Adhesion and biofilm inhibiting properties of Chinese Herbal Formula SanHuang decoction against antibiotic resistant staphylococcal strains


 Clinical study has shown that external socking and washing with the Chinese herbal SanHuang decoction (SH) can control the orthopedic-biofilm related infections. However, the antibiofilm activities of SH in vitro have not been investigated. The aim of the current study was to explore the effect of SH on adhesion and biofilm formation of antibiotic-resistant staphylococci on titanium surface, and to explore its probable mechanistic effects on staphylococcal strains. Biofilm-forming ATCC 35984 (methicillin-resistant Staphylococcus epidermidis, MRSE) and ATCC 43300 (methicillin-resistant Staphylococcus aureus, MRSA) strains were used in this study. The minimum inhibitory concentrations (MICs) of SH and vancomycin against planktonic bacterial strains were determined by the broth microdilution method. Different sub-MIC of SH with TSB (Tryptic soy broth) were used as the basis for experimental grouping (SH group). TSB culture medium alone (TSB group) or TSB containing vancomycin (vancomycin group) incubated with bacteria were considered as the negative or positive control group, respectively. The inhibitory effect of different treatment on bacterial adhesion and biofilm formation were observed by the spread plate method, CV (crystal violet) staining, scanning electron microscopy (SEM), and confocal laser scanning microscopy (CLSM). Real time PCR analysis was performed to determine the effect of SH on the expression levels of ica AD and ica R gene in ATCC 35984 during the biofilm formation. The strains were found to be susceptible to SH decoction with MIC and MBC values of 38.75 mg/ml and 77.5 mg/ml, respectively. The MIC values for vancomycin was 2.5 μg/ml. SH treatment with 1 MIC and 1/2 MIC could inhibit the bacteria adhesion on the titanium surface, showing only scattered bacterial adhesion from SEM. CLSM showed that SH with 1 MIC and 1/2 MIC could also inhibit the bacterial biofilm formation. The quantitative results of the spread plate method and CV staining showed that there was significant differences between the SH groups (P < 0.05). Further, with an increase in SH concentration, the inhibitory effect became more obvious at different culture time points, when compared with TSB control group (P < 0.05). Among the groups, vancomycin had the strongest inhibitory effect on bacterial adhesion and biofilm formation (P < 0.01). Meanwhile, with an increase in SH concentration, the expression levels of ica A and ica D decreased, and the expression of ica R increased correspondingly (P < 0.05). In conclusions, a certain concentration of SH can inhibit the adhesion and biofilm formation of antibiotic-resistant Staphylococcal strains on the titanium surface, but the effect was not as good as vancomycin. Its probable mechanistic activity may be through the inhibition of polysaccharide intercellular adhesin synthesis by down-regulating the expression of ica AD gene, thus inhibiting bacterial biofilm formation.


Introduction
Clinical study has shown that external socking and washing with the Chinese herbal SanHuang decoction (SH) can control the orthopedic-bio lm related infections. However, the antibio lm activities of SH in vitro have not been investigated. The aim of the current study was to explore the effect of SH on adhesion and bio lm formation of antibiotic-resistant staphylococci on titanium surface, and to explore its probable mechanistic effects on staphylococcal strains. Bio lm-forming ATCC 35984 (methicillinresistant Staphylococcus epidermidis, MRSE) and ATCC 43300 (methicillin-resistant Staphylococcus aureus, MRSA) strains were used in this study. The minimum inhibitory concentrations (MICs) of SH and vancomycin against planktonic bacterial strains were determined by the broth microdilution method.
Different sub-MIC of SH with TSB (Tryptic soy broth) were used as the basis for experimental grouping (SH group). TSB culture medium alone (TSB group) or TSB containing vancomycin (vancomycin group) incubated with bacteria were considered as the negative or positive control group, respectively. The inhibitory effect of different treatment on bacterial adhesion and bio lm formation were observed by the spread plate method, CV (crystal violet) staining, scanning electron microscopy (SEM), and confocal laser scanning microscopy (CLSM). Real time PCR analysis was performed to determine the effect of SH on the expression levels of ica AD and ica R gene in ATCC 35984 during the bio lm formation. The strains were found to be susceptible to SH decoction with MIC and MBC values of 38.75 mg/ml and 77.5 mg/ml, respectively. The MIC values for vancomycin was 2.5 μg/ml. SH treatment with 1 MIC and 1/2 MIC could inhibit the bacteria adhesion on the titanium surface, showing only scattered bacterial adhesion from SEM. CLSM showed that SH with 1 MIC and 1/2 MIC could also inhibit the bacterial bio lm formation.
