Optical density was recorded on a Multiskan™ FC Microplate Photometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA) and fluorescence was measured on an Infinite^® M1000 PRO Multimode reader (Tecan Group, Ltd., Mannedorf, Switzerland). Water was purified using a Milli-Q water purification system (EMD Millipore, Billerica, MA, USA).

Institute of Cancer Research (ICR) mice were obtained from Lingchang Biotechnology Co., Ltd. (Shanghai, China). All procedures were performed according to the Qiqihar Medical University guidelines, and approved by the Ethics Committee for Animal Care and the Use of Laboratory Animals at Qiqihar Medical University (Qiqihar, China). Specific pathogen-free ICR mice were bred in the animal facility of Qiqihar Medical University.

The MRSA strain, SA1199B, derived from a methicillin-susceptible S. aureus bloodstream isolate from a rabbit endocarditis model, previously described by Kaatz et al (16), which overexpresses the gene encoding NorA, was used in the present study (17). Minimum inhibitory concentrations (MICs) were determined following the guidelines of Clinical Laboratory Standards Institute (18). Mueller-Hinton broth (MHB; Oxoid; Thermo Fisher Scientific, Inc.), containing 20 mg/l Ca²⁺ and 10 mg/l Mg²⁺, was used in the assay. All bacterial suspensions were adjusted to 5×10⁵ Cfu/ml for the bioassay, cells were seeded at a density of 5×10⁴ cells/well in 96-well plates. Ofloxacine (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) served as the positive control and 2% DMSO in MHB as the negative control. Controls and treatments were assessed in duplicate. Following the incubation at 37°C for 18–24 h, 20 µl MTT solution (5 mg/ml; Biosharp, Solon, OH, USA) was added into each well and incubated at 37°C for 2–4 h. Concentrations that completely inhibited the visible bacterial growth (from light yellow to deep purple) were recorded as MIC values against the SA1199B strain. The optical density in each well was measured at 570 nm. The 50% effective concentrations (EC₅₀s) were calculated based on the percentage of bacteria growth inhibition following the treatment of the bacteria with different concentrations of compounds.

Using the strain SA1199B, a checkerboard microdilution assay was employed to examine for the presence of a synergistic interaction of antimicrobes (19). Jatrorrhizine or reserpine (both Dalian Meilun Biotech Co., Ltd., Dalian, China) in a concentration range of 2–128 mg/l (2, 4, 8, 16, 32, 64 and 128 mg/l) was tested in combination with NFX (Dalian Meilun Biotech Co., Ltd.) against the tested S. aureus strain; the concentration range of NFX was 0.5–256 mg/l (0.5, 1, 2, 4, 8, 16, 32, 64, 128 and 256 mg/l). Cells were seeded at a density of 5×10⁴ cells/well in 96-well plates. All treatments were tested in duplicate wells. Following the incubation at 37°C for 18–24 h, 20 µl MTT solution (5 mg/ml; Biosharp) was added into each well and incubated at 37°C for 2–4 h. MIC values were used to evaluate the effects of the combination by calculating the fractional inhibitory concentration index (FICI) according to the following formula:

‘Synergy’ was confirmed when the FICI was ≤0.5. When the FICI ranged from 0.5–4.0 the effect was regarded as ‘indifferent’, whereas an ‘antagonistic’ effect was confirmed when the FICI was >4.0.

The growth kinetics of SA1199B cells were assessed to elucidate the effect of NFX in the presence of jatrorrhizine in MHB, and evaluated using a time-growth curve method, as described previously (14,20). A bacterial suspension in its logarithmic phase (1×10⁶ Cfu/ml) was used as the inoculum. The growth kinetics were analysed following the treatment of NFX at 16 mg/ml (1/4 MIC) alone and in combination with jatrorrhizine at 16 mg/ml (1/4 MIC). The cfu/ml was determined by a serial dilution method (21) on Mueller-Hinton agar (MHA; Yeasen Biotech, Shanghai, China) plates at 37°C for 2, 4, 6, 8, 10, 12 and 24 h.

Jatrorrhizine was subjected for a bacterial drug efflux assay as described previously (22). The efflux activity of strain SA1199B was determined by measuring the accumulation of the fluorescent dye EtBr (25 µM; Aladdin Shanghai Biochemical Technology Co., Ltd., Shanghai, China) in the absence or presence of known efflux inhibitor carbonyl cyanide 3-chlorophenylhydrazone (CCCP; 100 µM; J&K Chemical, Ltd., Beijing, China) and in the absence or presence of jatrorrhizine (16 mg/l). Cells were seeded at a density of 5×10⁴ cells/well in 96-well plates. With an excitation wavelength of 530 nm and emission wavelength of 600 nm, the fluorescence of the suspension was monitored every 5 min for 1 h (slit width, 5 nm). The assay was performed in triplicate.

