Efficacy of levofloxacin against biofilms of Pseudomonas aeruginosa isolated from patients with respiratory tract infections in vitro

Abstract Microbial biofilms are formed in a variety of clinical situations and increase antibiotic resistance of the pathogen by almost ~1,000 times. The effect of levofloxacin (OFLX) on the biofilms of Pseudomonas aeruginosa strain PAO1 and the clinical isolates was investigated by crystal violet staining and confocal laser scanning microscope. The transcriptional alteration in the PAO1 biofilms upon OFLX treatment was also analyzed by RNA sequencing (RNA‐seq). We found that while OFLX significantly inhibited P. aeruginosa biofilm formation in a dose‐dependent manner, it could not completely eradicate preformed biofilms even at higher concentrations. RNA‐seq revealed that PAO1 genes related to metabolism, formation of secondary metabolites, and quorum sensing biosynthesis were differentially expressed in the biofilms treated with OFLX. Our data might be useful in determining the optimum OFLX concentration needed for P. aeruginosa biofilm inhibition and eradication in patients with respiratory tract infections.


| INTRODUC TI ON
Biofilm is a three-dimensional structured community of microbial cells that adhere to a biotic or abiotic surface and are enclosed in a self-produced polymeric matrix (Costerton, Stewart, & Greenberg, 1999). Biofilm formation is initiated by adherence of single planktonic cells to a surface, leading to the development microcolonies, which then further grow into mature biofilms that produce components of the extracellular matrix (ECM) to maintain the distinctive structures of cellular aggregates (Hentzer et al., 2003). They act as barriers to antimicrobial agents and protect the colonies from any environmental fluctuations. When encased in a biofilm, bacteria can be almost 1,000-fold more resistant to antibiotics compared with their planktonic counterparts, often rendering the antibiotic therapy ineffective (Chu et al., 2016;Drilling et al., 2014). This is probably due to the slow growth, metabolic shift, production of ECM, and different cell surface properties of the embedded bacteria (Stryjewski & Corey, 2014). According to published reports, over 80% microbial infections in humans are due to biofilms (Musk & Hergenrother, 2006).
With the increasing number of immunocompromised patients, the clinical relevance of opportunistic pathogens has increased in recent years. Pseudomonas aeruginosa is one of the most important opportunistic human pathogens (Sousa & Pereira, 2014).
These gram-negative bacteria kill thousands of people annually and are responsible for 10% of all hospital-acquired infections (Davies et al., 1998). The biofilms of P. aeruginosa play a substantial role in cystic fibrosis pneumonia, chronic wound infections, chronic otitis media, chronic bacterial prostatitis, and medical device-related infections (Rybtke, Hultqvist, Givskov, & Tolker-Nielsen, 2015).
Levofloxacin (OFLX) belongs to a new class of fluoroquinolones, and its activity on Stretococcus pneumoniae and other respiratory pathogens has been widely studied and documented (Marchetti & Viale, 2003). However, the inhibitory action of OFLX against P. aeruginosa biofilms is still controversial due to the different drug concentrations, duration of treatment, and the age of biofilms in different studies. To the best of our knowledge, there is no report so far that evaluates the effect of OFLX on biofilm transcriptomics. We have examined the inhibitory effects of OFLX on the biofilms of P. aeruginosa strain PAO1 and clinical isolates by crystal violet (CV) staining and confocal laser scanning microscopy (CLSM), along with the morphological changes in the biofilms associated with OFLX therapy.
To determine the potential mechanisms of the antibiofilm effect of OFLX, we analyzed the differentially expressed genes in the biofilms using high-throughput RNA-seq.  (Qu et al., 2016). All of the strains were stored at −80°C in whole milk culture. And Luria-Bertani (LB) broth (Solarbio, Shanghai, China) was used for bacterial culture in all experiments.

| Determination of the minimal inhibitory concentration
The minimal inhibitory concentration (MIC) was detected by the broth microdilution method as previously described by The CLSI 2015. Briefly, twofold dilutions of OFLX (ranging from 1,024 to 0.0625 μg/ml) were prepared in Mueller-Hinton broth, and 100 μl was dispensed per well in a 96-well plate. Each well was then inoculated with 10 μl of P. aeruginosa suspension, and the plates were incubated at 35°C for 16~18 hr. The MIC was determined as the lowest concentration without any visible bacterial growth. McFarland standard 0.5 was used as the reference for turbidity.

| Biofilm inhibition and eradication assay
The biofilm inhibitory assay was performed in microplates as previously described (Qu et al., 2016). Briefly, 4 μl overnight culture and 196 μl LB broth were dispensed per well in a 96-well microplate, and exposed to different concentrations of OFLX. After incubation for 24 hr at 37°C, biofilm biomass was determined by the CV staining method as described. For biofilm eradication (Christensen et al., 1985), overnight culture was diluted 1:200 with LB broth, and 100 μl bacterial suspension was dispensed per well in microplates.
After static incubation for 24 hr at 37°C, the planktonic cells were removed with saline, and 200 μl LB broth with OFLX was added to each well. After incubation for another 24 hr, biofilm biomass was detected with CV staining.