The quantitative results of the spread plate method and CV staining showed that there was signi cant differences between the SH groups (P < 0.05). Further, with an increase in SH concentration, the inhibitory effect became more obvious at different culture time points, when compared with TSB control group (P < 0.05). Among the groups, vancomycin had the strongest inhibitory effect on bacterial adhesion and bio lm formation (P < 0.01). Meanwhile, with an increase in SH concentration, the expression levels of ica A and ica D decreased, and the expression of ica R increased correspondingly (P < 0.05). In conclusions, a certain concentration of SH can inhibit the adhesion and bio lm formation of antibiotic-resistant Staphylococcal strains on the titanium surface, but the effect was not as good as vancomycin. Its probable mechanistic activity may be through the inhibition of polysaccharide intercellular adhesin synthesis by down-regulating the expression of ica AD gene, thus inhibiting bacterial bio lm formation.
Postoperative infection is one of the serious complications impacting the clinical results of orthopedic surgery, and its inadequate management will lead to improper union of fracture, loss of limb function and even amputation, which will inevitably impose an economic burden to the patients' families and social medical insurance 1 . According to the literature, the incidence of postoperative infections for closed fracture is about 1-2 %, but the incidence of an open fracture is signi cantly as high as 30 % 1,2 . Studies have shown that 80% of clinical infections are related to bacterial bio lm formation (BBF), which plays an important role in the pathogenesis of refractory orthopedic infections 3 . Firstly, bacteria in a bio lm are very resistant to antibiotics, and their antibiotic resistance concentration is 1,000-fold higher than the MIC of the planktonic form, which has further attributed to the metabolic dormancy, altered microenvironments within the bio lm, and active resistance mechanisms 3,4 . Secondly, the bio lm and its colonized pathogens can successfully defend the host's immune system, so that they cannot be phagocytized and cleared by polymorphonuclear leukocytes 5 . Therefore, inhibition of BBF is key to the successful treatment of orthopedic infections.
For refractory orthopedic infections related to BBF, in addition to thorough local debridement, another important option is a long-term application of systemic intravenous antibiotics 6 . In clinical practice, however, BBF-related infections are often caused by resistant Staphylococcus, especially MRSE and MRSA, which makes the treatment of implant-associated infections with antibiotics still one of the key challenges for orthopedic surgeons 2,4,6 . Hence, discovering alternative treatment methods for orthopedic infections linked to multi-antibiotic-resistant bacteria forming BBF is becoming more and more relevant.
For a long time, TCM (traditional Chinese medicine) has been widely used in the treatment of various infectious diseases in China, and many traditional or empirical prescriptions have been proved to have satisfactory clinical therapeutic effects [7][8][9][10] . The SanHuang decoction (abbreviated as SH) is one of the treatment methods widely used in our hospital for treatment of limb infections with bone and implant exposure. This decoction is composed of Scutellaria baicalensis Georgi, Coptidis rhizoma, Cortex Phellodendri chinsis and other Chinese herbs. According to the theory of traditional Chinese medicine, it has the heat-clearing and detoxing functions. Through external soaking and washing of the locally infected wound with SH, it was found that this decoction can effectively control the infections [8][9][10] .
However, there has been no study to date that has explored the antibio lm formation properties of the SH in vitro. Therefore, based on previously published clinical studies [8][9][10] , this study used MRSE and MRSA which are widely seen in orthopedic implant-associated infections, and investigated the effect of SH decoction on adhesion and bio lm formation of bacterial strains on titanium surfaces. Further, we also explored the probable mechanism behind its effect on bio lm formation.

