BBRC Biochemical and Biophysical Research Communications 317 (2004) 605–609 www.elsevier.com/locate/ybbrc
Oxidative stress involved in the antibacterial action of different antibiotics Ines Albesa,* M. Cecilia Becerra, Paola C. Batt an, and Paulina L. P aez Departamento de Farmacia, Facultad de Ciencias Quımicas, Universidad Nacional de Cordoba, Haya de la Torre y Medina Allende, Ciudad Universitaria, 5000 Cordoba, Argentina Received 4 March 2004
Abstract Staphylococcus aureus and Escherichia coli sensitive to chloramphenicol incubated with this antibiotic suffered oxidative stress with increase of anion superoxide (O 2 ). This reactive species of oxygen was detected by chemiluminescence with lucigenin. S. aureus, E. coli, and Enterococcus faecalis sensitive to ciprofloxacin exhibited oxidative stress when they were incubated with this antibiotic while resistant strains did not show stimuli of O 2 . Other bacteria investigated was Pseudomonas aeruginosa, strains sensitive to ceftazidime and piperacillin presented oxidative stress in presence of these antibiotics while resistant strains were not stressed. Higher antibiotic concentration was necessary to augment O 2 in P. aeruginosa biofilm than in suspension, moreover old biofilms were resistant to oxidative stress caused by antibiotics. A ceftazidime-sensitive mutant of P. aeruginosa, coming from a resistant strain, exhibited higher production of O 2 than wild type in presence of this antibiotic. There was relation between antibiotic susceptibility and production of oxidative stress. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Oxidative stress; Staphylococcus aureus; Enterococcus faecalis; Escherichia coli; Pseudomonas; Chloramphenicol; Ciprofloxacin; Ceftazidime; Piperacillin; Biofilm
Diverse substances affect the reactive species of oxygen (ROS) produced by cells during oxidative processes. The alteration of normal levels of ROS can cause injury by oxidative stress raising lesions in different tissues, for example, fluoroquinolones can stimulate the production of ROS in human cells. These antibiotics possess toxicity for skin and generate free radicals, this harmful effect was described as phototoxicity and was associated to the increased singlet oxygen and superoxide anion (O 2 ) [1]. The effects of chloramphenicol, ciprofloxacin, ceftazidime, and piperacillin have been well studied during several years; however, the antibiotic action has not been related with oxidative stress in bacteria. Important research is aimed to understand, at molecular level, the mechanism of antibacterial action in different microorganisms. Recently, an antecedent indicated that the effect of ciprofloxacin involves an alteration of O 2 production in S. aureus [2], this oxidative alteration * Corresponding author. Fax: +54-351-433-41-63. E-mail address:
[email protected] (I. Albesa).
0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.03.085
being studied by chemiluminescence (CL). This method was employed to detect oxidative stress in diverse cells and used to evaluate O 2 in bacteria [3] but CL was not applied to investigate the action of antibiotics. Escherichia coli is a microorganism that can react efficiently in presence of substances that alter the levels of O 2 , due to its great superoxide dismutase SOD activity [4], resulting in being attractive to investigate the capacity of ciprofloxacin to induce oxidative stress in this bacteria. Moreover, the study of oxidative stress in E. faecalis treated with this antibiotic could be convenient because this bacteria produces high amount of O 2 [5]. P. aeruginosa must also be selected because it is a non-fermentative species and is a problematic organism, as a consequence of its resistance to antibiotics. This bacteria gives persistent colonization of lungs and might exist as biofilms, which are structured communities frequent in patients with cystic fibrosis (CF). P. aeruginosa uses extracellular quorum-sensing (QS) signals to coordinate biofilm formation showing phenotypic variants that exhibit antibiotic-resistant and generate persistent
606
I. Albesa et al. / Biochemical and Biophysical Research Communications 317 (2004) 605–609
infections [6,7]. Bacteria growing in a biofilm express resistance factors in a different form than planktonic cells. In general, the explanation was accepted that the chemical structure of exopolysaccharide and the architecture of biofilm contribute to the exclusion of antibiotic agents [8]. In addition, it was observed that the antibiotics present little efficiency; principally when biofilms are old, young sessile cells in immature biofilms resulted in becoming more susceptible to antibiotic therapy [9]. Certainly, it was observed that microbial surface-attached communities present high resistance to diverse types of stresses [10]. Consequently, these observations induced us to study by CL the differences between bacteria suspension and biofilm during the incubation with antibiotics. The aim of this investigation was to study the possibility of generating oxidative stress in four bacteria species treated with different antibiotics, as a tool to know whether they can cause oxidative stress, comparing attached bacteria in a biofilm with free bacteria, trying to explain new aspects of antibiotics’ actions.
