Antimicrobial Pressure of Ciprofloxacin and Gentamicin on Biofilm Development by an Endoscope-Isolated Pseudomonas aeruginosa

This work aims at characterizing endoscope biofilm-isolated (PAI) and reference strain P. aeruginosa (PA) adhesion, biofilm formation and sensitivity to antibiotics. The recovery ability of the biofilm-growing bacteria subjected to intermittent antibiotic pressure (ciprofloxacin (CIP) and gentamicin (GM)), as well as the development of resistance towards antibiotics and benzalkonium chloride (BC), were also determined. The capacity of both strains to develop biofilms was greatly impaired in the presence of CIP and GM. Sanitization was not complete allowing biofilm recovery after the intermittent cycles of antibiotic pressure. The environmental pressure exerted by CIP and GM did not develop P. aeruginosa resistance to antibiotics nor cross-resistance towards BC. However, data highlighted that none of the antimicrobials led to complete biofilm eradication, allowing the recovery of the remaining adhered population possibly due to the selection of persister cells. This feature may lead to biofilm recalcitrance, reinforcement of bacterial attachment, and recolonization of other sites.


Introduction
Pseudomonas aeruginosa is an opportunistic pathogenic bacterium [1] widely investigated for its high incidence and extraordinary ability to form strong bio�lms in clinical equipment, medical devices, and wounds [2,3]. is microorganism is commonly associated with nosocomial infections and is a leading cause of severe and life-threatening infections, especially in immunosuppressed hosts [4].
P. aeruginosa is one of the most common microorganisms transferred by bronchoscopes, being the most frequent in gastrointestinal endoscopy [5]. Flexible endoscopes undergo repeated rounds of patient use and reprocessing. Studies related to endoscope contamination have reported the presence of bio�lms on the inner surface of endoscope channels [6,7], highlighting the importance of effective measures for cleaning and disinfection in endoscope reprocessing. Bio�lm removal is a crucial step to prevent lapses in reprocessing, being thus of clinical relevance in endoscopy [5,8]. Bio�lms represent a reservoir of pathogenic bacteria that can detach, resume their planktonic state, and contaminate new surfaces and patients. Moreover, microbial bio�lms are notorious for their high level of resistance towards antibiotic and biocide treatments [9]. Bacteria within bio�lms can easily live in the presence of high antibiotic concentrations similar to the ones that are prescribed during the course of therapies [10,11]. Bio�lm resistance mechanisms involve not only the reactiondiffusion limitation of antimicrobial access to the bio�lmentrapped bacteria [12,13] but also the expression of spatially heterogeneous, less susceptible phenotypes, caused either by growth as a bio�lm per se [14] or through the expression of high cell density [15] or starvation phenotypes [16]. Antibiotics and biocides are frequently used in hospitals with the purpose of controling the growth of, or to kill bacteria in, respectively, infection control and sanitation. e use of certain active substances in biocides in various settings may contribute to the increased occurrence of antibiotic-resistant bacteria. Cross-resistance between biocides and antibiotics and between different antibiotics has been reported previously [17], and there are several studies suggesting that if two antimicrobial compounds have similar mechanisms of action, they may also share resistance mechanisms [18]. It has also been demonstrated that highly antibiotic-resistant clinical isolates of Gram-negative bacteria are generally more resistant to disinfectants [19]. Although there is much concern regarding the risks of antibiotic resistance induced by the use of and resistance to biocides, there is a lack of studies evaluating the performance of disinfectants aer bacterial exposure to antibiotics. Studies related with the use of sessile bacteria to assess the efficacy of antibiotics and biocides are even lesser, even though bio�lm formation is an important aspect of many bacterial diseases. ese bio�lm tests should also include bacteria isolated from real scenarios as they can present genetic diversity and thus possess distinguishing virulence factors. e pathogenesis of P. aeruginosa is attributed to the production of several cellassociated and extracellular virulence factors that arise under certain environmental conditions [20].