Materials And Methods
Preparation of SH decoction. SH contains Scutellaria baicalensis Georgi (25 g), Coptidis rhizoma (25 g), Cortex Phellodendri chinsis (25 g), Sophora avescens (20 g), Lonicera Japonica (20 g), Forsythia suspensa (20 g) and Taraxacum mongolicum Hand.-Mazz. (20 g) ( Table 1). All the herbs were provided by the TCM pharmacy of Henan Luoyang Orthopedic-Traumatological Hospital following the requirement standards of the section one of Pharmacopoeia of the People's Republic of China (2015 edition). The procedure of preparing the SH aqueous extract is as follows: rstly, all herbs were tipped into a ceramic pot with 1000 mL pure water and soaked for 30 min. Secondly, the concoction was boiled at high heat for 30 min, and was maintained at a low ame until the aqueous extract reduced to 250 ml. Finally, aqueous extract was ltered with lter paper, and water extraction of the liquid was performed through Ø 22 μm lter. Based on the original herb dose, the initial liquid concentration of the extract solution was about 0.62 g/ml, which was stored in the refrigerator at -20 °C for the subsequent experiments.
Preparation of bacterial strains. In this study, the positive bio lm-forming strain ATCC 35984 (MRSE) and ATCC 43300 (MRSA) were used to evaluate the bacterial bio lm formation on titanium surface 11 . The strains frozen in a -80°C refrigerator were rapidly thawed and inoculated on a tryptone soy agar (TSA) plate and cultured in an incubator at 37 °C for 24 h. A single colony was selected from the TSA and inoculated in a 50 ml centrifuge tube containing 10 ml TSB, the tube was incubated for 10 h at 37 °C with agitation at 100 rpm. Then, 25 μl of bacterial suspension was transferred to another sterile centrifuge tube containing 10 ml TSB and incubated at 37 °C for 14 h, after which 1 ml suspension was transferred into a 1.5 ml centrifuge tube and the bacteria in the suspension was harvested by centrifugation for 5 min at 4°C and 1000 rpm (Sorvall TC6 centrifuge, Du Pont, Bad Nauheim, Germany). The precipitate was then washed three times with 0.15 M PBS (phosphate buffered saline) to remove the remaining TSB, and resuspended in sterile PBS to an optical density of 0.490 at 600 nm using Synergy HT multi-detection microplate spectrophotometer (Bio-tek, Winooski, VT), which corresponded to 10 9 CFUs (Colony forming units)/ml. Finally, the bacteria concentration was adjusted to 1×10 6 CFUs/ml by TSB, TSB-SH or TSB-Vancomycin.
Minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) determination. The broth microdilution method was used to determine the MIC and MBC of SH and vancomycin as recommended by the National Committee for Clinical Laboratory Standards Institute 12 . An overnight culture of the strains were diluted 10-fold in TSB and incubated at 37°C until they reached xponential growth phase. Serial two-fold dilutions of SH (155 mg/ml to 4.84 mg/ml) or vancomycin (8 μg/ml to 0.0625 μg/ml) in TSB were prepared in a 96-wells plate with 190 μl per well. Ten microlitres of bacterial inocula of each strain containing 5×10 6 CFUs/ml was added in all wells. A number of wells with only TSB (without inoculum and drugs) in each plate were designed as blank control, and a number of wells with TSB and inoculum (without SH or vancomycin) were designed as negative control. The plates were incubated for 24 h at 37 °C. The MIC was detected following addition of 50 μl of INT (2-4-Iodophenyl-3-4nitrophenyl-5-phenyl-2H-tetrazolium chloride) at a nal concentration 0.2 mg/ml in all the wells and incubated for 30 min at 37 °C. Bacterial growth was determined by observing the color change of INT in the wells. Biologically active bacterial cells will reduce the colourless tetrazolium salt which act as an electron acceptor to a red-coloured formazan product 13 . Inhibition of bacterial growth is observed when the solution in the well remained clear after incubation with INT. MIC was de ned as the lowest extract concentration that completely inhibits the growth of microorganisms. To determine MBC values, 10 μl of culture medium from the microtitre plate wells that showed no changes in color will be re-inoculated on agar plates. After 24 h incubation at 37°C, MBC was determined as the lowest concentration that yielded nil bacterial growth on agar plates. Each experiment was performed three times in duplicate.
Experimental grouping. Different sub-MICs of SH decoction were used as grouping basis. Only TSB incubated with bacteria were used as the negative control, and TSB containing vancomycin (2.5 μg/ml) and incubated with bacteria as the positive control. In total, our study had six experimental groups, namely the TSB group (bacteria cultured with only TSB), Vancomycin group (bacteria cultured with TSB containing vancomycin), 1/8MIC SH group (1MIC SH was diluted by 1/8 using TSB), 1/4MIC SH group, 1/2MIC SH group and 1MIC SH group.
