Accepted Article

Received Date : 30-Nov-2013 Revised Date : 17-Apr-2014 Accepted Date : 06-May-2014 Article type : Research Paper Editor : Thomas Bjarnsholt Manuscript category: Virulence Factors Catheter-Related Infections Caused by Pseudomonas aeruginosa: Virulence Factors Involved and their Relationships Olejnickova Katerina, Hola Veronika, Ruzicka Filip

Institute for Microbiology, Faculty of Medicine, Masaryk University and St. Anne's University Hospital Brno

Address of corresponding author: Veronika Hola Institute for Microbiology, Faculty of Medicine MU and St. Anne's University Hospital Brno Pekarska 53 CZ-65691 Brno Czech Republic e-mail: [email protected] phone: +420543183093 fax: +420543183089

Key words: Pseudomonas aeruginosa, virulence factors, biofilm, antibiotic resistance, urinary tract, blood stream

Running title: Catheter-related infections caused by Pseudomonas aeruginosa

Catheter-Related Infections Caused by Pseudomonas aeruginosa: Virulence Factors Involved and their Relationships Olejnickova Katerina, Hola Veronika, Ruzicka Filip

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/2049-632X.12188 This article is protected by copyright. All rights reserved.

Accepted Article

Abstract The nosocomial pathogen Pseudomonas aeruginosa is equipped with a large arsenal of cellassociated and secreted virulence factors, which enhance its invasive potential. The complex relationships among virulence determinants have hitherto not been fully elucidated. In the present study, 175 catheter-related isolates were observed for the presence of selected virulence factors, namely extracellular enzymes and siderophores production, biofilm formation, resistance to antibiotics, and motility. A high percentage of the strains produced most of the tested virulence factors. A positive correlation was identified between the production of several exoproducts, and also between the formation of both types of biofilm. An opposite trend was observed between the two types of biofilm and the production of siderophores. While the relationship between the submerged biofilm production (i.e. the biofilm formed on the solid surface below the water level) and the siderophores secretion was negative, the production of air-liquid interface biofilm (i.e. the biofilm floating on the surface of the cultivation medium) and the siderophores secretion correlated positively. All correlations were statistically significant at the level p=0.05 with the correlation coefficient γ≥0.50. Our results suggest that: (i) the co-production of the lytic enzymes and siderophores can play important role in the pathogenesis of the catheter-related infections and should be taken into account when the virulence potential is assessed; (ii) biofilm-positive strains are capable of forming both submerged and non-attached A-L biofilms; (iii) the different microenvironment in the submerged biofilm and A-L biofilm layers leads to opposite consequences for the production of other virulence factors.

Introduction Nosocomial infections are a serious problem of modern health care. The increasing number of immunocompromised patients leads to a great interest in the surveillance of nosocomial pathogens such as Pseudomonas aeruginosa. These infections are difficult to treat, which is connected with longer hospitalization and increased mortality (Palleroni, 2005). P. aeruginosa endangers especially vulnerable hosts, including CF patients, immunocompromised persons or patients with indwelling medical devices like catheters. P. aeruginosa is also one of the main causative agents of catheter-related nosocomial urinary tract infections (Hola et al., 2010; Mittal et al., 2009). The insertion of a catheter (either urinary or bloodstream) enhances the risk of bloodstream infection in already fragile patients (Trautner & Darouiche, 2004). The mortality in immunocompromised patients with indwelling catheters is high, ranging between one third to almost two thirds of cases (Rossolini & Mantengoli, 2005). The This article is protected by copyright. All rights reserved.

Accepted Article

pathogenesis of these infections is multifactorial and involves cell-associated factors as well as secreted virulence determinants (Stehling et al., 2008). The complex relationships among virulence determinants have hitherto not been fully elucidated, and the aim of our study was to extend the current knowledge within the field. The biofilm formation is crucial in the pathogenesis of catheter-related infections (Trautner & Darouiche, 2004). Biofilm was interpreted in many studies as a surface-adherent structure which forms at submerged solid surfaces (Deligianni et al., 2010; Singh et al., 2002). In our research, we referred to this structure as submerged biofilm. Several recent papers have also examined biofilm not attached to a solid surface (see the review Bjarnsholt et al., 2013). A floating pellicle at the interface of an air-culture medium is called air-liquid interface (A-L) biofilm, and recently it has been described in several microbial species (Koza et al., 2009; Martí et al., 2011). Air-liquid biofilm is formed on the surface of a liquid medium, where nutrient and oxygen gradients are opposing, with access to the oxygen above and nutrients below. The consequences of A-L biofilm for the pathogenesis of P.aeruginosa infections have not been observed and explained thus far, although the human body contains locations suitable for the development of such biofilm: lungs, where the solid surface of alveoli is continually moistened by blood which forms the air-liquid interface, or other locations accessible for air such as endotracheal tubes or other tubing present in the respiratory tract. The airways and urinary tract are the most important gates for bloodstream-related infections caused by P. aeruginosa; therefore, in this paper, correlations between the formation of submerged biofilm and A-L biofilm and the production of other virulence factors (proteolytic enzymes, haemolysins and siderophores, different types of motility) are presented. P. aeruginosa is equipped with several powerful enzymes. Proteases, such as LasB elastase, help to break down the host defence by cleaving the tissue and altering the relevant immune response. Two groups of haemolysins, namely phospholipase C and rhamnolipids, cause the lysis of various cells, protects the microbe against the host immune response, and enhances bacterial growth due to the release of iron ions (Mittal et al., 2006; van Gennip et al., 2009). Moreover, the production of rhamnolipids by colonising P. aeruginosa isolates is related to the development of ventilator-associated pneumonia (Köhler et al., 2010). Siderophores, i.e. yellow-green pyoverdines and brown pyochelin, facilitate the acquirement of iron in the host body. The blue-green pigment pyocyanine is involved in phosphate uptake (Palleroni, 2005). P. aeruginosa performs several types of motion, namely swimming powered by rotating flagella in liquid environment, twitching enabled by the extension and retraction of type IV pili, and swarming; the latter process is a rapid multicellular movement across a surface This article is protected by copyright. All rights reserved.

