Abstract: Multidrug Resistance (MDR) Pseudomonas aeruginosa is an emerging nosocomial pathogen worldwide. In severe infections, Extended Spectrum β-lactamases (ESBLs) and expression of various virulence factors may work in harmony, resulting in the treatment failure. An association between ESBL production and two virulence factors: cell surface hydrophobicity and biofilm formation in P. aeruginosa has been demonstrated previously. This study was designed to compare the antimicrobial resistance pattern and production of some virulence factors in ESBL producers ( ESBL+) and ESBL non-producers (ESBL-) clinical isolates of P. aeruginosa. ESBL production was detected by the double disk method and antimicrobial activity was tested by agar dilution methods. Antibacterial resistance pattern showed 48 different antibiotypes with 25 and 23 different types in ESBL+ and ESBL- isolates, respectively. The Minimum Inhibitory Concentrations (MICs) and percentage of isolates resistant to nearly all antibacterial agents tested, was significantly higher in ESBL+ isolates compared to ESBL-isolates ( p≤0.001). The activity of LasA in ESBL+ isolates was higher than the negative isolates and statistically significant (p≤0.05). Expression of mannose sensitive haemagglutination and production of pyoverdin in the ESBL+ isolates was significantly lower than the ESBL-isolates (p≤0.04). No significant difference in the expression of mannose resistant haemagglutination and production of LasB was found between the ESBL+ positive or negative isolates. The result of present study suggests a correlation between ESBL phenotype and production of some factors that are reported to be involved in the virulence in P. aeruginosa.
INTRODUCTION
Although, opportunistic in nature, Pseudomonas aeruginosa is now emerged as a major pathogen in patients with cystic fibrosis, weak or immunocompromised patients in the hospitals or nursing homes (Mesaros et al., 2007; Pearson et al., 1999). Multidrug resistance in P. aeruginosa is an increasing problem worldwide (Hocquet et al., 2006; Livermore and Woodford, 2006). Presence of low outer membrane permeability, expression of several efflux pumps and production of extracellular enzymes to inactivate antibiotics are the main reason of intrinsic resistance of P. aeruginosa to several antibacterial agents (Tomas et al., 2010; Abdi-Ali et al., 2007).
P. aeruginosa is able to form biofilm on different animate or inanimate surfaces; this property helps the survival and protects the bacteria from the immune system and antimicrobial agents inside the host (Deziel et al., 2001; O’toole and Kolter, 1998). Slime layer production by some strains of P. aeruginosa can protect the bacteria and keep them away from the action of the antibacterial agents. Slime producing isolates are reported to have a high rate of resistance to antibacterial agents. Production of slime layer on medical devices is an important factor in pathogenesis of device associated infections with P. aeruginosa (Farrag, 2001).
Regulation of growth, production of many virulence factors and resistance to several antibacterial agents in P. aeruginosa is under the control of a population density, cell-to-cell communication called Quorum Sensing (QS) (Aendekerk et al., 2005; Kong et al., 2005; Ohman et al., 1980). Expression of many virulence genes and acquisitions of genes involved in multidrug resistance are usually mediated by different efflux pumps and are under the control of QS (Deptula and Gospodarek, 2010). Different efflux pumps in P. aeruginosa are able to eject multiple antimicrobials from the cell, including β-lactamas and effects penicillins as well as non β-lactams such as fluoroquinolones (Livermore and Woodford, 2006). Other mechanism of resistance to β-lactamases is production of chromosomally encoded AmpC type β-lactamases and production of ESBLs (Hocquet et al., 2006). ESBLs production is often associated with the mobile genetic elements, carrying genes that encode resistant to other antimicrobials such as aminoglycosides, sulfonamides, quinolones and trimethoprim (Livermore and Woodford, 2006; McGowan et al., 1989). It has been shown that ESBLs producing isolates of Klebsiella pneumoniae are more invasive, producing more fimbrial adhesions and are more resistance to bactericidal activity of serum (Sahly et al., 2004, 2008). This study was aimed to detect any possible correlation between the virulence factors and ESBL production in P. aeruginosa isolated from different clinical specimens of hospitalized patients.