Effect of different SH concentration on bacterial adhesion. Titanium discs of 10 mm in diameter and 1 mm in thickness were placed into a 48-well plate (Costar 3548, Corning, NY, USA), and six technical replicates were used for each group. ATCC 35984 and ATCC 43300 were diluted to a density of 10 6 CFUs/ml with TSB, TSB-vancomycin, or different sub-MICs of SH, and 1 ml of the cell suspension was added to the well containing the discs. The plates were incubated in aerobic conditions at a low shaking rate of 100 rpm for 6 h. Then, the bacterial adhesion on the titanium surface were analyzed with the following methods.
The spread plate method: The discs were gently washed with PBS and transferred into a 10 ml glass tube containing 0.5 ml TSB. The tubes were then placed in an ultrasonic bath (B3500S-MT, Branson Ultrasonics Co., Shanghai, China) and the bacteria attached on the disc were dislodged by ultrasonication (5 min), then followed by rapid vortex mixing (Vortex Genie 2, Scienti c Industries, Bohemia, NY, USA) at maximum power for 1 min. The vortexed solutions were serially diluted 10-fold and the nal three dilutions were plated in triplicate onto TSA and then incubated at 37 °C for 24 h. Finally, the number of CFUs on the TSA were counted, and the number of bacteria in the bio lm was calculated and is expressed relative to the surface area of the sample (CFUs/mm 2 ).
SEM assay: A sterile pipette was used to carefully remove the TSB from each well. The discs were transferred with forceps into another fresh 48-well plate and gently washed with PBS three times to remove loosely adherent bacteria. The discs were xed in 2.5% glutaraldehyde for 2 h at 4 °C, then washed with PBS and dehydrated through a series of graded ethanol solutions (25, 50, 75, 95 and 100%). The samples were subsequently freeze-dried, sputter coated with gold, and observed using SEM (Joel JSM-6310LV, JEOL Ltd., Tokyo, Japan).
Effect of different SH concentration on bio lm formation. Titanium discs of 10 mm in diameter and 1 mm in thickness were placed into a 48-well plate, and six replicates were used for each group. ATCC 35984 and ATCC 43300 were diluted to a density of 10 6 CFUs /ml with TSB, TSB-vancomycin, or different sub-MICs of SH, and 1 ml of the cell suspension was added to each well. The plates were then incubated in aerobic conditions at low shaking rate of 100 rpm for 24 h. Bio lm formation on the titanium surface were analyzed by the following methods.
CV staining assay: A sterile pipette was used to carefully remove the TSB from each well. The discs were transferred into another 48-well plate and gently washed three times with PBS. Then, 1 ml of 2.5% glutaraldehyde solution was added into each well for 10 min to allow xation. The glutaraldehyde solution was removed and the wells were washed with PBS. Subsequently, PBS was removed and 2 ml of 0.1% (w/v) aqueous CV solution was added to each well followed by incubation for 20 min at room temperature. CV solution was discarded, and the wells were washed with PBS again and air-dried for 12 h in the dark. Next, the quantity of bio lm was analyzed by adding 30% acetic acid in a volume of 1 ml to each well to dissolve dye for 30 min, after which 200 μL of the dye solution were transferred into a 96-well microtiter plate, one group equaling one vertical row of the plate. The optical density at 492 nm was read using a plate reader (Synergy HT, BioTEK).
CLSM assay: After 24 h incubation, the TSB medium were removed carefully from each well, and the discs were transferred into another 48-well plate and gently washed with PBS three times. Then the wells containing the titanium discs were added with 500 μl combination dye (LIVE/DEAD Baclight bacteria viability kits, L13152; Molecular Probes, Eugene, OR, USA) in a dark environment at room temperature for Germany). The live/dead kit containing two kinds of uorescence dyes were used to distinguish the viable and non-viable cells under the uorescence microscope, because dye SYTO 9 can make the viable bacteria with intact cell membranes display green uorescence, whereas dye PI (propidium iodide) can make non-viable bacteria with damaged membranes display red uorescence. The Leica confocal software was used to analyze the bio lm images. The images were acquired from random positions on the disc surface.