Materials and methods Nitroblue tetrazolium reduction. Bacteria suspensions in 0.1 mL of Hanks’ balanced salt solution (106 cfu/mL) were incubated with one of the following antibiotics: chloramphenicol (0.5–8 lg/mL); ciprofloxacin (0.5–16 lg/mL); ceftazidime (0.5–80 lg/mL) or piperacillin (0.5– 80 lg/mL); then 0.5 mL of 1 mg/mL NBT was added and incubated for 30 min at 37 °C. The reaction was stopped with 0.1 ml of 0.1 M HCl and the tubes were centrifuged at 1500g for 10 min to separate cells from supernatants. The bacteria pellets were treated with 0.4 mL dimethyl sulfoxide (DMSO) to extract the reduced nitroblue tetrazolium (NBT) and then 0.8 mL of HBSS was added. Formazan blue obtained from cells was measured at 575 nm. Chemiluminescence of bacteria treated with antibiotics. Pseudomonas aeruginosa, S. aureus, E. coli, and E. faecalis were cultured in trypticase soy agar (TSA) for 24–48 h at 37 °C. Bacteria absorbance was adjusted to 1.0 OD600 with TSA, and 0.1 mL (107 cfu/mL) was incubated with 0.1 mL of 0.145 mM N,N0 -dimethyl-9,90 -bisacridinium dinitrate (lucigenin) and one of the following antibiotics were added at different concentrations: (a) 0.5–8 lg/mL chloramphenicol; (b) 0.5– 16 lg/mL ciprofloxacin; (c) 0.5–80 lg/mL ceftazidime or (d) 0.5–80 lg/ mL piperacillin. Then 100 lL of DMSO was added, as described in NBT assay. Normal or spontaneous production of ROS by each bacteria strain was determined and the results were compared with thse obtained in presence of antibiotic. Determinations by CL were performed at room temperature in a BioOrbit chemiluminometer mod 1253. The light emission results were expressed as relative light units (RLU) at different times in seconds, with subtraction of the background. Investigation of biofilm formation. Pseudomonas aeruginosa strains were studied in their capacity to form biofilms. Adhesion was investigated at 24, 48, and 72 h by optic microscopy and Gram stain to determine the amount of bacteria/l2 attached on slides placed vertically in tubes. Slime evaluation by staining. Slime formation was assayed in tubes of glass incubated with bacteria in TSB with 0.1% glucose. After 48 h the cultures were discarded and tubes were stained with safranin. Slimes were then detached with 50% acetic acid and absorbance at 470 nm was determined by spectroscopy.
Effect of piperacillin and ceftazidime on the production of O 2 in biofilm. Strains of Pseudomonas (10 ll with 108 –109 cfu/mL) were incubated in tubes with vertical slides of glass or plastic immersed in 5 ml trypticase soy broth (TSB). After 24, 48, and 72 h slides were washed three times with TSB and bacteria adhesion was determined according to previous description. Production of O 2 in normal conditions and in presence of antibiotics was investigated by CL. Slides with 106 cfu/slide were incubated with 10 lL TSB, 0.3 mL lucigenin (75 lg/mL), and 0.3 mL of each antibiotic. The light emission was measured by CL in a BioOrbit chemiluminometer mod 1253, and the results were expressed as relative light units (RLU) at different concentrations of each antibiotic. RLU of slides were compared with RLU produced by the same amount of bacteria in suspensions (106 cfu/slide and 106 cfu in 0.1 mL suspensions). Investigation of minimum inhibitory concentration. Minimum inhibitory concentration (MIC) was determined in M€ uller–Hinton medium using the standard tube dilution method. The strains (0.1 mL) coming from TSB were diluted in 9.9 mL buffer, incubated for 10 min at 37 °C, and then the antibiotics were added in different concentrations, between 0.5 and 128 lg/mL. Bacteria development was observed at 24 h of incubation. Oxidative stress caused in a sensitive mutant of Pseudomonas aeruginosa. A sensitive mutant M10 obtained after 10 consecutive cultures of 48 h in TSB more antibiotic under MIC concentration (0.2 lg/mL) was studied by CL. Generation of O 2 and MIC was determined in wild strain and M10 , according to description in the above section.
