Antioxidant and Antibacterial Study of 10 Flavonoids Revealed Rutin as a Potential Anti-Bio lm Agent in Klebsiella Pneumoniae Strains Isolated From Hospitalized Patients

Zhibin Wang The A liated Hospital of Southwest Medical University Zixuan Ding The A liated Hospital of Southwest Medical University Zhaoyinqian Li The A liated Hospital of Southwest Medical University Yinhuan Ding The A liated Hospital of Southwest Medical University Fan Jiang Shandong University Jinbo Liu (  liulab2019@163.com ) The A liated Hospital of Southwest Medical University https://orcid.org/0000-0002-7981-3666


Antioxidant capacity of avonoids
As a stable free radical the DPPH was used to evaluate the e ciency of antioxidant and antiradical of 10 avonoids. The method was adapted to previous study [25,26] and slightly modi ed. The DPPH assay was conducted as follows. Fabrication of standard curve-The 2.5 mg DPPH powder was weighted and dissolved in methanol, the standard stock solution of 2.5 ×10 -2 mg/mL was prepared. Then a series of solutions of 0, 5, 10, 15, 20 and 25 μg/mL were separately made. The absorbance of each solution at 515 nm was measured with a Beckman Model DU650 spectrophotometer. The standard curve was drawn according to the absorbance value. DPPH residual rate detection-The 12, 6, 3 and 1.5 mg/mL solutions were prepared separately of 10 avonoids.
Ascorbic acid was used as control. Reaction tubes of 10 mL were prepared and each was added in 100 μL avonoids solution and 3.9 mL DPPH solution (2.5×10 -2 mg/mL). Methanol was used as empty control. The absorbance was measured at different time points. The DPPH residual rate was calculated out according to the standard curve.

Determination of minimum inhibitory concentration
The K. pneumoniae ATCC700603 and E.coli ATCC25922 strain were cultured in LB agar medium for 24 h at 37℃. The antibacterial ability of 10 avonoids were determined by minimum inhibitory concentration (MIC). The MIC was evaluated through micro-broth dilution method performed in 96-well micro-plates according to the methods previously reported [27]. Brie y, K. pneumoniae ATCC700603 and E.coli ATCC25922 were cultured in LB solid medium and grown overnight at 37℃ without shaking. Then, 100μL bacterial suspension (10^6cfu/mL) was transferred to each well. Then, 100 μL of each dissolved avonoid stock was twofold serially diluted (ranging from 1μg/mL to 1024μg/mL) and added to each well, making the nal volume of 200 μL in each well. The empty control was performed by 3 wells lled with LB only and the positive control was conducted in 3 wells lled with LB and bacteria. Cipro oxacin was used as positive drug control. The 96-well plates were covered and incubated for16 h at 37℃ with shaking (150 rpm). The MIC was determined when no visible bacterial growth can be seen (absence of turbidity) in the well.

Bacterial strains and growth conditions
Bacterial species were identi ed by standard methods with a MicroScan WalkAway 96 Plus System (Siemens, Germany). K. pneumoniae ATCC700603 and E.coli ATCC25922 were purchased from China National Health Inspection Center.
All procedures performed were approved by the Ethical Committee of the A liated Hospital of Southwest Medical University and were in accordance with the tenets of the 1964 Helsinki declaration and its later amendments. All strains were cultured in LB medium at 37℃. Before testing, the microorganisms were transferred on Mueller-Hinton agar (MHA, Lio lchem, Italy) and cultured over night at 37 ˚C. Subsequently, 10 mL of Mueller Hinton broth (MHB, Scharlau, Spain) were inoculated with one representative colony of each organism taken from MHA, then cultured overnight (37˚C, 150 rpm) and used as source of inoculum for each assay.
Growth curve measurement Stationary K. pneumoniae cultures were diluted 1:1000 in LB and incubated at 37°C with shaking (220 rpm). Each well of a 96-well polystyrene microtiter culture plate (Costar #3524, Corning, NY, USA) were lled with 10μL of the overnight culture diluted to a nal concentration of 1.5×10^8 cfu/ml and 190 μL of LB broth mixed with rutin to a nal concentration of 512 μg/mL. Sterile LB was added to the last three wells and used as a negative control. The 96-well plates were covered and incubated for 24 h at 37℃ with shaking (150 rpm). The growth was measured every 2h at OD590 using an microplate reader. All experiments were performed in triplicate.