MATERIALS AND METHODS
P. aeruginosa isolates were collected from different clinical samples of patients admitted to 3 different hospitals located in the southeast (Kerman) and 2 hospitals in the southwest (Shiraz) of Iran from January 2008 to June 2009. The isolates were identified by conventional biochemical tests (Hall, 2007). Only one isolate per patient was included. Type of sample and the place of the patient in the hospital were recorded. Totally 204 isolates were identified as P. aeruginosa. The isolated bacteria were kept at -70°C in TSB glycerol until use. After determination of Minimum Inhibitory Concentrations (MICs) of the isolates to 9 antimicrobial agents by agar dilution method (Hanlon et al., 2007) the isolates showing resistance to 3 antibacterial agents from different classes were selected as the multiple drug resistance MDR. ESBL production in these isolates was determined by Combined Disc method (CD) as described previously (Norouzi et al., 2010). In order to have a more diverse group of isolates, the isolates were selected based on antibiotypes and ESBL production. Fifty ESBL+ isolates comprising of 25 antibiotypes and 50 ESBL-isolates, comprising of 23 antibiotypes were selected for the experiment.
For determination of Mannose Sensitive(MS) and Mannose Resistance (MR) haemagglutination pili the bacteria were serially sub cultured four times every 48 h in static brain heart infusion broth or nutrient broth for the detection of MS or MR, respectively (Evans et al., 1979; Sahly et al., 2008). Washed RBC prepared from freshly drawn guinea-pig blood or human blood group O+, was used for the detection of MS or MR, respectively. The haemagglutination was graded from negative to 3+ according to Sahly et al. (2008).
For determination of Elastase (LasB) and Staphylolysin (LasA) activity the bacteria were grown up to the early stationary phase of growth (OD of 1.7 at 600 nm) in LB medium in a shaker incubator (120 rpm). LasB activity was determined with Elastin Congo red (ESR, Sigma, St. Louis, Mo, USA) as a substrate (Ohman et al., 1980). The insoluble Congo red was removed by centrifugation after 6 h incubation and the optical density of the mixture was determined at 495 nm. For determination of Las A activity the boiled cell pellet prepared from an overnight culture of Staphylococcus aureus (ATCC 25923) was used (Kong et al., 2005). The activity of LasA and LasB were presented as the change in the optical density of the mixture per hour per microgram of protein (Musthafa et al., 2010). Protein concentration in the bacterial cell free supernatants was determined by Bradford (1976).
Production of pyoverdin by the bacteria was determined on the cell free supernatant of the bacteria grown on LB medium with shaking (120 rpm-24 h). The optical density of the isolates was recorded at 600 nm. The relative concentration of pyoverdin in the cell free supernatant was determined at wavelength of 405 nm (Singh et al., 2010).
To detect any possible contribution between the virulence factors tested and the ESBL encoding plasmids, sixteen ESBL+ isolates were selected randomly. These isolates were sub cultured in LB broths containing 2% SDS at 42°C. The isolates were tested for the loss of the ESBLs phenotypes and by the decrease in MIC to cefotaxime and ceftazidime to susceptible level every 48 h until 60 days. After then 8 isolates which became ESBL- were tested again for the virulence factors mentioned.
Statistical analysis: Expression of different virulence factors in ESBL+ and ESBL- isolates were compared using chi-square analysis using SPSS software (SAS Institute, Cary, NC, USA) Data are expressed as Mean±Standard Division (SD). The correlation between the antibacterial resistance patterns and virulence factors with ESBL phenotype were tested by the independent sample t-test. p<0.05 was considered statistically significant.
RESULTS AND DISCUSSION
In present study resistance to some antibacterial agents was high and 71-72% of the isolates was resistance to tetracycline and trimethoprim /sulfamethoxazole (SXT), respectively. The high rate of resistance to these antimicrobials in gram negative bacteria are increasing worldwide (Manikandan et al., 2011; Shahcheraghi et al., 2009; Giske et al., 2008). The ESBL+ isolates were significantly more resistance to antibacterial agents compared with the ESBL-isolates (p≤0.001), except for trimethoprim/sulfamethoxazole (SXT). SXT was the only antimicrobial agent that ESBL+isolate were more susceptible to it with (p≤0.04). Nearly all the ESBL- isolates were found to be sensitive to ceftazidime (100%), ciprofloxacin (98%) and imipenem (98%) while for ESBL+isolates the resistance rate was significantly higher (p≤0.002). The mean MIC for all the antibacterial agents tested (even for SXT) was statistically higher in the ESBL+compared to ESBL-isolates (Table 1). The same results has been reported for clinical isolates of enteric bacteria (Mansouri and Abbasi, 2010) and the gram negative bacteria from urinary tract infections (Selvakumar and Jasmine, 2007; Yousefi-Mashouf and Hashemi, 2006). This could be explained by the fact that ESBL+ bacteria usually had mobile genetic element coding gene for resistance to other antibacterial agent (McGowan et al., 1989). Statistically significant difference between the percentages of isolates resistance to antimicrobial agents in ESBL+ compared to non β-lactam antimicrobials is very important and ESBL production worked as a good marker for detection of MDR isolates. Bacterial isolates producing ESBL enzymes are increasing and they are among the most important problem contributing to significant mortality and morbidity in infected patients (Ho et al., 2010).
