Effects of exogenous glucose on Pseudomonas aeruginosa biofilm formation and antibiotic resistance

Abstract Pseudomonas aeruginosa is commonly found in nosocomial and life‐threatening infections in patients. Biofilms formed by P. aeruginosa exhibit much greater resistance to antibiotics than the planktonic form of the bacteria. Few groups have studied the effects of glucose, a major carbon source, and metabolite, on P. aeruginosa biofilm formation and on its metabolic pathways. In this study, we investigated the effect of glucose on the biofilm formation ability of P. aeruginosa and carried out a metabolomic analysis to identify whether glucose alters the metabolic activity of P. aeruginosa in biofilms. We found that glucose efficiently promoted P. aeruginosa biofilm formation by upregulating the expression of the extracellular polysaccharide‐related gene pslA. Treatment with glucose caused an increase in 7 metabolites (including 3‐hydroxypropionic acid, glucose‐6‐phosphate, and 2,3‐dimethylsuccinic acid) and a decrease in 18 metabolites (including myo‐inositol, glutamine, and methoxamedrine) in the biofilm. In addition, there was a synergistic effect between glucose and horse serum on biofilm formation when the two were added in combination, which also increased the resistance of biofilm to levofloxacin therapy. Thus, our work sheds light on the underlying mechanisms by which glucose may enhance biofilm formation and identifies novel targets for developing strategies to counteract biofilm formation.

patients with cystic fibrosis (CF) (Van Delden & Iglewski, 1998) and demonstrates extensive antibiotic resistance. In immunocompromised individuals, P. aeruginosa proliferates in a self-produced glycocalyx and forms a biofilm that imparts antibiotic resistance (Wei & Ma, 2013). Pseudomonas aeruginosa biofilm formation includes four major steps: surface attachment, microcolony formation, maturation, and dispersion (Whiteley et al., 2001). Published reports suggest that P. aeruginosa forms biofilms that confer a survival advantage to bacteria, such as increased antibiotic tolerance in the lungs of CF patients (Mah & O'Toole, 2001).
The extracellular matrix of P. aeruginosa is responsible for biofilm architecture and functions as an adhesive, holding biofilm cells together and protecting them against antibiotics, metallic cations, ultraviolet radiation, oxidizing biocides, and host immune defenses (Flemming & Wingender, 2010;Stoodley et al., 2002). Extracellular polysaccharide is a major component of the P. aeruginosa biofilm extracellular matrix (Flemming & Wingender, 2010). P. aeruginosa produces at least three different types of EPS: Pel, Psl, and alginate which contribute to biofilm formation and architecture maintenance (Ryder, Byrd, & Wozniak, 2007). Psl consists of a repeated pentasaccharide containing D-mannose, D-glucose, and D-rhamnose and plays a vital role in the adherence to abiotic and biotic surfaces and in the maintenance of biofilm architecture. Pel, a glucose-rich polysaccharide, is important for biofilm formation in air-liquid interfaces (Byrd et al., 2009). In contrast, alginate exists only in the mucoid P. aeruginosa and is essential for the initial steps of biofilm development (Byrd et al., 2009). Although glucose and its derivatives are the major components of EPS, only few reports have demonstrated the effects and mechanisms of exogenous glucose on P. aeruginosa biofilm formation. Understanding the molecular mechanism of glucose-induced biofilm formation may help in identifying novel targets to eradicate P. aeruginosa biofilms in CF patients or in environmental conditions associated with high levels of glucose and biofilms.
In recent years, many factors such as ethanol, glucosamine, temperature, and subinhibitory concentrations of certain antibiotics have been reported to influence extracellular matrix expression and biofilm formation in vitro (You et al., 2014). As a carbon source and a metabolite, glucose shows multiple effects on bacterial growth and biofilm formation. You et al. (2014) found that glucose induced Staphylococcus aureus biofilm formation through the accessory protein GbaAB in a polysaccharide intercellular adhesiondependent manner. Pan, Breidt, and Gorski (2010) reported that glucose combined with sodium chloride showed a synergistic effect on promoting Listeria monocytogenes biofilm formation through the accumulation of extracellular polymeric substances rather than by increasing the number of viable biofilm cells. However, the effects of glucose on P. aeruginosa biofilm formation and associated metabolic pathway alterations, as assessed by metabolomics, have not been analyzed.
In the present study, we investigated the effects of glucose on biofilm formation by P. aeruginosa strain PAO1 and clinical isolates using a multiphenotypic approach and metabolomics.