Accepted Article

(Rampioni et al., 2009). Motility facilitates the colonization and dissemination of the bacteria in the host body. The influence of motility, namely swarming, on the development of catheterrelated infections was previously described in Proteus sp. (Jones et al., 2004; Hola et al., 2012). P. aeruginosa virulence factors are usually associated with acute infection within the host tissue (Stehling et al., 2008). However, scarce information is available on the contribution of P. aeruginosa virulence factors to the development of catheter-related infection. We focused on the virulence factors potentially relevant to the development and persistence of catheterrelated infections (namely extracellular enzymes and siderophores production, biofilm formation, resistance to antibiotics, and motility). The determination of relationships among these virulence factors could enable us to identify their importance for catheter-related infections. Material and methods Bacterial strains The total of 175 P. aeruginosa isolates used in this study were obtained from patients with urinary or bloodstream catheter-related infections treated at St. Anne's University Hospital, Brno, the Czech Republic. All the strains were identified by biochemical tests (NefermTEST and OXItest, Lachema, CZ). The strains were stored in a cryoprotective medium at -75°C. For the tests, a fresh 24-hour culture of a particular strain was used.

Collection of patient-related, microbiological, and clinical data In all the isolates, the patient-related and clinical data (sex, age, type of catheter, number of isolated species, etc.) were monitored (Tab. 1). After the species were microbiologically examined, the microbiological data (antibiotic resistance, motility, biofilm formation, exoproducts secretion, etc.) were added to the dataset. The data were processed anonymously under specially assigned codes.

Biofilm assays An assay for the formation of submerged biofilm was performed in a microtiter plate according to Stepanović et al. (2007). The P. aeruginosa overnight culture was suspended in Brain Heart Infusion (Oxoid, UK) supplemented with 4% of glucose (BHI-g) to OD600 of 0.5. The culture was diluted 10 times in the microtiter plate. Following the incubation carried out at 37°C for 24hrs, the microtiter plate wells were washed three times with water, and the This article is protected by copyright. All rights reserved.

Accepted Article

biofilm was fixed by air drying. The adherent biofilm layer was stained with crystal violet (2% solution) and dried out; subsequently, the dye was redissolved with 150μl of ethanol and measured at OD595. Each strain was tested three times. The assay for the formation of A-L biofilm was performed using polystyrene tubes. The preparation of the P. aeruginosa suspension and the growth conditions were both adopted from Stepanović et al. (2007). The tubes with 200µl of the suspension and 1.8ml of BHI-g were incubated at 37°C for 24hrs. After the cultivation, the presence and character of the pellicle on the surface of the medium was evaluated (Fig. 1) and indexed in categories according to Koza et al. (2009). The A-L biofilm was quantified in half of the tested strains. The pellicle was gently attached to the tube wall; the tube and the pellicle were then washed three times with water and fixed by air drying. At the next stage, we performed crystal violet staining of both types of biofilm and let the samples dry for 24hrs (Fig. 2). Subsequently, ethanol (1.5ml) was gently added to dissolve the dye from the submerged biofilm. The ethanol was pipetted directly to the bottom of the tube so as not to affect the A-L biofilm layer. The tubes remained at rest for 1 min, and the solution was discarded. After this phase, the only dye present was that attached to the A-L biofilm, and this dye was redissolved with ethanol. In order to compare the formation of submerged biofilm in the microtiter plates and in the test tubes, we converted the surface area to ethanol volume so that the relevant area in the microtiter well would correspond to the relevant area in the test tube. The ratio of the surface area was calculated to be 1:4.11; thus, the relevant volume in the tube was determined as 660μl (4.11 x 150μl). The tubes were vortexed for 30s to eliminate all the dye, and 150μl of the solution was transferred into the microtiter plate. All the samples were measured at OD595; each strain was tested three times.