In the present work, the phenotype (early-stage adhesion, bio�lm formation, and sensitivity to antimicrobials) of P. aeruginosa isolated from a bio�lm formed on an endoscope was determined and compared with the reference strain. Furthermore, this study was also undertaken to determine whether exposure of P. aeruginosa bio�lms to intermittent cycles of antibiotic chemotherapy (cipro�oxacin (CIP), and gentamicin (GM)) could lead to regrowth and potential resistance and cross-resistance towards a disinfectant, BC, CIP and GM.

Test Organisms and Culture
Conditions. P. aeruginosa (ATCC 10145) (PA) and P. aeruginosa isolated (PAI) from a bio�lm formed in a medical device (gastrointestinal endoscope) were preserved at −80 ± 2 ∘ C in 10% glycerol stocks. Prior to each experiment, bacterial cells were grown on Tryptic Soy Agar (TSA, Merck) plates for 24 h, at 37 ∘ C.
To prepare the bacterial suspensions, one colony of each strain of P. aeruginosa (PA or PAI) was collected from the TSA plates and grown in Tryptic Soy Broth (TSB) for 24 h at 37 ∘ C, with agitation. Subsequently, bacteria were harvested by centrifugation and washed twice with sterilized ultrapure water (UP). Standardized cell suspensions were prepared in TSB unless otherwise stated.

Antibacterial and Antibiotic
Agents. Cipro�oxacin (CIP), a broad-spectrum synthetic chemotherapeutic antibiotic of the �uoroquinolone drug class, used clinically to treat P. aeruginosa infections, was purchased from Fluka. Gentamicin (GM), an aminoglycoside antibiotic, used to treat many types of bacterial infections, particularly those caused by Gram-negative organisms, was purchased from Sigma. Benzalkonium chloride (BC), a quaternary ammonium compound, widely used in clinical disinfectant formulations, was obtained from Calbiochem. e antibiotics concentrations used in the present work were determined using as reference the 3x MBC [21] of each product for the reference strain. So, for bio�lm disturbance cycles and bio�lm treatment, a concentration of 3 mg/L of CIP and 10 mg/L of GM were used. Concerning BC, the MIC concentration of 360 mg/L was used.

Early Bacterial Adhesion.
In order to determine the adhesion ability of both strains, a parallel plate �ow chamber (PPFC) and image analysis system as described by Sjollema et al. [22] were used to study the early-stage bacterial adhesion and detachment. Brie�y, the PPFC consists of a nickel-coated frame measuring 16 × 8 × 1.8 cm. Te�on spacers were placed between the plates, to separate them by 0.06 cm.
e PFFC device was mounted in a phase contrast inverted microscope (Diaphot 300; Nikon) equipped with a 40x ultralong working distance objective. e images were acquired in a CCD camera (AVC, D5CE; Sony) connected to the microscope and coupled to an image analyser (Image Proplus 4.5; Media Cybernetics).
Prior to each experiment, all tubes and the �ow chamber were �lled with sterile phosphate buffered saline (PBS), taking care to remove air bubbles from the system. To assess the rate of attachment of reference strain and isolated P. aeruginosa on polystyrene (PS), bacterial suspensions were put to circulate through the PPFC at 0.020 mL/s for 30 min to allow surface colonization under �ow. en, PBS was circulated through the equipment, at the same rate, to remove unattached and weakly adhering cells and thus to evaluate the detachment of the bacterial cells.
e initial increase in the number of adhering microorganisms with time was expressed as the so-called initial deposition rate [ 0 , cells/(cm 2 s)], that is, the number of adhering microorganisms per unit area and time. e number of adhering microorganisms aer 30 min of bacterial suspension �ow [ 30 min , cells/(cm s)] and the number of microorganisms aer PBS passage [ 60 min , cells/(cm s)] were also determined. e rate of detachment (%) denotes the percentage of P. aeruginosa cells that were detached upon the passage of PBS trough the �ow chamber.