Bio lm formation assay
Crystal violet (CV) staining was used in this study to quantify the bio lm formation. The assay was performed as previously described [28,29]. The procedures was slightly modi ed. Brie y, for each strain, three wells of the 96-well plate were lled with 180 μL LB broth and 20 μL of the overnight culture. The overnight culture was grown at 37 °C and diluted to a nal optical density of 0.1 at 600nm before use. Wells lled with nothing and sterile LB broth were used as empty control and negative control, respectively. After incubation for 24h at 37°C, each well was washed three times with phosphate-buffered saline (PBS). The 96well plate need to be dried at 60 °C for 30 minutes and stained for 15 minutes with 200 µL 0.1% (w/v) CV for 30min at room temperature. After that, each well was washed three times with distilled water and air-dried. The dye bound of the adherent cells was dissolved in 200 µL of 95% ethanol (Merck, Germany). The 590 nm absorbance was quanti ed using a microplate reader. All assays were performed in triplicate.

Quantitative reverse transcription (qRT)-PCR
The expression levels of 15 genes were determined from bio lm-grown cells using qRT-PCR. Brie y, bacteria were grown in LB broth at 37 °C in a 6-well microtiter plates(LG18-301C-1H, Lige Science, Guangzhou, China) and, after incubation for 24 hours, each well was gently washed and bio lm cells were removed from the well surface using a cell scraper and resuspended in saline solution. Total RNA was extracted using an RNeasy Mini Kit (Qiagen) and cDNA was synthesised using the RevertAid First Strand cDNA Synthesis Kit (Thermo Scienti c) according to the manufacturers' instructions. The qRT-PCR was performed with an Applied Biosystems 7500 RT-PCR system using a SYBR Green RT-PCR Kit (Qiagen) and the primer pairs listed in Table 1.
Primers were synthesized by Shanghai Biotech (Shanghai, China). The relative expression levels of tested genes were normalised to the expression of the 16sRNA gene. Each sample was run in triplicate and the means of Ct values were obtained for analysis. The relative gene expression was represented as fold change between the (-rutin) and (+rutin) treatment. Data were analysed using the 2 -ΔΔCt method.

Statistical analysis
The Student's t-test was used to analyse the difference in bio lm formation ability and gene expression levels. Statistical analysis was performed with GraphPad Prism 7.01 software and IBM SPSS Statistics 23 software.

Correlation analysis
The correlation analysis was conducted through Pearson's correlation analysis. Three separated biological replicates of each assay were used for analysis.

Antioxidant capacity of avonoids
To identify the correlation between the antioxidant capacity and the related molecular structures of active avonoids, and to develop new drugs that are less likely to induce resistance to treat drug-resistant bacteria, 10 components representing 4 major avonoids categories were screened and evaluated through DPPH assay in present study ( Table 2). The results showed that among the 4 major classes of avonids, avonols showed the highest free radical scavenging capacity, followed by avones and avanones. The polymethoxylated avones were supposed to have the lowest antioxidant activity ( Figure 1). Furthermore, within the avonols, rutin exhibited highest free radical scavenging ability, followed by kaempferol and quercetin, respectively.

Clinical characteristics and bio lm formation abilities of K. pneumoniae strains
The bio lm formation abilities of 177 clinical isolates of K. pneumoniae were evaluated by crystal violet staining. The results were shown in Table 3. Among the 177 isolates, the positive rate of bio lm producer was 88.14% (156/177), of which strong (+++), moderate (++), and weak (+) bio lm producing strains account for 10.73% (19/177), 16.95% (30/177) and 60.45% (107/177), respectively. The rate of strong bio lm producer was low among these clinically isolated K. pneumoniae strains, though the positive rate of bio lm producer was high.