Motility assays All motility types were evaluated by plate tests, with the medium solidified using different amounts of agar. The zones after the cultivation were measured and assessed. The swimming motility assay was carried out in a semi-solid Luria-Bertani (LB) medium (Oxoid, UK) with 0.3% of agar (Déziel et al., 2001). The inoculation was performed by inserting the tip of a sterile toothpick into the fresh medium, and the motility was assessed via diameter measurement (the zones were regular circles) after 20hrs of cultivation. Subsequently, the size of the growth zone was calculated. The plates for twitching motility contained a fresh LB medium solidified with 1% of agar. These plates were inoculated with a toothpick reaching to the bottom of the Petri dish and incubated at 37°C for 16hrs; then the incubation continued at This article is protected by copyright. All rights reserved.

Accepted Article

room temperature for another 72hrs (Zolfaghar et al., 2003). After the cultivation, the agar layer was removed, and the motility zone at the bottom of the Petri dish was fixed by air drying and stained with 2% crystal violet. Because of the irregular shape of the zones, the processing software ImageJ was used to facilitate the size assessment (Rasband, 2011). The medium applied for the swarming assay consisted of 0.5% of nutrient broth (Hi-Media, India), 0.005% of glucose, and 0.5% of agar. The plates were inoculated with a sterile toothpick contacting the top of the agar and incubated at 37°C for 48hrs (Deligianni et al., 2010). The surface zones were measured using ImageJ (Rasband, 2011). In all the motility assays, strains with the zone of < 20mm2 (the threshold value between colony and motility) were evaluated as non-motile. Assays for extracellular virulence factors Haemolysin production For the haemolytic activity assay, the method described by Rampioni et al. (2009) was utilized. The P. aeruginosa culture was grown in LB at 37°C; OD600 was adjusted to 3.0, and 1ml of the culture was mixed with 50μl of pre-washed sterile sheep erythrocytes. After 3hrs of incubation at 37°C, the debris was removed by centrifugation (20 min., 11,000 rpm, 20°C), and the released haemoglobin was evaluated by measuring OD545. The haemolysin-positive strains showed OD545 higher than the average of the OD545 of the three independent negative controls. LasB elastase production The elastolytic activity was determined by means of the Elastin Congo Red Assay (Hamood et al., 1996). The P. aeruginosa culture was grown in LB at 37°C; OD600 was adjusted to 3.0. Seven hundred microlitres of the culture was incubated with 20mg of Elastin Congo Red (Sigma-Aldrich, USA) in 250µl of Tris Buffer with 1mM CaCl2 at 37°C for the period of 17hrs. After centrifugation (20 min., 11,000 rpm, 20°C), OD492 was determined. The elastasepositive strains showed OD492 higher than the average of the OD492 of the three independent negative controls.

Siderophores production assay The secretion of extracellular pigments was detected via incubation using King’s A Agar (the production of pyocyanine and pyochelin) and King’s B Agar (the production of pyoverdine); both these media had been supplied by the Hi-Media company, India. The plates were cultivated at 37°C for 24hrs (Fonseca et al., 2008). The production of pyocyanine was

This article is protected by copyright. All rights reserved.

Accepted Article

confirmed by a colour change after acidification with 5µl 1M HCl dropped directly on the colony (Bachofen & Siegrist, 2008). Antibiotic susceptibility assay Antibiotic susceptibility was determined by the disc diffusion method (CLSI, 2011). The P. aeruginosa culture was diluted in sterile saline to OD600 of 0.5; the suspension was diluted a hundred times and plated on a Mueller-Hinton Agar (Oxoid, UK). The set of discs with antibiotics suitable for the treatment of P. aeruginosa infections was applied (30µg amikacin, 10µg gentamicin, 30µg ceftazidime, 75µg cefoperazone, 30µg cefepime, 105µg cefoperazone/sulbactam, 110µg piperacilin/tazobactam, 10µg imipenem, 10µg meropenem, 30µg aztreonam, 5µg ciprofloxacin, 5µg ofloxacin, 10µg colistin; Oxoid, UK). The plates were cultivated at 37°C for 20hrs; the inhibition zones were subsequently measured and compared with the zone diameter interpretive criteria for the used antibiotics. β-lactamases production assay Assays for the detection of extended spectrum β-lactamases (ESBL), AmpC, and metallo-βlactamase (MBL) were performed via a double-disc synergy test according to CLSI (CLSI, 2011). ESBL and AmpC β-lactamases (AmpC) were tested in ceftazidime and cefoperazoneresistant strains (N = 47); MBL was tested in imipenem and/or meropenem-resistant strains (N = 58). In 35 strains, all types of β-lactamases were tested. The presence of characteristic inhibition zone deformations was assessed. The production of MBL was kindly confirmed by Jaroslav Hrabak, Ph.D. (1st Faculty of Medicine, Charles University in Prague), who used a spectrophotometric assay with crude cell extracts. Statistical analysis The data were tested for normality where such testing was necessary. As most of the data were non-parametric, the complex statistical analysis was performed using non-parametric tests. The partial analyses of the data exhibiting normal distribution were performed by parametric tests. Correlations between the virulence factors within the group of blood and urinary tract isolates were analysed separately by the Statistica software for Windows 10.1 (StatSoft, Inc., 2012), using the Goodman-Kruskal correlation coefficient γ. The correlations attributed by p