2.�. Bio�lm �ormation. e bio�lm formation ability of PA and PAI on polystyrene (PS) was inspected along time (24,48,72, and 96 h) using the microtiter plate test developed by Stepanović et al. [23]. Brie�y, cell suspensions of both strains were diluted to obtain a �nal concentration of 1.0 × 10 7 cfu/mL. Aerwards, 200 L/well of the bacterial suspension was transferred to sterile 96-well �at-bottom tissue culture plates (�range Scienti�c). All the plates were incubated at 37 ∘ C, during 24 h for bio�lm development, with agitation.
Aer 24 h of bio�lm growth, the supernatant containing planktonic cells and media was removed. e wells were re-�led with fresh TSB, and this process of supernatant removal and media �lling was repeated for every 24 h until 96 h of bio�lm formation.
Aer 24, 48, 72, and 96 h of growth, bio�lms were characterized in terms of biomass, respiratory activity, and number of viable bio�lm-entrapped cells.

2.�. �io�lm Distur�ance and �eco�ery.
To assess whether the presence of antibiotics could interfere with the establishment of bio�lms by both strains on PS surfaces, bio�lms were allowed to form for 24 h in the presence of both antibiotics. To ascertain the postantibiotic effects (PAEs) of CIP and GM on P. aeruginosa, those challenged bio�lms were later sub�ected to intermittent cycles of antibiotic chemotherapy with CIP and GM.
Bio�lms were formed in microtiter plates with cell suspensions of both strains at a �nal concentration of 1.0 × 10 7 cfu/mL prepared in TSB containing CIP or GM in a �nal concentration of, respectively, 3.0 mg/L or 10 mg/L. Each cycle of antibiotic treatment was followed by a recovery period of 24 h, where bio�lms only developed in 200 L of fresh TSB.
Aer 24, 48, 72, and 96 h the content of each well was removed and bio�lms were washed and phenotypically characterized.
2.�. �io�lm Susce�ti�ility. e 96-hour-old bio�lms untreated and submitted to the intermittent antibiotic chemotherapy with CIP and GM were inspected regarding their susceptibility towards the same antibiotics as well as the potential occurrence of cross-resistance towards the disinfectant BC.
In order to determine the bio�lm response aer antibiotic therapy, bio�lms were treated with 200 L per well of 360 mg/L of BC, 3 mg/L of CIP or 10 mg/L of GM for 30 min. Nontreated wells were �lled with 200 L of UP sterilized water. Aer that, the content of each well was removed and bio�lms were washed with 200 L with ultrapure sterilized water (UP) being reserved for posterior analysis.

2.�. �io�lm Analyses
2.�.1. �io�lm Mass and Acti�ity. Biomass of P. aeruginosa bio�lms was quanti�ed by crystal violet (CV) staining method adapted from Stepanović et al. [23]. For that, the plates containing the bio�lms were le to air-dry for 30 min, and 200 L of 98% methanol was transferred to each well in order to �x the remaining attached bio�lm, for 15 min. Aerwards, the plates were emptied and le to air dry. e �xed bacteria were stained with 200 L of 1% (w/v) CV (Gram's staining; Merck) per well, for 5 min. Aer this staining step, plates were washed with running tap water, air dried, and �lled with 200 L of 33% (v/v) of acetic acid (Merck) in order to resolubilize the CV bound to the adherent bacteria. e quantitative analysis of bio�lm production was performed through the measurement of optical density at 570 nm (OD 570 ), of each well using a microtiter plate reader (Bio-Tek Synergy HT, Isaza), the bio�lm mass being presented as OD 570 . Control experiments to avoid false results were also performed. When the optical density was higher than 1.0, samples were diluted with 33% (v/v) of acetic acid. For each condition tested, eight different wells were used to quantify the mass of bio�lms.