Effect of different incubation times on bio lm formation. The bacterial bio lm formation on the titanium surface was also observed at different incubation time-points. Titanium discs (10 mm in diameter and 1 mm in thickness) were placed into 48-well plates, 1 ml of the cell suspension with 10 6 CFUs/ml adjusted by TSB, TSB-vancomycin, or different sub-MICs of SH were added to each well and incubated at 37 °C for 6, 12, 18, 24, and 48 h. At the speci ed time-point, the quantitative analysis of bacterial bio lm formation on the titanium surface were observed by CV staining assay, as described above.
Real-time PCR analysis of icaAD and icaR transcription. The experiment had ve groups, including the TSB, 1/8MIC SH, 1/4MIC SH, 1/2MIC SH and 1MIC SH group. ATCC 35984 was selected for the icaAD and icaR expression assays. Titanium discs (14 mm in diameter and 1 mm in thickness) were placed into 24-well plates, and 2 ml of the cell suspension with 10 6 CFUs/ml adjusted by TSB only or SH at different sub-MIC were added to each well and incubated at 37 °C for 24 h. Then, the bacterial bio lms on the disc surface were harvested together with the planktonic bacteria by sonication and transferred into the RNAprotect bacterial reagent (Qiagen, Germantown, MD, USA) to ensure RNA integrity. Bacteria were pelleted by centrifugation at 4 °C and 8000 g for 10 min. We further removed the protective reagent and resuspended the pellets in 0.8 mL TE buffer (10 mM TrisHCl, 1 mM EDTA, pH 7.0) containing 100 g/ml lysostaphin (Sigma, St Louis, MO, USA). Total RNA was isolated using Axyprep Multisource Total RNA Miniprep Kit (Axygen bioscience, Union City, NJ, USA) according to the manufacturer's instructions. RNA was reversely transcribed to cDNA using the PrimeScript® RT reagent kit (TaKaRa, Shiga, Japan). Realtime PCR was performed on an ABI 7500 Fast machine according to the SYBR ® Premix ExTM PCR kit (TaKaRa) kit. The reactions were performed using cDNA templates, and with speci c forward and reverse primers. The ampli cation conditions were as follows: 50 °C for 20 s; 40 cycles of 95 °C for 15 s, and 60°C for 1 min; then 95 °C for 10 min. The expression levels of icaAD and icaR were evaluated and normalized to the internal standard gene 16S rRNA. The quanti cation of gene expression was based on the CT value of each sample, which was calculated as the average of three replicate measurements for each sample analyzed. The primers used for the RT-PCR are shown in Table 2 14 .
Statistical Analysis. All the experiments were performed in triplicate. Data are expressed as mean ± standard deviation. The results were tested using a one-way analysis of variance with Tukey's multiple comparison tests. The differences observed between groups were considered to be signi cant at P < 0.05 using SPSS 19.0 software. (Table 3). Among all the pathogen, ATCC 35984 and ATCC 43300 were found to be susceptible to SH decoction with MIC and MBC values of 38.75 mg/ml and 77.5 mg/ml, respectively. Vancomycin was found to possess high inhibitory activity against bacteria under investigation, with MIC values of 2.5 μg/ml. Effect of SH decoction on bacterial adhesion. From the results of the spread plate method (Fig. 1), 1/2MIC SH, 1MIC SH and TSB-vancomycin groups could signi cantly inhibit the bacterial adhesion on the titanium surface (P < 0.01), compared with TSB, 1/8MIC SH and 1/4MIC SH groups. While among TSB, 1/8MIC SH and 1/4MIC SH group of ATCC 35984, no signi cant difference were observed (P > 0.05). For ATCC 43300, compared with TSB group, both 1/4MIC SH and 1/8MIC SH groups also inhibit the bacterial adhesion on the titanium surface (P < 0.05); while no signi cant difference between 1/4MIC SH and 1/8MIC SH groups were found (P > 0.05). We also observed statistically signi cant difference in the quantitative analysis of bacterial adhesion between SH groups under 1/4, 1/2 and 1MIC concentrations, which was more obvious with the increase of concentration of SH decoction (P < 0.01). On the other hand, vancomycin at MIC had the strongest inhibitory effect on the bacterial adhesion on the titanium surface, compared with other groups (P < 0.01).