| Bacterial strains and plasmids
The PAO1 strain (ATCC 15692) used in this study was kindly provided by Mingqiang Qiao (College of Life Sciences of Nankai University, Tianjin, P.R. China). P. aeruginosa PA47 is a clinical isolate and strong biofilm producer, as we previously reported (She et al., 2018). Other clinical isolates of P. aeruginosa were collected from patients with pulmonary infections at the Third Xiangya Hospital of Central South University (Changsha, Hunan, P.R. China) from January 2014 to December 2014 (Qu et al., 2016). Detailed information on other strains and plasmids used in this study is shown in Table 1.
All strains were stored at −80°C in whole milk culture. Glucose-free Luria-Bertani (LB) broth (Solarbio) was used for bacterial planktonic cell growth and biofilm formation in all experiments; Mueller-Hinton (MH) broth (Solarbio) was used for the antibiotic susceptibility assay.

| P. aeruginosa PA47∆pslA construction
A 573-bp DNA fragment containing the upstream region of pslA was amplified using primers PslA-MF1 and PslA-MR1. A 556-bp DNA fragment containing the pslA downstream region was amplified using primers PslA-MF2 and PslA-MR2. The two PCR products were then amplified using PrimerSTAR Max DNA Polymerase (TaKaRa). A 1,129-bp DNA fragment of the PCR product was ligated into pLP12 with Exnase II (ClonExpress II, Vazyme) to generate pLP12-∆pslA and was further confirmed by primers pLP-UF and pLP-UR. DH5α and PA47 wild-type strains were transformed with this plasmid. Deletion of pslA was confirmed by PCR using primers PslA-MF1 and PslA-MR2. Detailed information about the strains and primers is shown in

| Growth curve
Glucose was diluted with LB broth to a final concentration of 1%-4%.

| Confocal laser scanning microscope (CLSM) imaging and analysis
Biofilms of PA47 were grown on coverslips in the presence or absence of 2% or 4% glucose as described above. Three fluorescent dyes (SYTO9 for total biomass, propidium iodide for dead cells, and ConA for α-polysaccharides; Thermo Fisher Scientific) were added to each coverslip, and after staining in the dark, samples were imaged using a CLSM (Zeiss LSM 800). The excitation and emission wavelengths for the three dyes were as follows: SYTO9: 633 nm and 650-700 nm; propidium iodide: 458 nm and 460-500 nm; and ConA: 543 nm and 550-600 nm; ImageJ software was used to quantify the biofilm biomass (Qu et al., 2016).

| Extracellular polysaccharide analysis by phenol-sulfuric acid
Biofilms of PA47 with or without 4% glucose treatment were constructed in 6-well cell culture plates. The planktonic cells and free glucose were thoroughly removed by washing five times with saline and the biofilms were scraped into 0.5 ml saline with moist swabs. After mixing, 0.5 ml of phenol (5%) and 5 ml of concentrated sulfuric acid was added and incubated for 1 hr in the dark, and the A490 was measured (Musthafa, Sivamaruthi, Pandian, & Ravi, 2012).

| qRT-PCR
Gene expression was detected by qRT-PCR according to our previous study (Qu et al., 2016). Briefly, overnight cultures of PA47 grown in LB broth in the presence or absence of 2% glucose were collected at the A630 of 0.5-0.8. Total RNA was extracted using an E.Z.N.A.
Total RNA Kit II (Omega Bio-tek). The RNA purity and concentration was determined by the absorbance at 260/280 nm, and 1 μl of RNA was used for cDNA synthesis by TransScript All-in-One First-Strand cDNA Synthesis SuperMix (Transgene). qPCR was performed using TransStartTM Green qPCR SuperMix UDG (Transgene) using a real-time quantitative PCR system (Eppendorf). The oligonucleotide primers used to amplify the housekeeping gene 16S rRNA and EPS-related genes (pelA, pslA, and alg44) (Kim, Park, & Lee, 2015) are shown in Table A1.

