A marked Copy of the Manuscriptinfluence of Oxyr on Susceptibility of Planktonic and Sessile Escherichia Coli Cultures to Ciprofloxacin and Cefotaxime

Alternation of bacterial antioxidant defense pathways might affect susceptibility to antibiotics in dual ways. Using a relatively simple model based on wild-type and oxyR Escherichia coli mature biofilms, their counterpart planktonic cultures and exponentially growing planktonic cultures, we explored the role of OxyR-mediated metabolism alternations in modulation of susceptibility to antibiotics ciprofloxacin and cefotaxime. All three types of cultures were placed in fresh medium,1 h after antibiotics were added and incubation continued further for 2 h. Killing rates of antibiotics were determined, biofilm eradication using crystal violet assay was estimated, expression of rpoS, katG, sulA genes as well as HPI and HPII catalase activity were measured. Biofilms of both strains were more recalcitrant to ciprofloxacin and cefotaxime at all tested concentrations compared to exponentially growing planktonic cultures. In oxyR biofilms killing rate of ciprofloxacin was lower, and killing rate of cefotaxime was higher compared to the parental strain. Compared to biofilms, wild-type biofilm counterpart planktonic cultures showed higher tolerance to low doses of ciprofloxacin, while oxyR plankton demonstrated higher tolerance to cefotaxime. Higher recalcitrance of oxyR biofilms to ciprofloxacin may be caused by an increase in persister cells under conditions of enhanced oxidative stress and activated SOS response.


Introduction
Ciprofloxacin-and cefotaxime-treated urinary tract infections are often associated with biofilmproducing bacteria such as Escherichia coli which demonstrated resistance to both antibiotics in 49.9 and 28% of described cases, respectively [1]. Promising strategies to combat multidrug-resistant and antibiotic tolerant microorganisms include alternation of central bacterial metabolism [2,3]. Since formation of reactive oxygen species (ROS) is a key trigger of stress response, modulation of antioxidant defense pathways might affect development of antibiotic tolerance/susceptibility. ROS including H 2 O 2 , have been proposed to explain antibiotic-induced bacterial cell death, regardless of primary antibiotic targets. [4] This subject remains a matter of intense debate, as other scientists published conflicting results, including those about the role of OxyR-controlled defenses in lethal activity of antibiotics [5][6][7].
OxyR factor stimulates transcription of genes involved in bacterial defense against peroxide stress, decreasing the levels of H 2 O 2 and unincorporated iron, thereby reducing DNA damage caused by Fenton chemistry and preventing cell death [8]. OxyR role in biofilm formation varied from stimulation [9][10][11] to inhibition [12,13].
Iraqi Journal of Science, 2020, Vol. 61, No. 11, pp: 2812-2819 2813 Here, we used an experimental model based on mature biofilms, their counterpart plankton and exponentially growing planktonic cultures of Escherichia coli strains to study effect of OxyR deficiency on susceptibility of E. coli biofilm and planktonic cultures to bactericidal antibiotics of different classes: fluoroquinolone ciprofloxacin and β-lactam cefotaxime.

Materials and methods Bacterial strains and growth conditions
E. coli strains BW25113 (wt) and JW3933 (oxyR) were from Keio collection [14]. The strains carrying transcriptional gene fusions katG::lacZ and rpoS::lacZ were constructed by transformation of the parental strain and the oxyR mutant with plasmids pKT1033 [15] and pRS415 katF5, respectively. The strains with sulA(sfiA)::lacZ fusions were created by P1 transduction from the strain DM4000 [16]. Bacteria were grown overnight in M9 minimal glucose (2 g/l) medium [17]. After centrifugation the cells were resuspended in fresh M9 medium (4 g/l glucose) supplemented with 0.2% casamino acids and 10 µg/ml thiamine to an initial optical density at 600 nm (OD 600 ) of 0.1. Wells of 96-well polystyrene microtitre plates were inoculated with 200 µl of culture prepared as described above and incubated at 37°C during 22 h to produce biofilms. All reagents for determination -galactosidase and catalase activity, antibiotics ciprofloxacin and cefotaxime, thiamine, casamino acids, agar, Luria-Bertani broth, were from Sigma-Aldrich Chemical Co (St Lous, MO, USA). Other reagents were of analytical grade (Reachim, Russia).

