the 0.5 McFarland standard with sterile saline, then dilute 100 times with LB broth. 200 µl bacterial solution was added to each well, and 20 ~ 24 h was incubated at 37℃. After removing the planktonic cells, CR and COL of synergistic concentration were used alone or in combination, respectively. After incubating at 37℃ for 20 ~ 24 h, the culture medium was removed, washed twice, and dried. Then, as mentioned earlier, crystal violet staining was used, and the absorbance was measured at 595 nm [23]. The data from each group are represented by an average of ± SD (n = 3 for each group).
SEM
The effect of COL/CR on the biofilm was observed by scanning electron microscopy (SEM). FK3994 was used as an experimental strain based on the results of the bacterial biofilm formation ability in the pre-experiment. 1800 µl LB solution and 200 µl bacteria solution were added to the well of the 6-well plate with sterilized glass cover (batch number: NO10211818, Citoglas Co., Ltd., China) and first cultured in LB to the middle stage. Then, clean the cover with aseptic PBS twice. The biofilm cover slides were divided into a control group, CR single treatment group, COL single treatment group and combination group. The single treatment group was added with LB 2000 µl of 4 µg/ml COL or 64 µg/ml CR, and the combined group was cultured with LB 2000 µl of 4 µg/ml COL and 64 µg/ml CR at 37℃ for 24 h. Then wash the biofilm with aseptic PBS twice, fix the biofilm with 2.5% glutaraldehyde (Merck), then dehydrate for 15 min in several stages with multiple concentrations of ethanol (30%, 50%, 70%, 90% and 100% vol/vol) and air-dry naturally when connected to a vacuum container [24]. According to the manufacturer's instructions, scanning electron microscope observation was performed at 10 kV using HITACHISU8010.
Safety evaluation
In order to determine the safety of CR at the concentration used, the erythrocyte hemolysis test, in vitro cytotoxicity test and in vivo toxicity test of G.mellonella were carried out.
We verified the effect of CR on red blood cells (RBCs) and evaluated the drug's safety [18]. Briefly, fresh blood of mice was collected and centrifuged to remove plasma and monocytes. The red blood cells were washed with aseptic PBS three times and centrifuged at 3000 rpm/min for 5 minutes. The RBCs suspension was diluted with normal saline to prepare 5% RBCs suspension. Add 250 µl of CR (16–128µg /ml) of different concentrations to 250 µl of 5% erythrocyte suspension. After incubating in a 37℃ water bath for 1 h with 4000 rpm centrifugal 5 min, the supernatant was adsorbed into the 96-well plate. The absorbance at 545 nm was measured to reflect the hemolysis rate. Only sterile PBS was added in the negative control group, and 0.1% TritonX-100 was added in the positive control group.
RAW264.7 cells were cultured in Dulbecco-modified Eagle medium supplemented with 10% heat-inactivated fetal bovine serum and incubated in the 37℃ 5% CO2 incubator until fusion [13]. The fusion cells were obtained by trypsin treatment. The experiment was divided into the control, CR, and combined groups. First, a 100 µl cell suspension containing 1 × 105 cells was inoculated into a sterile 96-well plate.10µl CR (16, 32, 64, 128µg/ml), 10 µl COL (2, 4, 8 µg/ml), and 5 µl CR + 5µl COL (64 + 8, 64 + 4, 64 + 2)µg /ml was added to the inoculated 96-well plate. The group without reagent was the negative control. After incubating in a 5% carbon dioxide incubator at 37℃ for 18 h, 10 µl CCK-8 (Toshindo Laboratory, Japan) was added to each well and incubated in the dark for 1 h. Finally, the absorbance of each well at 450 nm is recorded with a miniature tablet reader.
The larvae weighing about 200 mg and having similar activity are selected for in vivo toxicity test [25]. Ten larvae in each group were injected with different concentrations of CR solution (32, 64, 128, 256µg/ml) through the lower part of the left foot of the larvae. The control group was injected with 10 µl PBS. The larvae were cultured in an incubator at 37℃. The larvae were observed continuously for seven days to record whether the larvae died or not. The larvae are considered dead when they do not respond to external physical stimuli.
Outer membrane permeability test
We determined the outer membrane permeability by measuring the uptake of fluorescent dye N-phenyl-1-naphthylamine (NPN), adapted from the previous study [26]. The FK6663 bacterial suspension was incubated overnight at 37°C, centrifuged at 5000 rpm for 10 min, and washed with aseptic PBS three times. Then fresh LB broth and different solution concentrations were added and incubated at 37℃ for 2 h and incubated in 10 M NPN at 37℃ for 30 min. Fluorescence was measured and recorded as F0 by using an FMI 4500 fluorescence spectrophotometer (Hitachi, Japan) under the following parameters: excitation λ "=" 350 " nm and emission λ" = "420" nm. The data from each group are represented by an average of ± SD (n = 3 for each group).
Reactive oxygen species (ROS) detection
The intracellular ROS level of bacteria was detected with 2',7'-dichlorodihydrofluorescein diacetate (H2DCFDA)(Solarbio, China)[27]. First, 0.5 McFarland turbidity FK6663 bacteria solution was centrifuged and washed twice with PBS, then re-suspended with LB broth. Then, CR and COL were added to each group individually or jointly, and they were shaken and cultured at 37℃ and 180 rpm for 2 h, then centrifuged and washed again, and then re-suspended with 1ml 30mM H2DCFDA. 200 µl of cell suspension was added to each well of the black 96-well plate. Subsequently, the fluorescence was measured at the excitation and emission wavelengths of 488nm and 535 nm using the multimode plate reader (VICTOR Nivo, PerkinElmer, USA). The data from each group are represented by an average of ± SD (n = 3 for each group).
Fluorescent staining of living cells and dead cells
We further verified the membrane permeability, according to the previously reported research [29]. In short, a 0.5 McFarland turbidity bacteria solution of K. pneumoniae FK3994 was cultured with CR, COL, or a combination of both for 12 h. The selection of drug concentration refers to the chessboard experiment. Then stained with Syto9 (100 µg/ml) and propidium iodide (50µg/ml) at 37°C for 30 min. Then the fluorescence intensity was evaluated by a microplate reader (BioTek) (excitation at 535 nm, emission at 615 nm). The survival of bacteria was directly observed by laser confocal microscopy (Laika®, Japan).
Statistical analyses
Statistical analysis was performed using GraphPad Prism 8. All data were presented as mean ± standard deviation (SD) of at least three independent trials. The statistical significance of the difference was evaluated by Student’s t test and One-way analysis of variant (ANOVA). P-values < 0.05 were considered statistically significant. (ns, not statistically significant; *, P < 0.05; **P < 0.01; ***,P < 0.001; ****, P < 0.0001)