| Metabolite extraction
Overnight culture of 8 P. aeruginosa clinical isolates was diluted with LB broth to ~10 6 CFU/ml in the presence/absence of 4% glucose in a 6-well microplate. After incubation at 37°C for 24 hr, the planktonic cells were removed; the remaining biofilms were collected with a cell scraper and washed four times with 1 × PBS, and then subjected to metabolomics analysis. Briefly, samples were collected in 2-ml tubes and extracted with 1 ml of extraction reagent (V Methanol : V Chloroform = 3:1). After vortexing for 30 s, samples were homogenized using a ball mill for 4 min at 45 Hz and then treated with ultrasound 5 times for 5 min each (incubated in ice water). After centrifugation for 15 min at ~12,000 g, and 4°C, the supernatant (0.9 ml) was transferred to a 2 ml GC/MS glass vial. After the sample was completely dried in a vacuum concentrator with no heating, 60 μl of methoxyamine hydrochloride (20 mg/ml in pyridine) was added to the sample and incubated for 30 min at 80°C. Next, 80 μl of BSTFA reagent (1% TMCS, vol/vol) was added to the samples and incubated for 1.5 hr at 70°C. All samples were analyzed by a gas chromatograph system coupled with a Pegasus HT time-of-flight mass spectrometer (GC-TOF-MS).

| GC-TOF-MS analysis
GC-TOF-MS analysis was performed using an Agilent 7890 gas chromatograph system coupled with a Pegasus HT time-of-flight mass spectrometer. The system utilized a DB-5MS capillary column coated with 5% diphenyl cross-linked with 95% dimethylpolysiloxane (30 m × 250 μm inner diameter, 0.25 μm film thickness; J&W Scientific). An 1 μl aliquot of the analyte was injected in splitless mode. Helium was used as a carrier gas; the front inlet purge flow was 3 ml per min, and the gas flow rate through the column was 1 ml per min. The initial temperature was kept at 50°C for 1 min, then raised to 310°C at a rate of 10°C per min, and then maintained at 310°C for 8 min. The injection, transfer line, and ion source temperatures were 280, 280, and 250°C, respectively. The energy was −70 eV in electron impact mode. Mass spectrometry data were acquired in full-scan mode with the m/z range of 50-500 at a rate of 20 spectra per second after a solvent delay of 6.27 min (Garcia & Barbas, 2011;Kind et al., 2009).

| MIC determination
The MIC was detected by the broth microdilution method as previ-

| MBEC determination
Minimal biofilm eradication concentration was assessed by the XTT staining method as described above. Briefly, overnight cultures were added to microplates with or without 4% glucose at 37°C for 24 hr.
After removing the planktonic cells with saline, LB broth containing different concentrations of OFLX (ranging from 1,024 to 0.0625 μg/ ml) were added to each well. After incubation for 24 hr, planktonic cells were removed by saline washing, and XTT combined with phenazine methosulfate was added to each well as described above. The MBEC 50 and MBEC 90 were defined as the eradication of 50% and 90% of the 24 hr preformed biofilm, respectively, when compared to the untreated controls.

| Statistical analysis
Data were analyzed by GraphPad Prism 7.0 software. For all figures with bar graphs, the data are represented as the mean ± standard deviation. ANOVA and Student's t test were used to determine the statistical significance. A p-value <.05 was considered indicative of statistical significance. For metabolomics data preprocessing and annotation, Chroma TOF 4.3X software (LECO Corporation) and the LECO-Fiehn Rtx5 database were used for raw peak extraction, baseline data filtering, and calibration of the baseline, peak alignment, deconvolution analysis, peak identification, and integration of the peak area. Both the retention index and mass spectrum match were considered in metabolite identification.

| Biofilms of P. aeruginosa clinical isolates are enhanced by glucose treatment
The biofilm biomass of P. aeruginosa was determined by crystal violet staining and XTT assay. Though the clinical isolates showed different sensitivities to glucose treatment, we found that glucose increased the biofilm formation in all of the tested P. aeruginosa strains (PAO1 and 12 clinical isolates) at a concentration of 2% (Figure 1a).