Bio�lm activity was determined with 2,3-bis(2-methoxy-4-nitro-5-sulfo-phenyl)-2H-tetrazolium-5-carboxanilide (XTT) colorimetric method as described by Stevens and Olsen [24], with some modi�cations. Aer bio�lm development and treatment, 200 L of a combined solution of XTT (Sigma) and PMS (phenazine methosulfate) (Sigma) was added to each well in order to obtain a �nal concentration of 150 mg/L of XTT and 10 mg/L of PMS. Aer that, plates were incubated at 37 ∘ C for 3 h, with agitation, in the dark. Bio�lm activity was determined through measurement of the optical density at 490 nm of the liquid content of each well using a microtiter plate reader, bio�lm activity being presented as OD 490 . Control tests, were also carried out, in order to avoid misleading results. For each condition tested, 8 different wells were used to determine bio�lm activity.

2.�.2. �io�lm �ell �numeration.
In order to determine the number of bio�lm-entrapped viable bacteria, bio�lm suspensions was prepared. Two-hundred microliters of UP sterilized water being were added to each well, the wellsattached bio�lms removed by ultrasonic bath in a Sonicor SC-52 (Sonicor Instruments) operating at 50 kHz, during 6 min. Aerwards, the bacterial suspensions of each 5 wells per condition were collected and gently vortexed for two min to disrupt possible cell aggregates (these parameters were previously optimized in order to promote the complete removal of all the bio�lm-attached cells without lysis). Bacterial suspensions were serially diluted, plated on TSA, and incubated at 37 ∘ C in an aerobic incubator for 24 h. e number of colony forming units (cfu) was enumerated, being the bio�lm cell numbers presented as log 10 (cfu/cm 2 ).

Statistical
Analysis. Statistical analysis was performed using GraphPad Prism, version 4.0 soware for Macintosh. Normality of data distribution was tested by the Kolmogorov-Smirnov method. Statistical signi�cance values of the groups' means of bio�lm mass, bio�lm activity, and cell number were evaluated using a one-way analysis of variance. Subsequent comparisons were performed using Tukey's post hoc test. Two-way analysis of variance with Bonferroni post hoc test were used to compare means of bio�lms obtained aer 96 h, and aer CIP and GM regrowth cycles aer treatment. e statistical analyses performed were considered signi�cant when 0.05.

Attachment and Detachment
Monitoring. e results obtained with early bacterial adhesion assay (Table 1) revealed that both strains had high ability to adhere to surfaces, but the rate of cell deposition ( 0 ) of the isolated strain was higher than that of PA ( 0.05). Aer 30 min of adhesion, the number of adhered cells ( 30 min ) is about twice for the isolated strain ( 0.05). Also, the number of PAI cells that remained attached in the PS surface ( 60 min ) was the double of PA cells ( 0.05), the percentage of detachment of PA cells aer PBS passage being of about 20%.

�io�lm �ormation.
In order to examine the bio�lm formation ability of both PAI and PA strains, the bio�lm phenotype was characterized in terms of mass, activity and 4 ISRN Biotechnology T 1: Initial deposition rate( 0 ), number of adhered cells, ( 30 min ), number of adhered cells aer PBS passage ( 60 min ), and percentage of detachment determined through the parallel plate �ow chamber. �alues are means ± SD for three measurements. 0 (10 3 cells/(cm 2 s)) 30   number of cells aer 24, 48, 72, and 96 h of growth ( Figure  1). In general, data showed that mass and activity of bio�lms increased along time, whereas the number of bio�lmentrapped cells was approximately in the same magnitude, for all the time periods of bio�lm formation. Comparing both strains, Figure 1 shows that, in general, PAI gave rise to bio�lms with more mass than PA ( 0.05) and activity ( 0.05). e number of bio�lm cells (Figure 1(c)) was identical to those �uanti�ed for PA bio�lms, except for 72-hour-old and 96-h-old bio�lms in which there was an increase in the number of cells ( 0.001). ese results indicate that both strains are good bio�lm producers although the isolate stands out relatively to the collection.
