Evaluation of antibacterial activity of some non-steroidal anti-inflammatory drugs against Escherichia coli causing urinary tract infection

1 Department of Microbiology and Immunology, Faculty of Pharmacy, Deraya University, Minia, Egypt. 2 Department of Microbiology and Immunology, Faculty of Pharmacy, Minia University, Minia, Egypt. 3 Department of Microbiology and Immunology Faculty of Pharmacy, Hail University, KSA, Saudi Arabia. 4 Department of Microbiology and Immunology, Faculty of Pharmacy, Ahram Canadian University, 12451Sixth of October City, Egypt.


INTRODUCTION
Urinary tract infections (UTIs) are the most common hospital acquired infection with a percentage of 35% of nosocomial infections (Stamm, 2002;Weinstein et al., 1997).Escherichia coli is the major pathogen causing urinary tract infections (UTIs) and represents more than 85% of recurrent cystitis and about 35% of recurrent pyelonephritis (Barnett and Stephens, 1997).UTIs remain the most common human bacterial infection, despite the high spread of antibiotics.The massive and irrational use of antibiotics and antibacterial agents for long periods has led to the emergence of multi drug resistant (MDR) microorganisms, and it is currently advised that the clinical administration of antibiotics against the pathogenic bacteria be gradually prohibited (Ray and Rice, 2004;Chowdhary et al., 1994).Another solution for this problem is to search for non-antibiotic compounds that have antibacterial activity through different mechanisms (Mazumdar et al., 2009).Recent studies have shown that some medicines have antibacterial activity in addition to their main function such as antihistamines, antipsychotics, tranquillizers, antihypertensives and local anesthetics (Rani et al., 2005;Dastidar et al., 1995).All these drugs with moderate to powerful anti-microbial activities have been known as "non-antibiotics" (Dastidar et al., 2000).Non-steroidal anti-inflammatory drugs (NSAIDs) are commonly used medicines for pain and inflammation management and previous studies have revealed that some NSAIDs have antibacterial activity (Wang et al., 2003;Hersh et al., 1991).NSAIDs exhibited strong antimicrobial activity when tested against a large number of Gram-positive and negative bacteria and the MIC ranged from 50-200 µg/mL in most of the cases and even lower in some cases (Annadurai et al., 1998;Sukul et al., 2015;Obad et al., 2015).The antibacterial agents, whether bacteriostatic or bactericidal, might act by inhibition of microbial cell wall synthesis, alteration of membrane function or membrane damage, inhibition of nucleic acid synthesis or inhibition of protein synthesis (Mazumdar et al., 2006).The aim of this study was to detect antibacterial activity of some NSAIDs (diclofenac sodium, aspirin, indomethacin and ibuprofen) against E. coli isolates causing UTIs, examine the effect of their combination with different antibiotics and finally detect the possible mechanism of antibacterial action of these NSAIDs if present.

Isolation of bacterial strains
One hundred urine samples were collected from UTI patients in Minia University Hospitals (MUH) in Minia, Egypt during the study period, from May 2014 to December 2015.Informed consent was obtained from all the subjects.Urine samples were inoculated on cysteine lactose electrolyte deficient (CLED) media (Lab, UK) (Winn and Koneman, 2006).All the samples were examined for the presence of E. coli by streaking them onto MacConkey agar (Lab, UK), EMB agar (Himedia, India) and incubating the plates at 37°C for 24 h.Identification of E. coli was based on fermentation of lactose giving pinkish colonies.Further identification was done by biochemical (citrate and triple sugar iron) tests.Bacteria were maintained by storage at -70°C on tryptone soy broth (TSB) medium (Himedia, India) enriched with 20% glycerol (Rusu et al., 2014;Nobmannn et al., 2010).