| Glucose increases EPS production by upregulating pslA gene expression
Extracellular polysaccharide is a major component of the biofilm extracellular matrix. Increased secretion of the polysaccharide was confirmed by a phenol/concentrated sulfuric acid assay in 6-well microplates ( Figure 3a). Moreover, since the fluorescent dye ConA binds to the α-mannopyranosyl and α-glucopyranosyl sugars in the biofilm matrix (Chen, Lee, Tay, & Show, 2007), the CLSM images and ImageJ analysis showed that the EPS was markedly increased in the presence of 4% glucose (Figure 3b). Using the housekeeping gene 16S rRNA as reference, qRT-PCR analysis showed that compared with the no glucose control, 4% glucose significantly increased the expression of pslA (PslA) and alg44 (alginate) genes by 6.0-fold (p < .05) and 3.2-fold (p < .05), respectively, without influencing pelA (PelA) gene expression (p > .05) (Figure 3c). Based on these findings, we constructed a pslA knockout strain of PA47⊿pslA. We found that compared to that of the untreated group, the biofilm biomass was significantly increased in the wild-type PA47 group in the presence of 2% and 4% glucose (p < .05) at 16 hr, but there was no significant difference in the PA47⊿pslA group in the presence of glucose ( Figure 3d). So, we assumed that glucose could increase the biofilm formation of P. aeruginosa by enhancing pslA gene expression in a moderate dosing-dependent manner.

| Glucose alters the metabolic pathways in P. aeruginosa
By metabolomics analysis, we found that P. aeruginosa metabolism was significantly affected by 4% glucose treatment (Figure 4).
Principal component analysis (PCA) was carried out to assess the quality of samples. Eight clinical isolates of P. aeruginosa in the presence or absence of glucose were divided into two groups based on their distinctive profiles in the PCA plots ( Figure 4a). By using oneway ANOVA, a total of 55 molecular features were identified to be significantly different (p < .05) between the two groups (Table A2).