MIC and MBC of SH decoction and vancomycin. SH was assessed on growth inhibitory ability against planktonic cells of standard strains. The potency was quantitatively assessed by the MIC and MBC values
Further, we performed SEM (Fig. 2) 6 h after incubation with bacteria and showed that scattered bacterial colonies were observed on the titanium surface in the 1/2MIC SH, 1MIC SH and vancomycin group. However, the bacteria in the TSB, 1/8MIC SH and 1/4MIC SH group obviously clumped together. For SH groups with 1/4, 1/2 and 1MIC concentrations, the bacterial adhesion on the surface of titanium disc became more obvious with the increase of dilution of SH decoction. Meanwhile, the number of adherent bacterial colony on the titanium surface in vancomycin group was the least in all groups. SEM qualitative analysis is consistent with the conclusion of the quantitative data in Fig. 1.
Antibio lm activity to bacterial strains of SH decoction. As shown in Fig. 3, compared with the TSB and 1/8MIC SH groups, the OD values of 1/4MIC SH, 1/2MIC SH, 1MIC SH and vancomycin groups were signi cantly decreased (P < 0.01), while there was no signi cant difference between TSB and 1/8MIC SH groups (P > 0.05). For the SH groups in the two strains, the OD values decreased with the increase of SH concentration, and the difference among the 1/8MIC, 1/4MIC, 1/2MIC and 1MIC SH groups was statistically signi cant (P < 0.01). For 1/4, 1/2 and 1 MIC of SH solution, the OD value was signi cantly decreased, among which the OD values of the 1MIC SH groups were lower than 0.120, indicating no bio lm formation on the titanium surface. Though the OD value of 1/2MIC SH group was signi cantly decreased; it was between 0.120 and 0.240, suggesting that there was still mild bio lm formation 15 . For 1/4MIC SH group, the OD value was more than 0.240, indicating the bio lm formation 15 . The OD value of vancomycin group was lower than that of 1MIC group (P < 0.01), which also indicated that there was no bio lm formation.
CLSM of three-dimensional structure of bacterial bio lms on the titanium surface is shown in Fig. 4. After culture for 24 h, a large number of bacterial colonies were clustered to form a dense bio lm structure on the titanium surface in the TSB and 1/8MIC SH groups, displaying high-intensity green uorescence. However, with the increase of the SH concentration, the intensity of green uorescence gradually decreased, suggesting that the bacterial bio lm formation was also gradually inhibited. For ATCC 35984, high-intensity green uorescence was still visible in 1/4MIC SH group, indicating the existence of severe bio lm formation; While in 1/4MIC SH group of ATCC 43300, in 1/2MIC SH groups of two bacterial strains, low-intensity green uorescence was shown, indicating that there was still mild bio lm formation on the titanium surface. In the 1MIC SH and vancomycin groups, only scattered green uorescence of the bacterial colonies were seen, showing no bio lm formation.
Effect of SH decoction on bio lm formation at different incubation time-point. CV staining (Fig. 5) showed that the OD value of bio lm had an increasing trend in TSB and 1/8MIC SH groups with the increase of incubation time, suggesting the strains adhered to the titanium surface and gradually formed bio lm; from 6 h to 24 h, we observed signi cant differences of the OD values between any two timepoints (P < 0.01), while no statistical difference of OD value was found for ATCC 35984 (P > 0.05) and statistical difference of OD value was found for ATCC 43300 (P < 0.01) between 24 h and 48 h. For the 1/4MIC SH group of ATCC 35984, the OD values increased and difference between any two time-point were found from 6 h to 18 h (P < 0.01), then no difference from 18 h to 48 h (P > 0.05); for ATCC 43300, the OD values of 1/4MIC SH group increased and difference between any two time-point were statistically signi cant from 6 h to 48 h (P < 0.05). On the contrary, the OD values from 6 h to 48h in the 1/2MIC SH, 1MIC SH and vancomycin groups were stable, and no obvious difference was found between any two time-points for 1MIC SH and vancomycin groups (P > 0.05). At different time point, the OD values in the TSB, 1/8MIC SH, 1/4MIC SH groups were higher than that in 1/2MIC SH, 1MIC SH and vancomycin groups (P < 0.01), and signi cant differences were also found among the 1/8MIC, 1/4MIC, 1/2MIC, 1MIC SH and vancomycin groups (P < 0.05).