Total RNA was isolated from the bacteria using the TRIzol reagent (Thermo Fisher Scientific, Inc.) following the manufacturer's protocol. The reverse transcription step was conducted using the RevertAid™ First Strand cDNA Synthesis kit (Thermo Fisher Scientific, Inc.) to synthesise cDNA following the manufacturer's protocol. The semi-quantitative PCR analysis was performed using the 2X Taq PCR MasterMix (PC1120; Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) and with an ABI 7300 real-time fluorescent quantitative PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). The PCR thermocycling parameters were as follows: 95°C for 1 min followed by 40 cycles of 95°C for 15 sec, 58°C for 25 sec and 72°C for 20 sec, and a final extension step of 72°C for 5 min. The relative gene expression was calculated by the 2^−ΔΔCq method (23). Each sample was assessed in duplicate and all the samples were analysed in parallel for the expression of the housekeeping gene 16s rRNA, which was used as an endogenous control for normalisation of the expression level of target genes. The fold induction was determined from the mean replicate values. At the end of the RT-PCR assay, equal amounts of the final amplified product of each sample was loaded in 2% agarose gel and subjected to electrophoresis and visualized using ethidium bromide (1610433; Bio-Rad Laboratories, Inc., Hercules, CA, USA) staining. The densitometry analysis was performed with ImageJ version 1.48 software (National Institutes of Health, Bethesda, MD, USA). The primer sequences for the analysis of NorA and 16s rRNA genes are described in Table I.

The three-dimensional structure of jatrorrhizine was built using ChemDraw Ultra 14.0 and ChemBio3D^® Ultra 14.0 (CambridgeSoft Corp., Waltham, MA, USA). The structure was constructed using the Prepare Ligands modules in Discovery Studio software (version 2016; Biovia Corp., San Diego, CA, USA). The amino acid sequence of NorA was investigated in the National Center for Biotechnology Information database (GI: 153055; Accession: AAA26658.1). The glycerol-3-phosphate transporter from Escherichia coli (Protein Data Bank ID: 1PW4) was selected from the Protein Data Bank database and used as template for NorA in the docking studies (14). The NorA homology model for molecular docking was created on the I-TASSER server (https://zhanglab.ccmb.med.umich.edu/I-TASSER/), which is one of the computational tools for automated protein 3D structure prediction (24–26), then refined by Fragment-Guided Molecular Dynamics simulation (27). The optimized model was further evaluated using server-based structural verification from the UCLA-DOE Institute for Genomics and Proteomics server. The probable binding site of the modeled structure was identified using the Define and Edit Binding Site Module, and then applied to the Flexible Docking Module for flexible docking studies with Chemistry at HARvard Macromolecular Mechanics force fields in Discovery Studio software.

The intramuscular infection model was based on methods described previously (28) using 4–6 week-old male ICR mice (18–20 g). A total of 70 mice were provided with free access to food and water, and housed at room temperature (25°C) with a 12-h light/dark cycle and humidity of 65–70%. They were rendered neutropenic by intraperitoneally (i.p.) injecting them with 150 and 100 mg/kg cyclophosphamide at 4 days and 1 day prior to inoculation, respectively. The colonies of MRSA SA1199B grown on MHA plates were suspended in MHB and then incubated at 37°C overnight. The animals were injected intramuscularly with 1×10⁵ CFU MRSA SA1199B and antimicrobial treatment was started 1 h following bacterial challenge. The mice were divided into five groups treated (n=10/group) with a drug-free solution [0.5% w/v sodium carboxymethyl-cellulose, 0.5 ml/20 g; oral administration (p.o.)], jatrorrhizine (50 mg/kg, i.p.), NFX [100 mg/kg, p.o.], a combination of jatrorrhizine (10, 25 or 50 mg/kg, i.p.) and NFX (100 mg/kg, p.o.) or ofloxacine (50 mg/kg, p.o.; Dalian Meilun Biotech Co., Ltd.). The mice that received the drug-free solution were deemed the control group. A total of 24 h after infection, the mice were euthanized, the muscle was removed and then homogenized in 1 ml PBS to recover the bacteria. In order to count the viable MRSA colonies, samples were serially diluted and plated onto MHA plates, and incubated at 37°C for 24 h. The bacterial colonies (~1 mm) on agar plates counted visually were used for determing viable bacteria in samples. Bacterial quantities are displayed as ln(Cfu/ml) in all treatment groups.

The data are presented as mean ± standard deviation of two or three separate experiments. Significant differences were established by one-way analysis of variance and post hoc Tukey's test using GraphPad Prism (version 5; GraphPad Software, Inc., La Jolla, CA, USA). P<0.05 indicated that the difference between groups was statistically significant.

Jatrorrhizine was subjected to an antibacterial assay against MRSA strain SA1199B. The MICs of jatrorrhizine and NFX were measured as 64 mg/l; the two agents exhibited low antibacterial activity against MRSA SA1199B (Table II). As the positive control, the antibiotic ofloxacine substantially inhibited the growth of MRSA SA1199B with an MIC of 1 mg/l. 2% DMSO (negative control) showed no inhibition on bacterial growth. The EC₅₀s of jatrorrhizine and NFX were determined to be 27.1 and 21.6 mg/l, respectively. The EC₅₀ values were high, supporting the antimicrobial activities of jatrorrhizine and NFX.