| CLSM observation
In order to investigate the biofilm characteristics affected by OFLX, glass slides were placed in six-well microplates containing 2 ml LB broth with OFLX, and 40 μl overnight culture was added per well. After incubation for 24 hr, the planktonic cells were removed and the biofilms remaining on the glass slides were stained with LIVE/DEAD BacLight Bacterial Viability Kit (Thermo Fisher Scientific, Shanghai, China) as recommended. Briefly, 1.5 μl of SYTO green and PI red dye mix was diluted in 1 ml PBS (pH 7.4), and 100 μl of the dye solution was dropped on each glass slide. After incubation in the dark for 15 min, the glass TA B L E 1 Classification of biofilm production Average A 570 nm value Biofilm production

≤A c Non
Note. Absorbance cutoff value (A c ) = average A 570 nm of negative control (ATCC27853) + 3 × standard deviation (SD) of negative control. slides were observed by CLSM (Zeiss LSM 800, Jena, Germany), and the biofilm biomass was quantified with ImageJ software.

| RNA sequencing
Pseudomonas aeruginosa PAO1 was incubated in LB broth with or without 1 μg/ml OFLX (MIC = 2 μg/ml) for 16 hr. After removing the planktonic cells, the biofilms were collected with moist swabs.
Three independent biological replicates were used for further anal-

| Statistical analysis
All experiments were repeated three times to validate the reproducibility. And the data were analyzed by using GraphPad Prism 7.0 software (GraphPad Software, San Diego, CA, USA), and Student's t test was used to determine the statistical significance of the data. The level of statistical significance was set at p < 0.05.

| Effects of OFLX on biofilm formation and eradication
The effects of OFLX on PAO1 biofilm formation and its eradication are shown in Figure 1. Treatment with OFLX at the MIC (2 μg/ml) significantly inhibited PAO1 biofilm formation and continued its inhibitory effect in a dose-dependent manner (Figure 1a). However, higher concentrations of OFLX were needed to eradicate the preformed biofilms. Although biofilms were significantly eradicated by OFLX at 8 μg/ml, there was no significant decrease in the biofilm biomass even at the concentration of 32 μg/ml (Figure 1b).

| Biofilm-forming capacity of pulmonary isolates
OFLX tolerance of the biofilms of 30 clinical isolates of P. aeruginosa, isolated from the sputum or respiratory tract irrigate of pulmonary infection patients, was tested. CV staining identified five (16.67%) strains as biofilm negative, four (13.33%) as weak, four (13.33%) as moderate, and 17 (56.67%) as strong biofilm producers (Figure 2a), which was consistent with the report of Olson et al. (2002).
In addition, significant differences were seen among different biofilm-forming groups by CV staining (Figure 2b). Of the 17 strong biofilm producers, four strains (PA01, PA07, PA09, and PA47) were selected for further experiments (Figure 2c).

| Antibiofilm effect of OFLX on pulmonary isolates
Biofilms of different clinical isolates showed different susceptibility patterns to OFLX. It significantly inhibited the biofilm formation of PA01, PA07, PA09, and PA47 strains at the respective MIC values (2, 0.25, 0.0625, and 0.25 μg/ml for PA01, PA07, PA09, and PA47, respectively; p < 0.05), as well as in a dose-dependent manner ( Figure 3a). PA47 was the strongest biofilm producer among them, but the biofilm of PA01 showed the highest resistance to OFLX.
However, biofilms of all strains showed higher resistance to complete eradication by OFLX. It could significantly eradicate the biofilms of PA01, PA07, PA09, and PA47 at the concentrations of 16, 4, and 2 μg/ml, respectively (Figure 3b). In addition, the 24-hr biofilm of PA47 was fully resistant to OFLX even at 128 μg/ml. Finally, there was no significant decrease in the preformed biofilm biomass with increasing concentration of OFLX.

| Morphological characteristics of P. aeruginosa biofilms
While the untreated biofilms were thick and relatively homogeneous, treatment with 2 μg/ml OFLX significantly decreased the biofilm F I G U R E 1 Antibiofilm activity of OFLX on PAO1. The effects of OFLX on biofilm formation (n = 3) (a) and eradication (n = 3) (b) at different concentrations, respectively. The data are reported as absorbance at 570 nm (A 570 nm ) of residual biofilm biomass and disrupted its structural integrity (Figure 4a,c). OFLX could slightly disrupt the integrity of preformed biofilms ( Figure 4b) and even eradicated the bulk of the biomass to some extent, but living cells remained even at the concentration of 32 μg/ml (Figure 4d).
In addition, significant morphological alterations were observed in the PAO1 strain and the clinical isolate PA47 following OFLX treatment ( Figure 5). Rounded or rod-shaped cells were observed in the untreated sample, while elongated cells with or without swelling appeared in the presence of OFLX.