�.�. �io��� �is�ur�an�e an� �e�o�er�. e presence of CIP and GM in the �rst 24 h of bio�lm development clearly hampered the establishment of bio�lms by both strains on PS surfaces (Figure 2). In fact, the phenotype of PA and PAI bio�lms grown for 24 h under antibiotic pressure was characterized by a large decrease in bio�lm mass and activity (about 95%) and a reduction of about 4 log in the number of viable bio�lm-entrapped cells. However, it must be emphasized that a considerable number of cells remained viable on the surfaces. ese data revealed that CIP and GM have a signi�cant in vitro anti bio�lm formation activity, this effect being similar for both strains.
Aer a recovery period of 24 h, where growth occurred in absence of antibiotics, those remaining less dense bio�lms recovered its levels of biomass, activity and number of cells ( Figure 2). Bio�lm recovery aer CIP pressure ( Figure 2I) gave rise to PAI bio�lms with higher values of biomass ( 0.001), activity, and viable cells ( 0.01), when compared with PA bio�lms. is trend was similar to that observed in the 48-hour-old bio�lms formed without any stress factor (Figure 1). Regarding the post-GM effect ( Figure 2II), the superiority of the bio�lms formed by the isolated strain is no longer evident as PAI bio�lms only showed higher biomass ( 0.001). e second cycle of bio�lm growth under antibiotics pressure clearly reduced the mass, activity, and number of bio�lm-encased cells, for both strains (Figure 2). However, these reductions were lower than those obtained aer the �rst cycle of antibiotic treatment, mainly when GM was used. ese results showed that both antibiotics have ability to disturb established P. aeruginosa bio�lms causing its removal and inactivation. Nevertheless, as Figure 2 shows, this sanitation was not total, allowing bio�lm regrowth during the second recovery period. In fact, the resulting 96-hour-old bio�lms recuperated again its levels of biomass and activity, although they are far from those observed in bio�lms developed by both strains in the absence of antibiotic stress (Figure 1). e numbers of viable bio�lm-cells were also restored reaching, however, values in the same order of magnitude of those determined in the 96-hour-old bio�lms formed in TSB. Comparing the behaviour of both strains, in general, PAI bio�lms showed higher biomass, activity and number of cells than the bio�lms formed by the reference strain. e postantibiotic effects observed aer the second cycle of antimicrobial treatment is similar to that observed aer the �rst bio�lm growth under antibiotic pressure, except for bio�lm activity (Figure 2). In fact, the activity of the 96hour-old bio�lms was higher than those observed aer the �rst 24-h recovery period, specially for bio�lms grown under GM pressure ( 0 00 ) and for those developed by PAI ( 0 00 ) (Figure 2(b)). e overall results highlighted that both antibiotics have good anti-bio�lm characteristics and ability to remove and inactivate established P. aeruginosa bio�lms. Nevertheless, sanitization was not complete allowing the resumption of the bio�lms immediately following antibiotic pressure.
�.�. �io��� �us�e��i�i�i��. e susceptibility of the 96-h-old bio�lms, sub�ected to the intermittent cycles of antibiotic pressure, towards antibiotics (GM and CIP) and biocide (BC) treatment can be observed in Figure 3. In the range of conditions tested, the 96-h-old bio�lms formed by both strains in TSB were practically tolerant to the action of antibiotics and susceptible to the toxic effect of BC. In fact, only treatment with BC of PA and PAI bio�lms promoted a signi�cant reduction of biomass (Figure 3(a)) ( 0 00 ), respiratory activity (Figure 3(b)) ( 0 00 ), and number of viable cells (Figure 3(c)) ( 0 00 ).