Jatrorrhizine alone exhibited little antibacterial activity with an MIC of 64 mg/l (Table II and Fig. 1). The strain SA1199B was resistant to the antibiotic, NFX (MIC, 64 mg/l; Table II). When the bacteria were incubated with a combination of jatrorrhizine and NFX, the MIC of NFX was markedly reduced compared with NFX alone. The synergistic effect between jatrorrhizine and NFX against the strain SA1199B was confirmed with an FICI of 0.375. However, reserpine demonstrated a better synergistic effect with an FICI of 0.250.

To further confirm the synergistic effect between jatrorrhizine and NFX, time-growth experiments of the SA1199B strain without treatment or treated with NFX, jatrorrhizine, reserpine or a combination of two (Fig. 2). Jatrorrhizine did not significantly inhibit the SA1199B strain at 16 mg/l (1/4 MIC), as shown in Fig. 2. The SA1199B strain did not markedly increase with the first 4 h of 1/4 MIC NFX treatment; however, proliferation markedly increased from 6–10 h. Compared with NFX alone, the SA1199B strain treated with NFX (1/4 MIC) combined with jatrorrhizine (1/4 MIC) grew at a slower rate. Jatrorrhizine markedly enhanced the antibacterial activity of NFX against MRSA, suggesting the synergistic effect between NFX and jatrorrhizine, even the effect was weaker compared with reserpine in combination with jatrorrhizine.

Measuring the accumulation of EtBr demonstrated that jatrorrhizine inhibited the efflux pump of the SA1199B strain (Fig. 3). Compared with the vehicle-treated bacteria, in which fluorescence was quickly reduced to <50%, the reduction in fluorescence was significantly slower in the jatrorrhizine-treated bacteria (P<0.05) as the fluorescence level plateaued at ~70%. The positive control CCCP significantly (P<0.01) inhibited the efflux of EtBr cells with the fluorescence retention of ~80%. Therefore, jatrorrhizine inhibited bacterial drug efflux in the SA1199B strain, although its inhibition was significantly weaker compared with the known efflux inhibitor, CCCP (P<0.05). Drug efflux is one of the important bacterial resistance mechanisms. Thus, jatrorrhizine may be a resistance-modifying agent against bacteria resistance.

The expression of the NorA mRNA was investigated (Fig. 4A). The expression of the NorA mRNA in vehicle-treated bacteria was set as a 1-fold induction. Jatrorrhizine (16 mg/l, 1/4 MIC) significantly inhibited the expression of NorA compared with the vehicle (P<0.05; Fig. 4B). Compared with the vehicle, the expression of NorA was significantly increased in the SA1199B strain following the incubation with NFX (16 mg/l, 1/4 MIC; P<0.001) and a combination of NFX and jatrorrhizine (P<0.01). However, the level of NorA mRNA expression in bacteria treated with a combination of NFX and jatrorrhizine was significantly higher compared with that of NFX-treated bacteria (P<0.01). Therefore, jatrorrhizine significantly decreased the antibiotic-induced expression of NorA at the mRNA level, thus assisting in the treatment of MRSA.

According to docking study, the NorA-jatrotthizine complex was formed by NorA-jatrotthizine interactions cluding hydrophobicity, as shown in Fig. 5A. The functional groups in NorA involved in the jatrorrhizine binding site were formed by Met263, Asn315, Phe259, Ala312, Tyr316, Leu269, Thr270, Phe271, Arg380, Lys377, Glu268, Asn319 and Lys384 (Fig. 5B). It was observed that Ala312 and Thr27 are associated with key interactions and hence serve an important role in ligand binding. A hydrophobic cleft formed by Met263, Asn315, Phe259, Tyr316, Leu269, Phe271, Lys377 and Lys384 provided extra stability to the complex. Also, the electrostatic interaction between the benzene ring and polar groups of the amino acid residues on Glu268 and Arg380 contributed extra stability to the NorA-jatrotthizine complex. The docking results suggest that jatrotthizine binds to NorA with hydrogen bonds, and hydrophobic and electrostatic interactions, which may be associated with its inhibitory efflux pump activity.

To validate the in vivo synergistic bactericidal activity of jatrorrhizine and NFX against the MRSA strain, SA1199B, a mouse model of thigh infection was used. The results revealed that ofloxacine (50 mg/kg), a positive control, significantly decreased the bacterial burden after 24 h compared with that of the PBS-treated animals (Fig. 6). No significant difference in the bacteria count was identified between the groups treated with jatrorrhizine (50 mg/kg), NFX (100 mg/kg), a combination of jatrorrhizine NFX (10 and 100 mg/kg, respectively) or PBS. However, the bacterial count in the groups treated with a combination of jatrorrhizine (25 or 50 mg/kg) and NFX (100 mg/kg) significantly decreased in what appears to be a dose-dependent manner (P<0.01 and P<0.001, respectively, suggesting that there is an in vivo synergistic effect.