SH effects on the expression of gene icaAD and icaR. As shown in Fig. 6, the expression level of icaAD was not signi cantly different between the TSB and 1/8MIC SH groups (P > 0.05). The expression level of icaA and ica D in the 1/4 MIC, 1/2MIC, and 1MIC SH groups decreased compared with that of the TSB and 1/8MIC groups (P < 0.01). In the SH groups, the icaA expression level showed a gradual decrease with the increase of SH concentration, showing statistically signi cant differences (P < 0.01). Further, no differences of ica D were found among the 1/4MIC, 1/2MIC, and 1MIC SH groups (P >0.05). For icaR expression level, there were no statistically signi cant differences between the TSB and 1/8MIC SH groups or the 1/2MIC and 1MIC SH groups (P > 0.05). Compared with the TSB and 1/8MIC SH groups, the 1/4MIC, 1/2MIC and 1MIC SH groups showed high expression level (P < 0.01). The expression levels of icaR were also different among the 1/8MIC, 1/4MIC and 1/2MIC (1MIC) SH groups (P < 0.01).

Page 10/21
The clinical treatment of postoperative implant-related infections still remains as a challenge for orthopedic surgeons, which is mainly related to the formation of bacterial bio lms at the infected site 3 . Bio lms are communities of bacteria that are attached to a surface and surrounded by a self-produced matrix, consisting of polysaccharide as well as proteins, nucleic acids, lipids, and humic substances 3 . Due to the formation of bio lms, intravenous antibiotics often fail to completely eliminate pathogenic bacteria of orthopedic implant-related infections. Even if the local symptoms are eliminated in a short time, the infection will recur again when the patient's immunity is reduced 16 . In addition, bio lms also increase bacterial resistance to antibiotics 17 . Therefore, the inhibition of bio lm formation and the killing of antibiotic-resistant bacteria are the key methods for effective treatment of the refractory infections related to the orthopedic implants. For these reasons, in addition to thorough debridement, intravenous use of sensitive antibiotics is necessary for the clinical treatment of infections. However, unfortunately, multidrug-resistant bacteria are often found in the bacterial culture of clinical orthopedic infection wounds, and based on the antibiotic sensitivity tests, vancomycin has often become the rst choice for clinicians 18 . Studies have shown that only a small portion of intravenous vancomycin can reach bone and joint tissues, and even less penetrates into bone tissues at the site of osteomyelitis; thus, vancomycin may not eliminate multidrug-resistant pathogens in bio lms 19 . In addition, the application of vancomycin has led to the emergence of vancomycin resistant Staphylococcus strains 20 . Linezolid is another novel antibiotic that has been found to be useful in the treatment of vancomycin failed MRSA infections, but safety concerns have limited its widespread use 20 .
Due to the aforementioned defects in antibiotics for the treatment of bio lm-related infections caused by multidrug-resistant pathogens, seeking other treatment methods for refractory orthopedic bio lm-related infections is urgent and necessary. One such treatment strategy is the external application of the TCM, which has been used in the chinese population to prevent and treat orthopedic infectious for long time, and many prescriptions have been proved to be therapeutically effective [8][9][10] . According to the literature review, the external application of TCM has bene cial properties such as anti-infection and local immune regulation in the infected wound, which has advantages in the control of acute and chronic local orthopedic infections. For local infections of extremities caused by multi-antibiotic-resistant bacteria with exposure of bone and implants, based on the TCM theory of "simmer pus promoting the regeneration", we have adopted SH decoction to directly soak and wash the infected wounds on the basis of the long-term clinical results of a large number of cases indicating that the local infection of the limbs was obviously controlled. Additionally, we observed that the swelling of the tissue subsided, pus and necrotic tissue decreased, which provided better tissue conditions for reoperation and wound healing. This decoction contains Scutellaria baicalensis Georgi, Coptidis rhizoma, Cortex Phellodendri chinsis and so on, which are heat-clearing and detoxifying drugs. Modern medical research have shown that these Chinese herbs had a broad-spectrum antibacterial and bacteriostatic effects, and no bacterial resistance was found in the combined application. The clinical application of SH prescription can play the role of anti-bacterial and anti-in ammatory in the wound , thus achieving the purpose of eliminating infections [8][9][10] .