Antibiotic treatments
Mature biofilms were washed twice with 0.9% NaCl, 200 µl aliquots of fresh М9 (4 g/l glucose) medium supplemented with 0.2% casamino acids and 10 µg/ml thiamine were added in each well. The plates were incubated at 37°C with shaking (340 rpm) in Shaker Thermostat Sky Line (ELMI, USA) during 1 h, then antibiotics 0.03 µg/ml (2 MIC) and 0.3 µg/ml (20 MIC) ciprofloxacin or 1 µg/ml (12.5 MIC) and 10 µg/ml (125 MIC) cefotaxime were added and incubation continued further for 2 h. Mass biofilm formation and colony forming units (CFU) were determined at time zero and every hour of cultivation.
Two types of planktonic cultures were used. One (wt-16, oxyR-16) was derived from 16 h night culture prepared as described above for biofilm production, another (wt-22, oxyR-22) was derived from 22 h planktonic counterparts of the biofilms by its dilution with the fresh medium to OD 600 of 0.1. 200 µl aliquots of these cultures were placed in wells of polysterene microtiter plates, incubated for 1 h, treated with antibiotics and further incubated for 2 h. OD 600 was measured every hour of incubation using xMark™ spectrophotometer (Bio-Rad, USA) and samples for CFU determination were taken.

Biofilm formation assay
Mass biofilm formation (BF) was monitored using the modified crystal violet microplate biofilm assay [18]. Wells of 96-well polysterene microtiter plates containing mature biofilms were prepared as described above. Control wells contained bacteria-free medium only. At time zero and every hour of cultivation OD 600 of each well was measured, broth was removed and wells were rinsed twice with 200 µl of sterile saline. The wells were air-dried and 150 µl per well of 0.1% crystal violet solution was added for 30 min. Then, the colourant was discarded and the wells were rinsed five times with distilled water. The plates were air-dried for 1 h. To quantify biofilms, 200 µl of 96% ethanol was pipetted into each well. After 5 min, 125 µl of the solution was transferred to a separate plate where the OD 540 was measured using xMark™ spectrophotometer. BF was calculated as follows: BF = AB -CW, where AB is the OD 540 of stained attached bacteria and CW is the OD 540 of stained control wells.

Determination of killing rates of antibiotics
To compare bactericidal activity of antibiotics in different cultures, killing rates after 2 h exposure with antibiotics were calculated from the CFU data for each culture. The rate of antibiotic-induced bacterial killing (ψ) was calculated based on density decline of viable bacteria (CFU/ml) over a defined period [19]. Colony-forming ability (CFU/ml) was estimated as previously described [18]. Preliminary experiments revealed a decrease in plating efficiency of oxyR mutant on LB agar plates under aerobic conditions. Colony-forming ability of this strain was restored in the presence of catalase (500 U/ml) during serial dilutions and in plates with LB agar. Therefore, all colony-forming assays were performed with the addition of catalase.