Susceptibility testing
Bacterial cultures were tested against some NSAIDS (diclofenac sodium, aspirin, indomethacin and ibuprofen) by agar dilution method (CLSI, 2005).Mueller-Hinton agar (MHA) plates (Lab, UK) contained two fold serial dilutions of NSAIDS from 0.25 to 1024 μg/mL.Bacterial suspensions of isolated bacteria were made in sterile saline and matched with McFarland index 0.5 tubes.Each bacterial suspension (1 μl) was inoculated (3×10 5 CFU/spot) on drug containing plates and incubated at 37°C for 24 h.

Determination of interaction between NSAIDs and antibiotics by checkerboard dilution technique
Two drugs combined effects were determined by the Checkerboard dilution technique to determine the fractional inhibitory concentration (FIC) indices.Definition of FIC is as follows: MIC of substanceA tested in combination/MIC of substanceA tested alone + MIC of substanceB tested in combination/MIC of substanceB tested alone.The FIC index (FICI) was calculated using the following formula: Synergism is shown as FIC index of ≤0.5, while indifference is showed as an FIC index of >0.5 ≤4 and antagonism is shown as an FIC index of>4.FIC index was an average of two independent experiments (Lorian, 2005).

Membrane-permeability assay
Membrane-disruptive activity of Amoxicillin, diclofenac sodium, aspirin, indomethacin, ibuprofen and amoxicillin/aspirin on in vitro grown E. coli (ATCC 8739) was determined by measuring the fluorescence enhancement of ethidium bromide (Sigma) (Paixão et al., 2009).To this end, E. coli were grown in tryptone broth medium in the presence or absence of amoxicillin as positive control, NSAIDs and Amoxicillin combined with aspirin for 24 h.The bacterial culture was incubated with ethidium bromide for 20 min at room temperature in the dark.Membrane permeability was  (Bink et al., 2012).

Loss of 260 nm absorbing material
The release of UV-absorbing material concentrations were determined by UV spectrophotometer (Zhou et al., 2008).Loss of 260 nm absorbing material released from bacteria was measured by the technique performed by Devi et al. (2010).Overnight broth cultures of E. coli ATCC 43889 in tryptone broth medium were adjusted to OD 600.Cells were harvested by centrifugation at 400 rpm for 15 min, supernatant was discarded, and pellet was washed twice and re-suspended in phosphate buffer saline (PBS) at pH 7.4.Different concentrations of NSAIDs [1\2MIC, MIC, 2MIC and 4MIC] were added to the cell suspension.Amoxicillin (1.6 mg/ml) was used as positive control.The experiment was done in triplicates.Cells without NSAIDs treatment were used as control.All the samples were incubated at 37°C for 60 min.After treatment, the cell suspension was centrifuged at 13,400 rpm for 15 min and OD 260 value of the supernatant was taken as a percentage of the released extracellular UV-absorbing materials.All the measurements were done in triplicates in Jenway 7305 UV spectrophotometer (UK).

Scanning electron microscopy (SEM)
E. coli (ATCC 8739) cells were suspended in saline solution containing 0.2% Tween-80 and incubated at 37°C with Amoxicillin, diclofenac sodium, aspirin, indomethacin, ibuprofen and Amoxicillin/aspirin at 2× MIC at room temperature.After 24 h, the bacterial cells were centrifuged at 8000 rpm for 15 min.The bacterial cells were then washed with 0.1 mol/l tris-acetate buffer (PH 7.1), fixed in tris-acetate buffer containing 1.5% glutaraldehyde, and then freeze-dried.Each bacterial culture was observed by SEM (Hitachi, Japan) at magnifications of 10000, 7500 and 15000x.The bacterial cell suspension in saline with no NSAIDs treatment served as a negative control (Soboh et al., 1995).

Statistical analysis
Statistical analysis was performed using one way Anova test.P values of <0.05 were considered indicative of statistically significant differences.