| OFLX-inhibited biofilms exhibit a novel transcriptome
In order to avoid the effect of cell growth inhibition on the ex-

| D ISCUSS I ON
Pseudomonas aeruginosa can adhere to the surface of the respiratory tract and form a biofilm, which is considered to be major cause  Fluoroquinolones have been used traditionally against the slow-growing and nongrowing bacteria (Eng, Padberg, Smith, Tan, & Cherubin, 1991), and OFLX has shown a considerable killing effect on P. aeruginosa biofilms in experimental studies (Mikuniya et al., 2005). In the present study, OFLX significantly inhibited biofilm formation of both PAO1 and respiratory isolates of P. aeruginosa, but showed moderate or slight inhibitory effects on the preformed biofilms. In addition, in accordance with the previous studies (Monden, Ando, Iida, & Kumon, 2002;Nielsen & Nielsen, 1987), the round PAO1 and PA47 cells were elongated in the presence of OFLX.
Finally, RNA-seq indicated differential gene expression in PAO1 biofilms upon OFLX treatment.  (Mulcahy, Burns, Lory, & Lewis, 2010), and (d) formation of small colony variants (Evans, 2015), as well as other unknown resistance mechanisms. Once the antibiotic is removed, the remaining biofilm resumes growth and returns to its initial size (Drago et al., 2011).
CLSM showed significant elongation of P. aeruginosa cells in the biofilms following OFLX treatment. In a study by Monden et al. (2002), scanning electron microscope, transmission electron microscope, and CLSM were used to investigate the morphological changes in P. aeruginosa biofilms in the presence of OFLX and fosfomycin. They also demonstrated a clear OFLX-induced elongation of the surface biofilm cells, but this change was not clear in the embedded cells. Since fluoroquinolones disrupt the membrane potential or electron transport (Celesk & Robillard, 1989), it is possible that changes in the bacterial outer membrane lead to changes in cell wall synthesis and fluidity (Kahan, Kahan, Cassidy, & Kropp, 1974).
However, little is known about the specific mechanisms and consequences of the shape change in these cells.
Pseudomonas aeruginosa biofilms showed significant changes in their transcriptomic profiles after OFLX treatment compared to the untreated cells, with a total of 129 DEGs. Contrary to our expectations, there were only a few of these DEGs that were associated with the well-known biofilm formation-related genes like the quorum sensing genes (lasI/R, rlhI/R, pqsR etc.) (Lee & Zhang, 2015) and cyclic diguanylate cyclase expression-associated genes (siaA and siaD) (Chen et al., 2015). Instead, most of the DEGs did not have any specific function. It is possible that instead of a unique influential factor, the inhibitory effects of OFLX on P. aeruginosa biofilm synthesis depend on multiple factors. KEGG pathway analysis revealed that the DEGs mainly contributed to metabolic processes, secondary metabolites, and amino acid biosynthesis.
Interestingly, though OFLX showed significant inhibitory effects on biofilm formation, there was more or equal amount of upregulated genes than downregulated genes among these significantly changed pathways. It is possible that the working concentration of OFLX we used was 1 μg/ml. And the sub-MIC of OFLX used in the transcriptomic analysis resulted in stronger expression of antibiotic resistance-related genes rather than inhibition of biofilm synthesis-related genes. Cipriani, Giordano, Magni, Papa, and Filadoro (1995) (2017) showed that sub-MIC levels of antibiotics can lead to resistance and cross-resistance across several classes of antibiotics in wild strains of S. aureus, possibly by free radical production. Finally, Gupta, Chhibber, and Harjai (2016) found that sub-MIC of ciprofloxacin could inhibit P. aeruginosa biofilm formation and virulence factor production by targeting the QS system. However, in our study, the QS-related pathway was significantly upregulated in the presence of sub-MIC OFLX, indicating different mechanisms of inhibiting P. aeruginosa biofilm formation by ciprofloxacin and OFLX at sub-MIC, though both are fluoroquinolones.
In summary, OFLX can inhibit P. aeruginosa biofilm synthesis in a dose-dependent manner, but cannot fully eradicate the preformed biofilms even at high concentrations. In addition, 1 μg/ml OFLX significantly altered the expression of PAO1 genes related to metabolic processes, cellular processes, single-organism processes, molecular binding, and catalytic activity. Our data can help determine the optimum OFLX concentration needed for biofilm inhibition and eradication of P. aeruginosa infection.

ACK N OWLED G M ENTS
This study was financed by The New Xiangya Talent Project of The Third Xiangya Hospital of Central South University, China (grant no. 20150309).

CO N FLI C T O F I NTE R E S T
None declared.

AUTH O R S CO NTR I B UTI O N S
She P and Wu Y designed experiments. She P and Luo Z carried out experiments. Chen L analyzed experimental results. She P and Wu Y wrote the manuscript.

DATA ACCE SS I B I LIT Y
The raw data of transcriptome sequencing for control group and OFLX-treated group were shown in Supporting Information Tables   S1-S3 and Supporting Information Tables S4-S6, respectively. And the gene different expression analysis was shown in Supporting Information Table S7. Pathogenic effects of biofilm on Pseudomonas Aeruginosa pulmonary