Determination of β-galactosidase and catalase activity
We explored the role of OxyR, RpoS and SOS regulons in susceptibility to antibiotics measuring βgalactosidase activity in reporter strains carrying fusions of the genes katG, rpoS and sulA with the gene lacZ [17] using a SmartSpec Plus Spectrophotometer (Bio-Rad, USA). Aliquots of 340 µl of similar planktonic culture samples were collected from two similar wells and OD 600 was measured. Then 170 µl of these aliquots were transferred into two tubes each containing 690 µl of reducing buffer [17]. Then, 5 µl of sodium deoxycholate and 5 µl of toluene were added to each tube and incubated for 30 min at 37°C. Then 170 µl of 2-nitrophenyl-β-D-galactopyranoside (ONPG) were added to each test tube and incubated for 5 min. "Blank" tube was supplied with 170 µl of reducing buffer instead. The reaction was stopped by addition of 170 µl of K 2 CO 3 and OD 420 and OD 550 were measured. β-galactosidase activity was expressed in Miller units, calculated using the formula: (OD 420 -1.75 × OD 550 / ∆OD 600 × t) × 6000, where OD 420 and OD 550 -optical density of the samples, ∆OD 600 is the difference between OD 600 of bacterial culture and OD 600 of bacteria-free medium, tduration of exposition with ONPG, 6000coefficient taking into account dilution of the culture. In biofilms, broth was removed and the plate wells were rinsed to remove nonadherent bacteria. Fresh M9 was further added to each well and biofilms were sonicated by two 1 min pulses (37 kHz, 30W) with 1 min pause time between them in a water bath sonicator (Ultrasonic cleaning unit Elmasonic S10 H, Elma, Germany). OD 600 and β-galactosidase activity were measured as described above. HPI and HPII catalase activity in planktonic cultures was determined by the spectrophotometric method [20].

Statistical analysis of the data
Each result is indicated as the mean value of at least five independent experiments ± the standard error of the mean (SEM). Significant difference was analyzed by Student's t-test. A P-value of 0.05 was used as the cut-off for statistical significance. Results were analyzed by means of Statistica 6 (ver. 6, 2001; StatSoft Inc.).

Effect of oxyR deficiency on biofilm eradication in the absence and presence of antibiotics
Similarly to previously published data [9], oxyR mutant demonstrated an increase in biofilm formation after 22 h of static incubation compared to the wild-type strain. Biofilm formation, calculated as crystal-violet staining (OD 540 ), normalised to total bacterial growth (OD 600 ), was 0.160 ± 0.016 and 0.258 ± 0.016 for the wild-type strain (wt-bf) and the oxyR mutant (oxyR-bf), respectively. Subsequent incubation in fresh M9 without antibiotics resulted in significant biofilm eradication, although wt-bf retained 60% of the initial biofilm compared to only 25% in oxyR-bf ( Figure-1), apparently due to changes in the structure of biofilm and/or the signal transduction pathways associated with OxyR. Addition of 0.03 μg/ml of ciprofloxacin (CF) after 1 h incubation in fresh medium even stimulated biofilm formation by wt-bf, whereas a higher dose (0.3 μg/ml) had no significant effect. In contrast, cefotaxime (Cef) accelerated biofilm eradication in a dose-dependent manner (Fig. 1A). In oxyR strain, both doses of CF did not affect biofilm eradication, while both concentrations of Cef reduced OD 540 to the same extent (Figure-1B).
Biofilms of both strains were more recalcitrant to CF and Cef at all concentrations tested compared to the wt-16 and oxyR-16 planktonic cultures (Fig. 2 A,B). A distinctive feature of wt-22 was a higher tolerance to low doses of both antibiotics compared to wt-16 ( Fig. 2 C,D), which was in agreement with our previous findings that regardless of reasons causing changes of growth rate, an increase in the growth rate led to an increase in the ciprofloxacin-induced killing rate [21]. The functional OxyR protein was necessary for the development of this tolerance to low doses of CF, while the presence of the oxyR mutation was not significant in the case of Cef. Notably, killing rate of low CF doses in wt-22 (ψ = -0.08) was even lower than that in wt-bf (ψ = -1.31). Similarly, oxyR-22 was less susceptible to low doses of Cef (ψ = 0.47) than oxyR-bf (ψ = -1.57). Expression of rpoS gene was minimal in wt-16 and oxyR-16 planktonic cells (Figure-3A), which corresponded to high growth rates of these cultures. The level of expression of the rpoS gene increased 1.5-2 times in slow-growing wt-22 and oxyR-22 and more than 4-fold in wt-bf and oxyR-bf. Biofilms also demonstrated a higher level of sulA expression, indicating an increased expression of SOS regulon compared to planktonic cultures (Figure-3B). Earlier, recA and other SOS response genes were significantly induced in mature biofilms compared to exponentially grown planktonic cells [22]. Absence of oxyR increased sulA gene expression by 2.5 times in both planktonic cultures compared to the wild type strain (Figure-3B), revealing an increased oxidative stress and DNA damage in the oxyR mutant under these conditions. The expression of katG gene of the OxyR regulon, was not significantly different in both planktonic cultures, but it increased 2.1-fold and 2.7-fold in wt-bf and oxyR-bf compared to planktonic cells (Fig. 3C) which was consistent with the findings that cells in biofilms undergo endogenous oxidative stress [23]. The expression of katG can be controlled by both OxyR and RpoS transcriptional regulators [24]. Therefore, enhanced expression of katG gene in oxyR mutant was presumably caused by RpoS-controlled expression. Determination of catalase activity in planktonic cultures of the wildtype and oxyR strains confirmed the role of RpoS in induction of catalase HPI (encoded by katG), especially in oxyR-22 planktonic cells (Fig. 3D). Total catalase activity in these cells was higher than in wt-22 planktonic culture due to the increased activity of both HPI and HPII (encoded by RpoSregulated katE gene) catalases. Unfortunately, we were unable to measure catalase activity in biofilms because of the small number of cells.