In vitro antimicrobial action of NSAIDs
NSAIDs were tested against a total of 48 isolates of E. coli as shown in Table 2. Diclofenac sodium showed the lowest MIC 50 and MIC 90 : 8 and 256 g/ml, respectively.Aspirin showed MIC 50 of 64 g/ml, while both indomethacin and ibuprofen showed MIC 50 of 256 g/ml.Indomethacin, aspirin and ibuprofen showed the same MIC 90 of 1024 g/ml.But for the standard strain as illustrated in Table 3, indomethacin showed the lowest MIC: 128 g/ml, followed by aspirin: 256 g/ml.Diclofenac sodium and ibuprofen showed the same MIC: 1024 g/ml.

Determination of interaction between NSAIDs and antibiotics by checkerboard dilution technique
The combined effects of the four NSAIDs (diclofenac sodium, indomethacin, aspirin and ibuprofen) and five antibiotics (amoxicillin, augmentin, cefotaxime, ciprofloxacin and gentamicin) were tested on five resistant clinical E. coli strains.All the tested NSAIDs significantly lowered the MICs of antibiotics against the tested bacteria.The synergistic effects of NSAIDs and five antibiotics combination are shown in Tables 4, 5, 6 and 7. FICIs for this combination ranged from 0.03 to 0.5 against the tested bacteria.All the examined E. coli showed high reduction in MIC values with NSAIDs and the five antibiotics.
On the other hand, the combined effects of the four NSAIDs and the five antibiotics on standard E. coli strain are shown in Table 8.These results showed that NSAIDs have a synergistic effect when combined with antibiotics and this combination effectively inhibit UTIs causing bacteria.

Membrane-permeability assay
These results suggested an effect of pretreatment of NSAIDs on the E. coli activity.It is hypothesized that NSAIDs affect membrane permeability of the tested E. coli cells, exhibited by the use of the fluorescent dye ethidium bromide.It is revealed that E. coli treated with different concentrations of NSAIDs during the growth phase resulted in a significantly increased membrane permeability of E. coli cells compared to the untreated ones as found by the significant increase in fluorescence of NSAIDS-treated cells (Figure 1).The presented fluorescence values are corrected with those obtained from untreated bacteria.

Effect of NSAIDs on leakage of 260 nm absorbing materials from E. coli
The measurement of release of UV-absorbing materials is an index of cell lysis (Zhou et al., 2008).The leakage of cytoplasmic membrane was analyzed by determining the release of cellular materials including nucleic acids, metabolites and ions, which were absorbed at 260 nm into the bacterial suspensions (Bajpai et al., 2014).After treatment with different concentrations of NSAIDs, the OD significantly increased up to 1.87 from 0.00 (P value ˂ 0.05) as shown in Table 9.These results suggest that NSAIDs damage cytoplasmic membrane and cause subsequent leakage of intracellular constituents.

Scanning electron microscopy (SEM)
SEM images showed differences in cell structures between NSAIDs-treated bacteria and the non-treated control bacteria.Non-treated bacteria were intact (regular rod shaped) and showed smooth surfaces as seen in Figure 2A, while bacterial cells treated with the individual NSAIDs underwent considerable structural changes as shown in Figures 2B to G. SEM observations confirmed the damage to the structural integrity of the cells and considerable morphological alteration to the tested bacteria.In Figure 2G, combined NSAIDs treatments altered the outer membrane, the structures of the cells and made them more permeable.