Results The assays of NBT indicated stimuli of oxidative stress with increase of ROS in sensitive strains of all bacteria species studied. Table 1 shows the mean values of relation ROS with antibiotic/ROS without antibiotic observed in sensitive strains of each species studied. Moreover, Cl demonstrated augment of O 2 only in sensitive strains; no significant increase was observed in resistant strains during the assays with ciprofloxacin in E. faecalis, S. aureus, and E. coli. Similarly, chloramphenicol stimulated oxidative stress in sensitive strains of S. aureus and E. coli. Sensitive strains of P. aeruginosa treated with ceftazidime or piperacillin augmented the production of O 2 while resistant strains did not exhibit increment of this anion (Table 2). There was dose-dependent response in oxidative stress caused by
Table 1 Increment of reactive species of oxygen (ROS) in bacteria of different species incubated with diverse antibiotics with respect to controls without antibiotic ROS with antibiotic/ROS without antibiotic E. coli E. faecalis S. aureus E. coli S. aureus P. aeruginosa
Ciprofloxacin 1.2 1.4 1.5 Chloramphenicol 1.6 2.1 Ceftazidime 1.9
Piperacillin 1.8
I. Albesa et al. / Biochemical and Biophysical Research Communications 317 (2004) 605–609 Table 2 Augment of O 2 generated by ciprofloxacin, chloramphenicol, ceftazidime, and piperacillin on sensitive strains of different bacteria species Increase of O 2 (RLU) with respect to normal level produced by E. coli E. faecalis S. aureus E. coli S. aureus P. aeruginosa
Ciprofloxacin 1.33 1.99 5.69 Chloramphenicol 1.98 5.88 Ceftazidime 4.67
Piperacillin 3.55
chloramphenicol or ciprofloxacin, the maximum stimuli was obtained with 0.5 lg/mL of each antibiotic. Oxidative stress generated by ciprofloxacin in sensitive strains of S. aureus was characterized by a maximum augment of O 2 , followed by a subsequent reduction of RLU. The effect caused by chlorampheni-
Fig. 1. Stimuli of anion superoxide in a sensitive strain of S. aureus treated with ciprofloxacin (j-j) and chloramphenicol (r-r) showing in both cases a maximum with subsequent reduction. Control without antibiotic (d-d).
Fig. 2. Oxidative stress in a sensitive strain of E. coli treated with ciprofloxacin (j-j) and chloramphenicol (r-r). Control without antibiotic (d-d).
607
col exhibited similar evolution as function of time, as presented in Fig. 1. Escherichia coli in presence of ciprofloxacin or chloramphenicol suffered an oxidative stress with less increase of O 2 and more pronounced decrease than S. aureus (Fig. 2). Otherwise, E. faecalis showed an important augment of O 2 followed by a non-pronounced decrease during the time observed (Fig. 3). Pseudomonas aeruginosa was treated with antibiotics recommended to this bacteria. Ceftazidime and piperacillin induced increase of O 2 in sensitive strains with a subsequent decrease, in the same form as was observed with ciprofloxacin or chloramphenicol in the other species (Fig. 4). Selection of P. aeruginosa strains able to form important biofilm was performed by means of tubes stained with safranin and microscopic observations. The strain that exhibited highest adherence to good biofilm formation was incubated in glass slides to perform the assays of CL. The increase of O 2 detected by CL in slides incubated with ceftazidime or piperacillin was comparable to controls non-treated with these antibiotics (Table 3).
Fig. 3. Enterobacter cloacae susceptibility to stress caused by ciprofloxacin (j-j). Control without antibiotic (d-d).
Fig. 4. Oxidative stress produced by ceftazidime (j-j) and piperacillin (r-r) in P. aeruginosa. Control without antibiotic (d-d).