| Synergistic effect of glucose and HoS on increased resistance of P. aeruginosa biofilm to OFLX
Many infections occur in the presence of serum. By crystal violet staining, we found that the presence of 3.125% HoS significantly increased P. aeruginosa biofilm formation (Figure 5a). CLSM imaging and ImageJ analysis revealed that HoS increased both the coverage and thickness of the biofilm biomass ( Figure 5b). However, clinical isolates showed a different response to 2.5% HoS. While there was a significant inhibitory effect of 2.5% HoS on biofilm formation in PA07, HoS increased the biofilm formation by PA09 and PA47 and did not influence PA01 (Figure 5c). Interestingly, there was a significant synergistic biofilm-promoting effect of HoS and glucose when added together (Figure 5d). By XTT staining assay, we found that the combined glucose and HoS treatment caused a significant increase in resistance to OFLX when compared to glucose treatment alone (p < .05) (Figure 5e).
This finding was further confirmed by CLSM and ImageJ analysis. Additionally, resistance to OFLX increased as biofilm formation increased in PA47, and at the same time, the percentage of live cells in the biofilm biomass also increased (Figure 6a,b). While there was a minimal effect of 2% glucose on the MICs of the clinical isolates (PA01, PA07, PA09 and PA47) in the planktonic form, glucose increased the MBEC 50 and MBEC 90 of these clinical isolates ranging from 2-to >256-fold in the biofilm forms (Table 2). Recently, evidence about the signaling role of matrix EPS has begun to emerge. Irie et al. (2012) found that the major matrix component of P. aeruginosa, Psl polysaccharide, serves as a signal for the production of cyclic-di-GMP (c-di-GMP), a key second messenger that drives a switch toward the biofilm phenotype in P. aeruginosa (Ha & O'Toole, 2015). Intracellular levels of c-di-GMP may be induced by supplementing cell cultures with purified Psl and by increasing psl expression using an inducible promoter (Irie et al., 2012).
C-di-GMP, like cAMP, c-di-AMP, cGMP, and (p)ppGpp, is an important intracellular signaling molecule (Bharati & Chatterji, 2013;Kalia et al., 2013). C-di-GMP is a ubiquitous bacterial second messenger (Hengge, 2009) involved in the regulation of surface attachment, aggregation, and biofilm formation in a wide range of bacteria (Duvel et al., 2012). Moreover, c-di-GMP also interacts with the quorum sensing (QS) system (Sharma, Petchiappan, & Chatterji, 2014), a cell-to-cell communication phenomenon, involved in the regulation of genes that control biofilm formation, sporulation, bioluminescence, and virulence (Hammer & Bassler, 2003), and is mediated by a variety of small diffusible auto-induced signaling molecules (Ueda & Wood, 2009). Taken together, these results suggest that the biofilm-promoting effects of glucose on P. aeruginosa may be due to the Psl-dependent upregulation of c-di-GMP.
The metabolomics analysis revealed that glucose treatment significantly altered metabolite production in P. aeruginosa. We found that glucose treatment mainly enhanced the production of 3-hydroxypropionic acid and glucose-6-phosphate and reduced the production of myo-inositol and glutamine (Figure 4). 3-Hydroxypropionate is an intermediate of the 3-hydroxypropionate and 3-hydroxypropionate/4-hydroxybutyrate cycle, two of the six pathways for autotrophic carbon dioxide fixation in microorganisms (Kumar, Ashok, & Park, 2013). However, few of the pathways described in the previous sections have been studied with glucose as a substrate. Therefore, the addition of exogenous glucose, which is a major carbon source, may increase the production of 3hydroxypropionate. The catalysis of glucose to glucose-6-phosphate by hexokinase is the first step of glycolysis (https ://www.kegg.jp/kegg-bin).
To some degree, the production of glucose-6-phosphate is positively corrected with the amount of the glucose substrate. As a consequence, the production of glucose-6-phosphate is promoted by the increase in exogenous glucose; Myo-inositol is synthesized from glucose-6-phosphate using two sequentially acting enzymes: Inositol-3-phosphate synthase converts glucose-6-phosphate to inositol-3-phosphate, and then inositol monophosphatase dephosphorylates inositol-3-phosphate to generate myo-inositol (Reynolds, 2009). In the inner biofilm, the cells prefer to acquire nutrition by transferring carbohydrate metabolism to glycolysis due to the lack of oxygen. Therefore, with the decrease in myo-inositol, more glucose-6-phosphate will accumulate to participate in glycolysis to produce ATP. Similarly, the production of glutamine and methoxamedrine is probably also inhibited by P. aeruginosa in the same way. These findings suggest that targeting the enzymes associated with these metabolites may influence glucose-induced biofilm growth and warrants further investigation.
Once the sugars (including polysaccharide and monosaccharide) are digested by gastrointestinal tract, the commonest existent form is glucose. And when patients are infected with bacteria, the circumstances around bacteria are blood or other body fluids also containing glucose. Moreover, published reports suggest that P. aeruginosa forms biofilms in the CF lung (Mah & O'Toole, 2001). And the CF patients are commonly complicated by glucose tolerance, hyperglycemia, or diabetes mellitus. The prevalence of CF-related diabetes affects more than 1/4 CF patients over the age of 20 (Moran, 2000). It has recently been reported that uncontrolled hyperglycemia in diabetics increase airway glucose, providing a richer growth medium that facilitates P. aeruginosa infection (Baker et al., 2006). And the airway glucose is an important determinant of increased bacterial loads during diabetes (Gill et al., 2016). Because HoS is easy to be obtained and has similar effects on P. aeruginosa biofilms as human serum ( Figure A1), many studies have tested the antimicrobial effects of agents by adding HoS to neutralize the differences between the in vitro properties of antibiotics. To simulate the body fluid environment of the human body, we detected the biofilm-promoting activity of glucose in the presence of HoS. In this study, we found that the free glucose in the heat-inactivated serum could be synergy with serum components enhancing P. aeruginosa biofilm formation and antibiotic resistance (Figures 5e and 6). These findings are in agreement with those of Samaranayake, Anil, Hashem, Vellappally, and Cheung

| CON CLUS ION
This work adds to a growing understanding of the positive role of glucose in promoting P. aeruginosa biofilm formation. This is the first study to demonstrate that glucose causes metabolic changes associated with biofilm formation. Thus, our work sheds some light on the underlying mechanisms by which glucose enhances biofilm formation and identifies novel targets for developing strategies to counteract biofilm formation. Our findings also imply that higher dose of OFLX is needed to eliminate biofilms formed in the presence of glucose.

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

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

E TH I C S S TATEM ENT
None required.

DATA AVA I L A B I L I T Y S TAT E M E N T
All primer sequence data and the metabonomic analysis data are provided in Tables A1 and A2, respectively. TA B L E A 2 (Continued)