ISRN Biotechnology
e 96-hour-old bio�lms formed by both the ATCC and the isolated strain under intermittent cycles of CIP or GM pressures were also clearly disturbed by the action of BC. It appears that antibiotic pressure during bio�lm growth gave rise to bio�lms more susceptible to the antimicrobial action of benzalkonium chloride.
e response of PA bio�lms towards CIP and GM treatment depended on the antibiotic used. e PA bio�lms challenged by CIP pressure (Figure 3I) are practically indifferent to the posterior aggression with the same antibiotic, as the values of bio�lm mass (Figure 3(a)I) and activity ( Figure  3(b)I) were higher than those observed in bio�lms without treatment ( ), although there was a reduction in the number of viable bio�lm-cells ( ) (Figure 3(c)I). e response of these bio�lms to the action of GM was somewhat different, as GM caused the reduction of bio�lm metabolic activity and number of bio�lm-entrapped cells.
e 96-h-old bio�lms formed under GM intermittent pressure appeared to be more tolerant towards both antibiotics as the aggression of CIP and GM for 30 min only decreased the number of viable bio�lm cells ( ) (Figure 3(c)I).
Regarding PAI bio�lms ( Figure 3II), those developed under CIP pressure were more sensitive to the action of GM than CIP since only signi�cant reductions in bio�lm mass ( ) (Figure 3(a)II) respiratory activity ( ) (Figure 3 ). Based on Figure 3, it can be stated that none of the conditions led to complete sanitation of the bio�lms, with, in general, the action of the antimicrobials being more effective in bacteria inactivation than in bio�lm removal. �ata also highlighted that cross-resistance between antibiotics and the biocide did not occur.

Discussion
Adhesion and bio�lm formation are two important aspects of many bacterial diseases, especially those related with medical devices [25], as �exible endoscopes. �hen bio�lms are identi�ed as the main cause of infection, treatment becomes very difficult since bacteria within bio�lms adopt special features that confer them increased resistance to antimicrobial agents [9]. is resistance usually makes sessile microorganism more difficult to kill and remove from surfaces than planktonic counterparts. Furthermore, in many cases incomplete removal of the bio�lm allows it to quickly return to its equilibrium state.
In this work, some phenotypic characteristics (early-stage adhesion, bio�lm formation ability and sensitivity to different antimicrobials) of P. aeruginosa isolated from an endoscope were inspected. e ability to adhere of the isolated strain was superior to that of reference strain as and 3 min are around two-fold higher than those observed with the reference strain. Besides revealing higher rate of adhesion ( ), the isolated strain also showed higher number of cells adhered to the surface aer 30 min of contact (Table 1). Furthermore, the strength of adhesion of this bacterium on PS surfaces was stronger as PBS circulation failed on cell detachment. In fact, for the reference strain the detachment of cells was around 20%, while for the isolated strain no signi�cant cell detachment was observed. �nowing that PAI was obtained from a real bio�lm formed on an endoscope, it is conceivable to speculate that the isolated strain has been exposed to mechanical and chemical stress conditions, namely, during endoscope reprocessing. So, as a survival strategy, this isolated strain may have acquired phenotypic and physicochemical changes that allowed it to adhere easily on PS and with superior strength. In fact, other authors [26] reported that organisms isolated from any given niche, medical, environmental, or industrial, have different mechanisms of adhesion and retention, mainly due to changes in their structural components, such as pili, �mbriae, and adhesive surface proteins that have adapted differently over time through selective pressures. Furthermore, exposure to antimicrobials may as well induce changes in cell surface hydrophobicity and surface charge that can alter bacterial adhesion properties [27].
e prominence of PAI was also visible in terms of bio�lm formation ability as it has developed bio�lms with more biomass, respiratory activity and number of cells than PA (Figure 1). is feature together with the greatest capacity to adhere allows to speculate that this isolated strain is more pathogenic than ATCC strain, as bacterial attachment and bio�lm formation are considered important virulence factors of bacterial pathogens [20,28]. In fact, the formation of thick bio�lms gives bacteria, amongst other advantages, protection from external aggressions, as host defences and antimicrobials, due to the lipopolysaccharides that constitute the EPS matrix [29]. is EPS matrix acts also as a diffusion barrier that can reduce antimicrobial efficacy by diminishing its penetration into the deeper layers of the bio�lm. Protected within this niche, bacteria can detach, proliferate and furthermore disseminate in large amounts making possible the spread of pathogens [30].
