Antibacterial potential of Forsythia suspensa polysaccharide against resistant Enterobacter cloacae with SHV‐12 extended‐spectrum β‐lactamase (ESBL)

Abstract In this study, a homogenous polysaccharide (FSP), with an average molecular weight of 9.08 × 104 Da, was isolated from Forsythia suspense and its antibacterial potential against Enterobacter cloacae producing SHV‐12 ESBL was investigated. Growth kinetics, in vitro competition and biofilm formation experiments demonstrated that SHV‐12 ESBL contributed to a fitness benefit to E cloacae strain. The antibacterial activity of FSP (2.5, 5.0 and 10.0 μg/mL) was tested against E cloacae bearing SHV‐12 ESBL gene using bacterial sensitivity, agar bioassay and agar well diffusion assays. It was found that the addition of FSP demonstrated potent antibacterial activities against this bacterial as showed by the decrease of bacterial growth and the increase of the inhibition zone diameter. Furthermore, SHV‐12 ESBL gene expression was decreased in E cloacae strain following different FSP treatment in a concentration‐dependent manner. In conclusion, these data showed that FSP exhibited potent good antibacterial activity against E cloacae producing SHV‐12 ESBL via inhibition of SHV‐12 ESBL gene expression, which may promote the development of novel natural antibacterial agents to treat infections caused by this drug‐resistant bacterial pathogen.


| Isolation and purification of polysaccharide
The fruit of F suspensa (0.5 kg) was mechanically chopped into small pieces and preliminarily treated with 95% ethanol (6000 mL) there times and 3 hours each time at 90°C to remove low molecular components, including oligosaccharide, lipids and pigments in the samples. The residue was further immersed in boiling water (8000 mL) for three times and each time for 2 hours. Then, the extract was filtered and collected by centrifugation (1700 × g, 10 minutes) at room temperature. All supernatant were collected, concentrated and precipitated with adding 4 volumes of 95% ethanol at 4°C for 24 hours.
The precipitate was subsequently re-dissolved in distilled water and deproteinized five times by Sevag reagent (chloroform: butanol 4:1, v/v). After removing Sevag reagent, the remaining water layer was concentrated and added with 5 volumes of 95% ethanol (v/v) to precipitate the crud polysaccharide (CFSP, 22.7 g).
Each aliquot of 80 mg crude polysaccharides was added with 5 mL deionized water, centrifuged (1700 × g, 10 minutes) and filtered by a 0.45 μm membrane. The resulting filtrate was loaded onto a column (2.0 cm × 40 cm, Cl − form) of DEAE Sepharose Fast Flow and eluted with distilled water and a stepwise gradient of NaCl aqueous solution (0.2, 0.5 and 1 mol/L) at a flow rate of 4.0 mL/min to yield FSP1, FSP2, FSP3 and FSP4, respectively. Fraction with of 8 mL in each tube was collected using an automated fraction collector and monitored by phenol-sulphuric acid method at 490 nm. 15 Subsequently, the main fraction FSP1 eluted with distilled water was collected for further purification by gel permeation chromatography on a Sepharose CL-6B Fast Flow (2.6 cm × 100 cm) column with 0.1 mol/L NaCl as the mobile phase at a flow rate of 2.0 mL/min, affording only one purified fraction (FSP).

| Chemical composition, monosaccharide analysis, homogeneity and molecular weight determination
The total carbohydrate content of polysaccharide sample was measured by the phenol-H 2 SO 4 method, with glucose as the standard. 15 Uronic acid content was determined according to a m-hydroxydiphenyl colorimetric method by using d-galacturonic acid as the standard. 16 In addition, protein content in the polysaccharide fraction was measured as previously described by Bradford's method 17 using bovine serum albumin (BSA) as the standard.
Monosaccharide composition was determined using gas chromatograph (GC) as described by Liu et al 18 In short, the polysaccharide samples were hydrolysed with 4 mL of 2 mol/L TFA at 115°C for 2.5 hours in a sealed test tube. After excessive TFA was completely removed, the hydrolysates were mixed with 10 mg of NaBH 4 for 3 hours at room temperature and supplemented with each 0.5 mL of anhydride and pyridine for 1 hour at 60°C to be converted into the alditol acetates before GC analysis.
The polysaccharide homogeneity and its molecular weight evaluation were determined by high-performance gel filtration chromatography (HPGPC) on an Agilent1100 HPLC instrument, matched with an Agilent RID-10A refractive index detector and TSK-GEL G4000SW column (7.8 mm × 300 mm). The samples were dissolved in 0.1 mol/L NaNO 3 and filtered (0.45 μm) before injection. The flow rate was 0.5 mL/min at 40°C, with 1.6 mPA. The molecular mass was estimated by reference to a calibration curve made from a set of Dextran T-series standards.