Minimum inhibitory concentration of 10 avonoids
The minimum inhibitory concentration, termed as MIC of avonoids against K. pneumoniae ATCC700603 and E.coli ATCC25922 were evaluated. Flavonols showed higher level of inhibitory effect than other categories of avonoids, maybe due to their strong antioxidant capacity. However, most of the avonoids are not strong inhibitory agents against neither K. pneumoniae, nor E.coli, as the lowest MIC was 512 µg/mL of rutin against E.coli ATCC25922 and 1024 µg/mL of rutin against K. pneumoniae ATCC700603. Other avonids were found no obvious inhibitory effects even when the concentration of these avonoids increased to 1024 µg/mL (Table 4).
Inhibitory effect of rutin on K. pneumoniae growth From the results above, we found rutin was the most potential antimicrobial agent. In order to con rm the inhibitory effect of rutin against K. pneumoniae strains, the growth of two drug-resistant K. pneumoniae isolates (Kpn495 and Kpn499 ) were evaluated by morphological observation and growth curve changes. Furthermore, naringin and the combination of rutin and naringin were also used to evaluate their inhibitory effect against K. pneumoniae ( Figure 2). The concentration of both rutin and naringin used in this session was increased to 50 mg/mL as no obvious inhibitory effect was found at low concentration.
In solid media, both rutin and naringin showed inhibitory effect on K. pneumoniae growth. Rutin showed better inhibitory effect than naringin both in Kpn495 and Kpn499 strains. In addition, rutin showed signi cant inhibitory effect on the 24-h growth curve of both Kpn495 and Kpn499 (Figure 2 c and d).
The combination of these two components were found no signi cant inhibitory effects. In contrast, the combination of rutin and naringin even promoted the growth of Kpn495 strain.
To identify the correlation between the concentration of rutin and its anti-bio lm capacity, ve different concentrations were used to treat the bio lm and the inhibitory effects were compared (Figure 3c and d). Rutin signi cantly decreased the bio lm production when the concentration was at 512 μg/mL and 256 μg/mL. However, when the concentration lowered to 128 μg/mL, the inhibitory effect was not signi cant. When the concentration decreased to 64 μg/mL or 32 μg/mL, no inhibition was found.

Gene expression analysis
To identify key factors that involved in bio lm formation process, totally 15 genes related to type I and III mbriae biosynthesis, virulence factors, lipopolysaccharide (LPS) biosynthesis, poly-β-1,6-N-acetyl-d-glucosamine polysaccharide secretion, quorum-sensing(QS), capsular polysaccharide(CPS), colonic acid (wcaJ), allantoin, capsule-associated virulence gene and siderophore-associated gene were analyzed in bio lm cells before and after the rutin treatment, respectively. In Kpn140, 7 genes (mrkA, wzm, wbbM, luxS, treC, wabG, entB) were detected to show signi cant expression differences with and without rutin treatment. Among these genes, wzm, wbbM and entB were up-regulated while, mrkA, luxS, treC and wabG were down-regulated after rutin treatment in bio lm cells.The expression level of luxS, wabG and mrkA gene were down regulated 69-fold, 61-fold and 8.8-fold after rutin treatment, respectively. However, the up-regulated genes were found to have very low expression level though the changes were signi cant after rutin treatment (Figure 4a). In

Correlation between bio lm and gene expression
To identify the function of these genes in bio lm cells, in particular their response to rutin treatment and bio lm formation, the correlation between gene expression and bio lm biomass was studied (table 5). Of these genes, 7 of them were highly correlated with bio lm biomass. The expression of mrkA gene was positively correlated with the bio lm production(R=0.694). However, the results showed the genes, wzm, wbbM, magA, wcaJ, allS and entB were negatively correlated with bio lm biomass, as their coe cients were -0.846, -0.815, -0.903, -0.823,-0.669 and -0.737, respectively. Though the luxS gene showed highest level of expression, the correlation between its gene expression and the bio lm biomass was not signi cant. Furthermore, co-expression of genes were found in this study.