Although the isolated strain develops thicker bio�lms, it must be referred that luckily the capacity of both strains to develop bio�lms was greatly impaired in the presence of antibiotics. e concentration of antibiotics used to cause antimicrobial stress (3 mg/L of CIP and 10 mg/L of GM) was similar to those referred to in the literature [31] to have high bactericidal activity (16 x MIC or 3 x MBC); however, it must be highlighted that a substantial number of viable cells remained adhered on PS surfaces. CIP is known for being initially very effective against P. aeruginosa. GM is also used to control P. aeruginosa growth and has been described in several works as a potential antibiotic to treat bio�lm associated infections [32][33][34]. However, the diminished bio�lm-forming capacity shown by PA and PAI under CIP and GM pressure may be related with other actions than bactericidal activity. Bio�lm formation, by P. aeruginosa is hypothesized to follow a developmental pattern involving essentially four steps [35]: surface attachment, irreversible attachment, microcolony formation and differentiation into a mature population encased in a polymeric matrix. e presence of the antibiotics during bio�lm formation may have interfered in the transition from reversible bindings to stable and irreversible interactions [36], affecting the transition from microcolonies to bio�lms and thus delaying the mature bio�lm development [37]. With the cessation of the antibiotic pressure, those less dense bio�lms resumed their developmental process and gave rise to thicker bio�lms, with again the PAI bio�lms being superior to those formed by the reference strain. During endoscopy procedure, the external environment surrounding the medical device provides optimal conditions for microbial adhesion and bio�lm growth [6,7]. If 8 ISRN Biotechnology the disinfection procedures implemented during endoscopy reprocessing are not fully effective, bio�lms may form and persist. ese bio�lms can later release bacterial cells that can spread to other locations, contaminating new surfaces and infecting the patients that underwent endoscopy. A vicious circle of bio�lm growth, antimicrobial treatment, partial killing or inhibition of some susceptible population and regrowth of resilient cells can thus be created. is vicious circle was clearly observed in Figure 2, when bio�lms obtained a�er the �rst period of PAE were submitted to a new cycle of antibiotic pressure with, a decrease in their density and activity being observed again. e number of bio�lmentrapped cells has also decreased but a substantial number of cells remained viable and adhered, mainly when GM was used. is latter event may be related with GM difficulties to diffuse across the bio�lm matrix [38], limiting thus its access to the bio�lm-growing cells. To augment its efficacy it has been recommended in-use concentrations of GM higher than those used to other antibiotics. Conversely, cipro�oxacin is known for its ability to penetrate rapidly [39]. Based on these data, it can be concluded that both antibiotics have great ability to disrupt established bio�lms but poor capacity to completely inhibit bio�lm-growing bacteria. Furthermore, when the level of antibiotics dropped, in the PAE periods, the population of adhered cells was able to multiply and to repopulate the bio�lm observed during the second recovery period when the remaining bio�lms recovered its metabolism becoming thicker.