| Bacterial strains and growth conditions
The clinical isolates used were two strains of E cloacae: one is SHV-12 ESBL-negative and another is SHV-12 ESBL-positive strain. Both of them were kept and identified using RT-PCR in our laboratory of Pathogenic Biology, Guangdong Medical University.

| Growth curve assays
To assess in vitro fitness costs under noncompetitive conditions, the growth kinetics of E cloacae strain with or without SHV-12 ESBL were constructed as described previously 19 in microplates coupled to a Multiscan spectrophotometer (Thermo Scientific). Growth curves were made by diluting ~2 × 10 4 bacteria into 200 μL of LB broth in a 96-well microplate, which was kept at 37°C with constant shaking. Absorbance was measured at 600 nm at 60-min interval post-dilution. Each curve was performed three times in the same microplate.

| Biofilm assay
Biofilm formation assay under static conditions was carried out as described previously, with minor modifications. 21 Briefly, two strains were grown for 24 hours at 37°C in TSB medium and overnight cultures were diluted 1:100 in TSB medium supplemented with 0.5% glucose. Then, 200 μL of diluted cell suspension was transferred into flat-bottom 96-well plates and incubated overnight at the room temperature without shaking. Some wells filled with fresh TSB were used as blank controls. At the end, unattached cells were carefully wiped off with gently phosphatebuffered saline (PBS) washing for three times. Subsequently, the plates were air-dried, stained with 200 μL of crystal violet (0.5%) for 15 minutes, and washed to remove the unbound stain. After the plates were air-dried for 2 hours, 200 μL of dimethyl sulfoxide (DMSO) was seeded into the wells to dissolve the attached dying material and the absorbance was measured at 600 nm. The assays were performed for three independent times.

| Sensitivity assay
For determining the susceptibility of SHV-12 ESBL-positive E cloacae strains to FSP, 1 mL of LB at pH 7.4 was inoculated with a single clone and cultured to logarithmic growth phase. After that, 300 μL of the culture was added into pre-warmed fresh LB broth (30 mL) for 24 hours. Thereafter, different growth of bacteria was assessed per hour through determining the optical density of the culture at 600 nm using the Lambda 25 UV-visible spectrophotometer. The culture without FSP was used as control, and levofloxacin (0.5 μg/ mL) was used as antibacterial positive control. All the experiments were determined at least three independent tests.

| Agar bioassay
The effect of FSP on SHV-12 ESBL-positive and SHV-12 ESBLnegative E cloacae strains was investigated by culturing the bacteria on LB agar plates (10 5 CFU of each strain per plate) supplemented with FSP (2.5, 5 and 10 μg/mL). Plate without strains was defined as a control. Thereafter, the plates were subject to incubation at 37°C for 12 hours and the number of colonies was counted.

| Agar well diffusion assay
The antibacterial activity of FSP was also confirmed by agar well diffusion assay, 22 with slight modification. In brief, E cloacae strains (10 5 CFU/mL) at exponential growth phase were poured uniformly into LB agar plates supplemented with different concentrations (2.5, 5 and 10 μg/m) of FSP and then cultured at 37°C for 12 hours.
The same amount of sterile water was taken as the control, and levofloxacin (2.5 μg/disc) was used as antibacterial positive control.

| RNA extraction and quantitative real-time PCR (qRT-PCR) analysis
The expression levels of the SHV-12 ESBL genes were determined by qRT-PCR. For RNA extraction, bacteria were grown aerobically in LB broth until mid-log phase. The culture without FSP was used as control, and levofloxacin (0.5 μg/mL) was used as antibacterial positive control. Next day, total RNA was collected using TRIzolR Reagent as per the manufacturer's instructions, and total RNA concentrations were quantified spectrophotometrically using a UV-visible NanoDrop 2000 spectrophotometer. cDNA was synthesized from 2 μg of total RNA using QIAGEN qPCR Kit and was used for qRT-PCR assay with specific primers (forward primer: 5′-GGT TAT GCG TTA TAT TCGCC-3′and reverse primer: 5′-TTA GCG TTG CCA GTG CTC-3) to examine the SHV-12 ESBL level. The comparative CT method was applied to assess the relative level of mRNA. For relative quantification, the level of SHV-12 ESBL was expressed as -fold changes normalized to the Ct value of GAPDH. All assays were carried out in triplicates to ensure reproducibility.

| Statistical analysis
All results are the means of 3 experiments ± standard error (SE).
Statistical analysis was performed using Student's t test or ANOVA using GraphPad software package for Windows (Prism version 3.00).
A value of <0.05 was considered statistically significant.