DISCUSSION
E. coli is the major bacterial uropathogen in the world (Miragliotta et al., 2008).In the study on 100 urine samples, 48 (48%) E. coli strains were detected.This is similar to findings from studies done in other countries such as India (50, 59 and 68%) (Ranjini et al., 2015;Kothari and Sagar, 2008;Tambekar et al., 2006) and Madagascar (67%) (Randrianirina et al., 2007).Another study performed in Egypt reported that E. coli was in 36% of UTIs patients (Alabsi et al., 2014).A study performed in South Africa revealed that E. coli was present in 75% of uncomplicated and 59% of complicated UTIs and it was similar to this study (Agpaoa et al., 2015).In this study, 23 strains (47.9%) of E. coli isolates were normally resistant to most of the antibiotics showing multi-drug resistance (MDR).The antibiogram resistance pattern of the strains was: Amoxicillin (66.7%),Ampicillin (56.3%),Cephalexin (43.8%),Augmentin (41.7%),Cephradin (41.7%),Cefotaxime (20.8%),Ciprofloxacin (18.8%) and Gentamicin (6.3%).A study done in Egypt revealed the same percentage of MDR E. coli: 40% (Alabsi et al., 2014).A very high degree of MDR of 82.5% among E. coli isolates was reported by Ranjini et al. (2015).This  study is in accordance with the study of Mazumdar et al. (2006) who reported the antibiogram resistance pattern of the E. coli strains as: ampicillin (74.4%), augmentin (59%), cefotaxime (38%), and these findings were similar to the results of Samsygina et al. (2000) and Khan et al. (2002).Alabsi et al. (2014) from Egypt reported 89 and 57% resistance among urinary E. coli isolates to ampicillin and gentamicin, respectively.NSAIDs are commonly used medicines for the treatment of pain and inflammation.Many studies found that some NSAIDs have good antibacterial activity especially diclofenac sodium (Wang et al., 2003).In this study, some NSAIDs (diclofenac sodium, aspirin, indomethacin and ibuprofen) were tested against a total of 48 isolates of E. coli.Diclofenac sodium showed the lowest MIC 50 and MIC 90 ; 8 and 256 g/ml, respectively.Annadurai et al. (1998) reported that the MIC in most of the cases ranged from 50-200 g/ml and even lower in some cases and diclofenac was bactericidal in its action.Dutta et al. (2007a) studied 32 isolates of E. coli, 8 were inhibited at 50 g/ml diclofenac, 9 at 100 g/ml, 5 at 400 g/ml and the remaining isolates of E. coli were resistant to diclofenac (MIC≥ 800 g/ml).In this study, aspirin showed MIC 50 of 64 g/ml, while both indomethacin and ibuprofen showed MIC 50 of 256 g/ml.Indomethacin, aspirin and ibuprofen showed the same MIC 90 of 1024 g/ml.Wang et al. (2003) tested the MICs of aspirin for 66 H. pylori isolates and the MIC 50 of aspirin was 256 g/ml, MIC 90 was 512 g/ml, and the range of MIC values was 256 to 512 g/ml and this finding is close to the current study results.Activity of ibuprofen on E. coli was proximally studied by Al-Janabi (2010) and showed susceptibility to the tested agent at MIC of 2.5 mg/ml, which is higher than this results.There is an ongoing trial in Germany evaluating reduction of the use of antibiotics for uncomplicated UTI by giving initial management with ibuprofen (Gágyor et al., 2012).NSAID is equally effective as an antibiotic, and this may lead to a reduction in the use of antibiotics and reduce antibiotic resistance.This is good to the environment and will reduce the costs in health services internationally (Vik et al., 2014).The combined effects of the four NSAIDs (diclofenac sodium, indomethacin, aspirin and ibuprofen) and five antibiotics (amoxicillin, augmentin, cefotaxime, ciprofloxacin and gentamicin) were tested on five resistant clinical E. coli strains by checkerboard dilution technique.All the tested NSAIDs significantly reduced the MICs of antibiotics against the tested bacteria and FICIs for this combination ranged from 0.03 to 0.5 with respect to synergism.Dutta et al. (2007a) used the checkerboard technique giving a FIC index for E. coli of 0.49 for diclofenac and streptomycin, thereby showing a synergistic effect and another study showed that the combination effect of diclofenac with gentamicin/ampicillin which was examined by using checkerboard technique yielded FIC index ranging from 0.4 to 0.5 for diclofenac + gentamicin and values >1 for diclofenac + ampicillin (Dutta et al., 2009).
In the present study, NSAIDs alone recorded antimicrobial activity, but NSAIDs in combination with antibiotics exhibited significant synergistic effect and the drugs were bactericidal.These data suggested that NSAIDs in combination with antibiotics could be useful for the treatment of complicated bacterial infections.In addition to yielding these synergistic effects, the combinations of two or more compounds are important to prevent or suppress the developing of resistant strains, to decrease dose toxicity and to perform a broad spectrum activity (Eliopoulos and Moellering, 1996).The bacterial membrane is a structural component which may be damaged during a bactericidal challenge.Therefore, release of intracellular components is an indicator of membrane integrity.Small ions such as potassium and phosphate when treated with a suitable antimicrobial agent leach out first, followed by large molecules such as DNA, RNA and other materials.These substances have strong UV absorption at 260 nm, they are known as "260nm absorbing materials" and this method is widely used in the determination of membrane integrity parameters (Denyer, 1990;Hugo and Snow, 1981).In this study, leakage of intracellular components suggests that the NSAIDs effect on E. coli can be through pores formation in the bacterial plasma membrane.The bacterial surface morphology alteration and cell damage could be confirmed thoroughly by SEM (Benli et al., 2008).In this study, SEM images showed differences in cell structures between NSAIDs-treated bacteria and the non-treated control bacteria.In addition, combined NSAIDs treatments altered the outer membrane as the structures of the cells made them more permeable.Thus, the mode of bactericidal action of NSAIDs against E. coli is through membrane disruption and so blocking the bacterial growth.The exact mechanism of antibacterial activity of diclofenac and ibuprofen is unclear.However, studies have suggested inhibition of bacterial DNA synthesis (Dutta et al., 2004) or impairment of membrane activity that agree with results obtained by SEM in this study (Hersh et al., 1991;Dutta et al., 2007a, b;Mohsen et al., 2015;Sikkema et al., 1995).
In conclusion, diclofenac sodium, aspirin, indomethacin and ibuprofen showed antibacterial activity against E. coli causing UTIs.This study results indicate that a combination of these NSAIDs and antibiotics exhibited good synergism against E. coli associated with UTIs, and the mechanism of their action was by damaging the bacterial cell membrane.This new finding of combination treatment with NSAIDS and antibiotics might provide an alternative way to overcome antibacterial drug resistance.However, further in vivo and clinical studies will be required to support this suggestion.