Discussion
Deletion of oxyR was reported to increase the susceptibility of Klebsiella pneumoniae to antibiotics due to a decrease in expression of acrB, encoding a component of multiple drug resistance system AcrAB [25]. Here we showed that OxyR deficiency significantly influenced susceptibility to CF in biofilms and oxyR-22. The effect was opposite: oxyR-bf became more recalcitrant than wild-type biofilms, whereas oxyR-22 lost ability to withstand low doses of CF better than oxyR-16. Alternatively, Cef demonstrated increased killing efficiency in oxyR biofilms compared to wild-type biofilms, while oxyR-22 retained higher tolerance to low doses of antibiotic than oxyR-16. Specific growth rate of bacterial cultures contributes to susceptibility of E. coli to killing by antimicrobials [21] and is inversely proportional to intracellular concentrations of global regulator of gene expression guanosine tetraphosphate (ppGpp) and general stress response regulator RpoS. Typically associated with starvation and stress, RpoS controls many defensive genes, which overlap partially with those that are under OxyR control (katG, ahpCF, dps) [24]. ppGpp-mediated stringent response and RpoS can be involved in generation of highly tolerant bacterial persisters to antibiotics [26,27]. Here, both wt-22 and oxyR-22 grew 2 times slower and showed 1.5 and 2.4 times higher levels of rpoS expression than wt-16 and oxyR-16. Observed low growth rate and RpoS regulon induction might explain decreased lethal effect of low Cef dose in wt-22 and oxyR-22 and low CF dose in wt-22. Under these conditions, tolerance of wt-22 and oxyR-22 planktonic cultures was the same or even higher than those of the corresponding biofilms. Similarly, earlier studies showed that stationary cultures of Pseudomonas aeruginosa had the same or even higher tolerance to antibiotics than biofilms [28]. In case of CF, oxyR deficiency in oxyR-22 prevented an increase in tolerance, probably because of lack of some vital function controlled by OxyR. It does not seem to be katG gene coding for catalase HPI, since katG expression and catalase activity in oxyR-22 were even slightly higher than in wt-22. Significant increase in rpoS, sulA and katG expression in both wild-type and oxyR biofilms compared to wt-16 and oxyR-16 may indicate an increase in persister cells which led to a decrease of CF and Cef killing efficiency in wild-type biofilms. oxyR deletion resulted in a strong biofilm eradication that was accelerated by Cef, which could be the reason for the decrease in oxyR biofilm recalcitrance toward Cef compared to wild-type biofilms. By contrast, lack of oxyR may enhance endogenously induced oxidative stress in biofilms, thereby promoting formation of persister cells with participation of toxinantitoxin modules and the SOS response [23,27]. Eventually, an increase in tolerance to CF in oxyR biofilms might occur as we have observed. Data presented here highlight the need of in-depth exploration of involvement of bacterial antioxidant pathways in antibiotic tolerance.