Discussion
As an opportunistic Gram-negative pathogen, K. pneumoniae can cause a wide range of severe infections like pneumonia, sepsis and urinary tract infections [37]. Furthermore, this strain is associated with invasive disease in community, while in hospital, it has a marked propensity to acquire drug resistance [38]. From previous studies we observed, associated with multi-drug resistance, K. pneumoniae is resistant to all, or almost all, available antibiotics [39]. As a result, it is urgent to develop new antibacterial agents or novel strategies to ght the strain.
Plant-derived avonoids have potential properties in the discovery of new antibacterial agents [40]. In some cases, it was showed that avonoids exhibited up to 6-fold stronger antibacterial activities than standard drugs in the market [41]. However, in this study, we evaluated 10 avonoids and found the MICs of most tested avonoids were higher than 1024 µg/mL, indicating that the antibacterial ability of avonoids was not strong especially when the avonoids were in low concentration. The lowest MIC values were observed in rutin of 1024 µg/mL and 512 µg/mL against K. pneumoniae ATCC700603 and E.coli ATCC25922, respectively. This result showed the inhibitory effect varied among different bacteria species. Regarding to bio lm, no MBIC was observed but the inhibitory effect of rutin against the bio lm formation was proved when the concentration of rutin was at 256 µg/mL or higher than that.
Flavonoids are secondary metabolites with C6-C3-C6 skeleton and widely distributed in plant kingdom [42], including more than 8,000 compounds [43]. As a result, it is not possible to evaluate them one by one. For our rst step, in present study, 10 avonoids representing 4 major avonoid categories were screened and studied.
Previous study showed free hydroxyl can donate hydrogen and electron and hence responsible for the radical scavenging ability [44]. In this study, we found the free radical scavenging ability was also correlated with the number of hydroxyl. Flavonols contain more (at least 4 hydroxyl) hydroxyl and they showed stronger free radical scavenging ability than others. In addition, luteolin showed higher free radical scavenging ability in comparison with apigenin, may be due to its 4 hydroxyl, though both of them are in avones group.
Flavonoids were found to perform antibacterial actions through a series of mechanisms, such as bacterial toxin production, inhibition of nucleic acid synthesis and bio lm formation [45]. Bio lm formation ability is increasingly considered when develop new antibacterial drugs as bacteria inside bio lms can increase 10-1000 fold resistance [46]. Compared with conventional antibiotics, avonoids are rich in nature and not easily to generate drug-resistance.
However, the inhibitory effects of avonoids on K. pneumoniae are still unclear. In this study, 15 genes related to bio lm formation, involving in various biological processes were studied with and without rutin treatment. The mrkA gene was revealed to tightly correlated with the bio lm formation. This result indicated that the mrkA gene may play vital role in bio lm formation.
The luxS gene is a key factor involved in quorum sensing (QS) system in bacteria [47]. QS is a universal mechanism of cell density and interspecies communication through which the bacteria can regulate various gene and biological process, including bio lm formation [48]. In both Kpn140 and Kpn173, the luxS gene's expression was high in bio lm cells without rutin. However, luxS gene's expression decreased sharply when the bio lm was inhibited by rutin. Furthermore, we can conclude the importance of luxS, as several bio lm related genes were found to be co-expressed with luxS, though the correlation between luxS gene expression and bio lm formation was not signi cant. These results shed light on the inhibitory mechanism of rutin on the bio lm formation in K. pneumoniae strains, which could aid in developing new strategies for infection control.

Conclusion
Among these avonoids, rutin showed stronger free radical scavenging capacity and antibacterial activity. Of the bio lm related genes, the expression of luxS gene was signi cantly induced by rutin treatment in bio lm cells. The bio lm biomass was positively correlated with the expression of a type III mbriae biosynthesis gene, mrkA. Rutin could be a potential agent to inhibit the K. pneumoniae bio lm formation.