Figure 1 .
Figure1.Membrane-disruptive activity of NSAIDs on in vitro grown E. coli.Data presented are the mean and standard error of the mean of 3 independent biological repeats.Statistical analysis was performed using one way ANOVA test and relative to untreated bacteria (P < 0.05).

Table 1 .
Distribution of minimum inhibitory concentrations and the prevalence of antibiotics resistance among the isolated E. coli.
determined by measuring the ethidium bromide fluorescence (excitation at 518 nm, emission at 605 nm).Fluorescence values presented are corrected with those obtained from untreated bacteria

Table 2 .
Distribution of minimum inhibitory concentrations of NSAIDs among the isolated E. coli.

Table 3 .
Distribution of MICs of NSAIDS and antibiotics against the standard strain.

Table 4 .
Synergistic effect of diclofenac sodium combination with five antibiotics on resistant clinical E. coli strains.

Table 5 .
Synergistic effect of indomethacin combination with five antibiotics on resistant clinical E. coli strains.

Table 6 .
Synergistic effect of aspirin combination with five antibiotics on resistant clinical E. coli strains.

Table 7 .
Synergistic effect of ibuprofen combination with five antibiotics on resistant clinical E. coli strains.

Table 8 .
Distribution of NSAIDs/antibiotics FICindexagainst the E. coli standard strain.

Table 9 .
Effects of NSAIDs at different concentrations on membrane integrity in E. coli standard strain measured by release of UV absorbing components at 260 nm.