is cycle of bio�lm decrease�bio�lm recovery may be explained by the existence, within the entire P. aeruginosa bio�lm population, of a subpopulation of dormant cells that survive antibiotic treatment ensuring population survival. ese cells that do not grow in the presence of an antibiotic, but neither do they die, are known as persister cells [40]. is subpopulation is recognized as "drug tolerant" as it remain metabolically inactive in stressful conditions, but it can resort to normal growth rates and susceptibility in the absence of antibiotic. e data gathered in this study allows speculating that bio�lm development in the presence of antibiotics can be problematic as antimicrobial pressure can select persister cells and encourage bacterial adhesion and bio�lm development. is resilient sessile population, that is normally able to sustain an antimicrobial attack, could account for the prevalence of bio�lm-associated infections and for recalcitrance of surface contamination [40,41]. e environmental pressure exerted by CIP and GM did not contribute to the development of P. aeruginosa tolerance to the same antibiotics. In fact, the 96-hour-old bio�lms developed only in TSB are practically tolerant to the toxic action of CIP and GM� however, those bio�lms formed under antibiotic selective pressure are to a certain extent susceptible to the same antibiotic attack. e sensitivity to the antibiotics of the bio�lms formed by the isolated strain is similar to that displayed by the bio�lms developed by reference strain. Based on this evidence, it can be referred that the environmental stresses to which the isolated P. aeruginosa have been submitted during endoscope reprocessing did not cause the development of a resistant phenotype towards the antibiotics studied. It is accepted that antimicrobial selective pressure may result not only in selection of persister cells but also in the development of cross-resistance towards other antimicrobials. In this study, data showed that BC was quite effective against P. aeruginosa bio�lms, despite not having caused complete sanitation. All P. aeruginosa 96hour-old bio�lms developed or not under antibiotic pressure were similarly susceptible to BC attack ( Figure 3). ese data highlighted that bio�lm-growing bacteria subjected to CIP and GM pressure did not exhibit cross-resistance to benzalkonium chloride. e use of this cationic surfactant is not advised in endoscope washing procedures, but this product is still used in clinical practice for surface disinfection, antisepsis preservation, and cleaning [42].
e exposure of cells within the bio�lm to antibiotics pressure did not further promoted antimicrobial resistance to any of the antimicrobials tested ( Figure 3). However, it must also be referred that none of the conditions caused complete sanitation of the bio�lms, with, in general, the action of the antimicrobials being more effective in bacteria inactivation than in bio�lm removal. is fact is of upmost importance as cells may then detach from the remaining bio�lms and disseminate infection elsewhere.
e antimicrobial treatment of bacterial bio�lms may lead to eradication of most of the susceptible or metabolically active population but again the small fraction of persister cells or bacteria in the deeper bio�lm layers that are just exposed to subinhibitory concentrations can survive and be able to reconstitute the bio�lm a�er discontinuation of antimicrobial therapy [29].
Since bio�lms do not develop or mature in stress conditions but rather maintain a remaining adhered population and are exceptionally complex to eradicate, they are considered recalcitrant [38]. e persister cells give rise to a new diverse bio�lm community with high genetic variability. is "new" bio�lm is not as persistent as the cells that where in its foundation, being on the contrary as sensitive to external aggressions as a bio�lm developed in normal conditions [38]. e persister cell role in bio�lm survival will undoubtedly drive the effort to understand the mechanisms of their remarkable recalcitrance [43].
In this study, it was showed that the isolated endoscope strain possesses the ability to adhere in higher extent than the reference strain, developing a�er thicker bio�lms. Its increased ability to adhere may be due to its previous stress exposure to cleaning agents and disinfection procedures. Moreover, bio�lm development in the presence of high doses of antibiotics might lead to the eradication of the most part of the bio�lm population, selecting just a small fraction of persister cells, which can survive, being able to reconstitute the bio�lms following discontinuation of antibiotic therapy. is may represent an increased risk of infection to patients, requiring careful surveillance. As persister cells, which survive within the bio�lm a�er treatments, can develop new bio�lms and recolonize other accessible sites, the antibio�lm efficacy of a cleaning agent or antibiotic should not be just related with the reduction of bio�lm mass or number of cells, but depends largely of its ability to kill all bio�llm-cells, and also to completely eradicate bio�lms from surfaces.