| Isolation, purification and characterization of polysaccharide
The polysaccharide CFSP, with a yield of 4.54% of the starting raw material, was isolated from defatted fruit of F suspensa by hot-water extraction, ethanol precipitation, deproteinization and lyophilization. Total CFSP was first fractioned through a DEAE Sepharose Fast

| Growth kinetics, in vitro competition and biofilm formation experiments of E cloacae strain with or without SHV-12 ESBL
To assess whether SHV-12 ESBL contributed a fitness defect to cloacae control was lower than that of E. cloacae strain producing SHV-12-type ESBL, but the difference in biofilm formation was not statistically significant.

| Antibacterial activity of FSP on E cloacae strain with SHV-12 ESBL
The growth curve of E cloacae strain with SHV-12 ESBL in the presence of FSP (2.5, 5.0 and 10.0 μg/mL) was shown in Figure 3A.  Figure 3B). In addition, the average dimension of inhibition zone resulting from FSP application against SHV-12 ESBL-positive E cloacae strain ranged from 8.71 ± 0.95 to 30.41 ± 2.54 mm with the increasing concentration ( Figure 3C). These results revealed that FSP was more effective against the growth of SHV-12 ESBLpositive E cloacae strain.

| Effect of FSP on SHV-12 ESBL gene expression
The transcript levels of SHV-12 ESBL were examined in all FSPtreated E cloacae strains. The qRT-PCR result in Figure 4 showed

| D ISCUSS I ON
Drug resistance with wide spread of antibiotic-resistant strains is a consequence of the worldwide antibiotic abuse, and this has caused acute challenge for human health. 24,25 Notably, with the increase of bacterial resistance to antibiotics, effective and nontoxic antibiotics for infections caused by multidrug-resistant Gram-negative pathogens become decreasing. 26 Furthermore, using antibiotics and other synthetic compounds have mountains of adverse effects to human. 27 Currently, much interest has been focused on searching for novel natural antibacterial agents from TCM that are rich in bioactive substances and well known for their antimicrobial and antibiofilm properties. 28,29 SHV-12 ESBL plays a crucial role in E cloacae resistance to β-lactam antibiotics, which can efficiently hydrolyse antibiotics and render them ineffective. 30  Next, with the aim of searching for a potent antibacterial substitute, we isolated and purified one homogenous polysaccharide F I G U R E 3 In vitro antibacterial analysis of the polysaccharide. A, Bacterial growth curve of SHV-12 ESBL-positive Enterobacter cloacae in LB media in the presence of FSP (2.5, 5.0 and 10.0 μg/mL). B, Agar bioassay for SHV-12 ESBL-positive E cloacae in the presence of FSP (2.5, 5.0 and 10.0 μg/mL). C, Agar well diffusion assay for SHV-12 ESBL-positive E cloacae in the presence of FSP (2.5, 5.0 and 10.0 μg/mL). The mean ± SE for three replicates are illustrated F I G U R E 4 Relative mRNA levels of SHV-12 ESBL in FSP-treated Enterobacter cloacae strains measured by qRT-PCR. The mean ± SE for three replicates are illustrated. * P < .05, ** P < .01 vs control from defatted fruit of F suspense, with an average molecular weight of 9.08 × 10 4 Da. FSP was a kind of glucan, as identified by GC analysis. To further explore whether E cloacae strain with SHV-12 ESBL was sensitive to FSP, antibacterial activity of FSP was evaluated by sensitivity, agar growth and agar diffusion assay.
The control cells grew in a fast rate with increasing DO value at 600 nm, kept stable until end. In contrast, the OD 600 value kept a slower growth rate than the control at each time point after 3 hours and even remained at a low value till the end. This trend was also observed in agar plates with less bacterial colony in the presence of SPF. Furthermore, FSP showed a high inhibitory effect on SHV-12 ESBL-positive E cloacae strain with the bigger diameter of inhibition zone than control. These results apparently indicated that E cloacae strain with SHV-12 ESBL was more susceptible to FSP treatment. Consistent with these observations, qRT-PCR indicated that the inhibitory effect of FSP on SHV-12 ESBL-positive E cloacae strain was negatively related to the SHV-12 ESBL gene expression.

| CON CLUS IONS
In conclusion, our present study provided the first evidence that the acquisition of SHV-12 ESBL promoted the fitness benefit of E cloacae strain and FSP possessed a potent antibacterial activity towards SHV-12 ESBL-positive E cloacae strain, which was achieved by suppressing SHV-12 ESBL gene expression, suggesting that FSP might be a new natural antibacterial alternative in pharmaceutical industries. Further testing of this plant material in animal models will be required to uncover their suitability for clinical use in humans.

ACK N OWLED G EM ENTS
This work was supported by Natural Science Foundation of