608
I. Albesa et al. / Biochemical and Biophysical Research Communications 317 (2004) 605–609
Table 3 Increase of lucigenin-CL due to stimuli of superoxide anion in P. aeruginosa and concentration required of ceftazidime and piperacillin to stress bacteria in biofilm and in suspension O 2 (RUL with antibiotic/RUL without antibiotic)
Antibiotic (lg/mL)
Ceftazidime Biofilm Suspension
1.6 4.8
80 0.5
Piperacillin Biofilm Suspension
1.4 3.9
8 0.5
P. aeruginosa in suspension was more susceptible to oxidative stress generated by ceftazidime or piperacillin than in biofilm, 0.5 lg/mL ceftazidime being sufficient to generate the maximum stimuli of O 2 in suspension; while 80 lg/mL ceftazidime was necessary to arrive at maximum oxidative stress in biofilm. Similarly, attached bacteria suffered oxidative stress with higher piperacillin concentration than bacteria in suspension. O 2 suffered maximum increase with 0.5 lg/mL piperacillin in suspension of bacteria and it was necessary for 8 lg/mL piperacillin to obtain oxidative response in biofilm. It was evident that attached bacteria were more resistant to oxidative stress generated by both antibiotics. The biofilms exhibited low oxidative stress after 48 and 72 h in presence of antibiotics. Aging makes attached bacteria more resistant to oxidative effect generated by the antibiotics, oxidative stress decreasing after 24 h of biofilm formation. Table 4 shows the augment of O 2 generated to piperacillin at different times of biofilm formation. A strain of P. aeruginosa able to form important biofilm presented a mutant (M10 ) with little capacity to form biofilm and more sensitive to antibiotics. This mutant showed less exopolysaccharide production and
Table 4 Effect of time of biofilm formation on susceptibility to oxidative stress generated by piperacillin in a selected strain of P. aeruginosa Time of biofilm (h)
O 2 (RUL with antibiotic/RUL without antibiotic)
24 48 72
1.53 1.10 1.02
Table 5 Increase of O 2 measured by chemiluminescence in wild type and mutant strain of P. aeruginosa treated with antibiotics
Wild type Mutant
Ceftazidime
Piperacillin
1.1 2.1
1.2 2.5
in presence of antibiotics exhibited higher O 2 production than wild type strain (Table 5).
Discussion The results obtained indicated that diverse antibiotics can generate increase of O 2 in bacteria of different species; only strains sensitive to these antibiotics presented oxidative stress. Ciprofloxacin, ceftazidime, piperacillin, and chloramphenicol stimulate oxidative stress in bacteria, probably this observation could be useful to explain aspects of antibacterial action that need more explanations. Recently, it was described that S. aureus suffered oxidative stress in presence of ciprofloxacin with stimuli of superoxide dismutase (SOD) [2]. It is possible to think that the alteration of oxidative metabolism and increase of O 2 could be an important effect involved in the mechanism of action of different antibiotics, because augment of O 2 in all sensitive bacteria and all antibiotics selected for this study was observed. Indeed, oxidative stress was appreciated in species with different degrees of defense against oxidative metabolism, i.e., P. aeruginosa, a non-fermentative bacteria, E. faecalis, a fermentative microorganism, E. coli, an anaerobic facultative with important intracellular SOD, and S. aureus, with lower SOD activity in the cytoplasm than extracellular [11]. In addition, resistance to oxidizing agents like H2 O2 was observed by other authors in biofilm of P. aeruginosa, this bacteria possesses two catalases, KatA and KatB. During the purification of KatB, it showed cellular localization and demonstrated that it is essential for optimal resistance to hydrogen peroxide. However, it was observed that catalase activity was greater in suspension than in biofilm-producing organisms, because these last bacteria reduced the oxidative metabolism and required less activity of catalase and superoxide dismutase as consequence of their little production of oxidative radicals [12]. According