This study investigated the in vitro effect of SH decoction on adhesion and bio lm formation of bacteria on the titanium surface, and explored the potential value of this prescription as an effective method to treat orthopedic infections. Since the majority of orthopedic implant-related infections are caused by S. epidermidis and S. aureus 21 . Therefore, we selected the ATCC 35984 and ATCC 43300 to understand bio lm formation in this study. Previous research have demonstrated that bio lm formation proceeds in two phases: primary attachment of staphylococcal cells on a biomaterial, which is followed by bacterial accumulation in multiple layers and glycocalyx formation leading to a mature bio lm, the attachment of Staphylococcus to solid surfaces is strongly seen as an essential step in the formation of bio lm 3,6 . If the primary bacterial adhesion is interfered, then the bio lm formation could be prohibited consequently 3,14 .
The results in this study showed that SH decoction could inhibit the bacterial adhesion on the titanium surface, and there was a trend that the inhibition effect was more obvious with the increase of dilution concentration of SH decoction. The results also showed that SH decoctiom could inhibit the formation of dense structural bio lms of the antibiotic-resistant staphylococcus on the titanium surface by either quantitative or qualitative analysis. Speci cally, the bacterial strains cannot form bio lms when incubated with the 1MIC or 1/2MIC of SH decoction, only showing loose adhesion of the bacterial colonies. Although bio lm was formed at 1/4 MIC, its OD values were obviously less than that of TSB group. At 1/8MIC of SH decoction, bio lm formation was not inhibited compared to the TSB control group. It can be seen from above analysis that SH decoction could inhibit the bio lm formation on the titanium surface, which is dependent on the dilution concentration of the SH decoction.
The results in the work showed that SH decoction with high concentration has an obvious inhibition of the bacterial adhesion and bio lm formation on the titanium surface. However, the mechanism that allows the inhibition of bio lm formation is not clear. Many studies have shown that polysaccharide intercellular adhesin (PIA) is the primary determinant of the accumulation phase of staphylococcal bio lm formation [21][22][23] . Although bacteria with PIA de ciency have been able to adhere to the surface of medical implants at the initial stage of bio lm formation, the ability of a large number of colonies to gather and form bio lm at the later stage were greatly reduced, and even bacterial bio lms could not be formed 3,6,23 . It has been reported that the ica A and ica D partially overlaps in sequence and they together mediate the synthesis of N-acetylglucosamine using UDP-N-acetylglucosamine as a substrate, which composes β-1,6-linked homoglycan to synthesize PIA 24 . It can be seen that the co-expression of the functional ica A and ica D gene can catalyze the synthesis of PIA, and the down-regulation of ica AD expression leads to the reduced activity of N-acetylglucosaminyl transferase, which also leads to reduced amount of PIA [24][25][26] . In addition, studies have shown that the complex gene regulation system of staphylococcus may ultimately in uence the bio lm formation by regulating the expression of ica operon 23,26 . Therefore, in order to verify the role of ica operon in the inhibition of bacterial bio lm formation on the titanium surface by SH decoction, ATCC 35984 with property of positive bio lm formation and ica AD gene expression was selected in this study 11 . The expressions of ica AD were detected by real-time uorescence quantitative PCR, and the results showed that during the process of bio lm formation, the expression level of ica A and ica D genes in the bio lm and culture medium was inhibited by the SH decoction at different dilution concentrations, and the inhibition became more obvious with the increase of SH concentration. In addition, the expression of ica operon is negatively regulated by its upstream inhibitory gene ica R 26 . In our study, we found that SH decoction could not only down-regulate the expression of ica AD, but also up-regulate the expression level of ica R in a concentration dependent manner. Therefore, it could be inferred that the SH decoction activates ica R to inhibit the transcription of ica AD, thus leading to the reduced synthesis of PIA and formation of bio lm on the titanium surface.

Conclusion
It is evident that SH decoction can inhibit bio lm formation of antibiotic-resistant Staphylococcus on the titanium surface based on our in vitro results. Further, there appears to be a concentration dependent relationship between SH decoction and bio lm formation. The mechanism may be that the SH decoction activates ica R to inhibit the transcription of ica AD, thus, leading to reduced synthesis of PIA and prevention of bio lm formation.  with TSB, 1/8MIC SH and 1/4MIC SH group, $ P < 0.01.   CV staining at different culture time. Compared with TSB and 1/8MIC SH groups, * P < 0.01; Compared with 1/4MIC SH group, # P < 0.05.