to our results attached P. aeruginosa produced less O 2 than in suspensions, which coincided with previous investigations that indicated which sessile bacteria present constitutive SOD while planktonic cells showed high SOD activity because bacteria in suspension have high necessity of this enzyme to counteract the elevated production of toxic ROS [13]. The antioxidant enzyme SOD can be induced by O 2 its activity in suspensions being more intense because free bacteria must neutralize higher amount of O 2 than in biofilm. Biofilm’s resistance to ceftazidime and piperacillin can be derived from physiological differences between biofilm and planktonic organisms. Different phenotypes can appear with diverse sensitivity to oxidative stress during the infection. Changes of phenotypes were
I. Albesa et al. / Biochemical and Biophysical Research Communications 317 (2004) 605–609
described during the process of biofilm development like production of alginate, alteration of growth rate, and oxidative metabolism. It was observed that QS governs the control of defenses against substances that stimulated oxidative stress [14]. However, there are few antecedents in this subject and more studies must be performed to know the consequences of oxidative stress in bacterial biofilms treated with antibiotics. The bacteriostatic or bactericidal effects could be related to induction of O 2 in response to exposure to antibiotics, future investigation must be done to know if oxidative stress contributed to the antibacterial activity. The fact that slides of 72 h exhibited low stimuli of O 2 than slide of 24 h suggested that aging increased the resistance of P. aeruginosa to oxidative stress and a possible relationship can be suggested between this observation and the well-known phenomenon of induction of anaerobic metabolism in biofilms [15]. Bacteria recently attached generate more ROS than old biofilm because aerobic growth produced more ROS than anaerobic growth, and antibiotics generated low oxidative stress when bacteria reduced the oxidative metabolism. This new aspect might be considered in the treatment of CF and other serious infections in which bacteria form biofilm resistant to antibiotics; probably assays of CL could be useful to determine the susceptibility to antibacterial effect in attached bacteria.
Acknowledgments The authors thank Agencia de Promoci on Cientıfica y Tecnol ogica (BID 1201/OC-AR. 06-07522), Agencia C ordoba de Prom-
609
oci on Cientıfica y Tecnica, and Secretarıa de Ciencia y Tecnica de la Universidad Nacional de C ordoba, for their support and collaboration.
References [1] Y. Hiraku, S. Kawanishi, Arch. Biochem. Biophys. 382 (2000) 211–218. [2] M.C. Becerra, I. Albesa, Biochem. Biophys. Res. Commun. 297 (2002) 1003–1007. [3] P. Andre, S. Bilger, P. Remy, S. Bettinger, D.J. Vidon, Biochem. Biophys. Res. Commun. 304 (2003) 807–811. [4] A. Battistoni, G. Donnarumma, R. Greco, P. Valenti, G. Rotilio, Biochem. Biophys. Res. Commun. 243 (1998) 804–807. [5] M.M. Huycke, V. Abrams, D.R. Moore, Carcinogen 23 (2002) 529–536. [6] H. Kobayashi, Am. J. Med. 99 (1995) S26–S30. [7] P.S. Stewart, Int. J. Med. Microbiol. 292 (2002) 107–113. [8] S. Favre-Bonte, J.C. Paghe, J. Robert, D. Blanc, J.C. Pechere, C. van Delden, Microb. Pathog. 32 (2002) 143–147. [9] T.C. Mah, G.A. O’Toole, Trends Microbiol. 9 (2001) 34–39. [10] R.A. Burnne, R.G. Quivey Jr., R.E. Marquis, Methods Enzymol. 310 (1999) 441–460. [11] P. Bogdanov, M. Martınezs de Bertorello, I. Albesa, Biochem. Biophys. Res. Commun. 244 (1998) 561–566. [12] S.M. Brown, M.L. Howell, M.L. Vasil, A.J. Anderson, D.J. Hassett, J. Bacteriol. 177 (1995) 6536–6544. [13] J.R. Frederick, J.G. Elkins, N. Bollinger, D.J. Hassett, T.R. McDermott, Appl. Environ. Microbiol. 67 (2001) 1375– 1379. [14] D.J. Hassett, J.F. Ma, J.G. Elkins, T.R. McDermott, U.A. Ochsner, S.E. West, C.T. Huang, J. Fredericks, S. Burnett, P.S. Stewart, G. McFeters, L. Passador, B.H. Iglewski, Mol. Microbiol. 34 (1999) 1082–1093. [15] J.W. Costerton, Mol. Cell 10 (2002) 699–700.