| Table 1: | Comparison between antimicrobial susceptibility of 50 ESBL producers with 50 non-ESBL producer clinical isolates of Pseudomonas aeruoginosa |
| *p-values are related to the differences between the mean of MIC between the two groups of isolates, producing and non-producing ESBLs | |
In the present study the most active agents against bacterial in both ESBL+and ESBL- were imipenem, ceftazidime and ciprofloxacin. In northwest of Iran (Urmia), the resistance rate of clinical isolates of P. aeruoginosa to CIP and CAZ was reported to be 50 and 78% which is higher than this study. However, resistance to IPM in this study is much higher than Uremia (Jazani and Babazadeh, 2008). Imipenem is an antibiotic that fights serious infections caused by the bacteria. The current shift in the treatment of pseudomonad infections with this agent may be the cause for higher rate of resistance to it. Although, the most active agent in this study was IPM, simultaneous resistance of IPM resistance isolates to all other antibacterial agents tested is a matter of concern. Therefore, use of this agent should be restricted in order to avoid the emergence of extremely drug resistance isolates. Multidrug resistance in P. aeruginosa is increasing worldwide (McGowan et al., 1989). Since ESBL+ is very common in this area, therapy with compounds other than β-lactams such as polymyxins (polmyxinB, or colistin, polmyxin E) must be considered for the treatment of severally ill patients infected with these bacteria.
ESBL+ expressed lower percentage of mannose sensitive haemagglutination (MS) compared to ESBL-isolates. MSHA was detected for 44% of the ESBL+ and 26% of ESBL negative isolates and the difference was significance (p≤0.04). Moderate and strong type of reaction was also higher in these isolates (36%) compared with ESBL+ isolates (52%, p≤0.04) (Fig. 1). In this study all the isolates except one (a strain of P. aeruginosa with ESBL+ phenotype isolated from a blood culture) expressed either MR or MS haemagglutination. These results are in accordance with the results obtained with K. pneumoniae and indicate that at least one fimbrial adhesion is required for colonization irrespective of the presence of ESBL plasmid (Sahly et al., 2008). P. aeruginosa isolates producing MS type haemagglutination are reported to be less hydrophobic than the organisms producing MR type haemagglutination (Sharma et al., 2007). Clinical isolates of P. aeruginosa producing ESBL are reported to be more hydrophobic and had a better ability to form biofilm compared to ESBL-isolates (p≤0.01) (Norouzi et al., 2010). The combination of these results suggests a correlation between production of these virulence factors and ESBL production.
Table 2 shows the difference between the activity of virulence factors in ESBL+ and ESBL-isolates. Mean activity of protease LasA was higher and LasB and pyoverdin production was lower in ESBL+ compared to ESBL-isolates (p≤0.05).
| Fig. 1: | Percentage of isolates producing of Mannose sensitive (MR) or Mannose resistance (MS) haemagglutination in ESBL producing (n = 50) and non-ESBL (n = 50) producing clinical isolates of Pseudomonas aeruginosa. Various degree of guinea-pig RBC (MS) or human O+ RBC haemagglutination (MR) were graded from negative to 3+++ |
| Table 2: | Comparison between the activity of LasA protease, LasB elastase and pyoverdin in Pseudomonas aeruginosa clinical isolates producing (50 isolates, ESBL +) and non-producing ESBL (50 isolates, ESBL-) and the standard strains used as the control |
| aAll the assays were performed on cell free supernatant of bacteria grown in LB medium and are presented as the Mean±SD. bControl strains | |
The activity of Las A was lower in the control strain PAO1 which was positive control for the experiments compared to standard strain ATCC 27853. Las B and pyoverdin production was higher in PAO1 compared to standard strain ATCC 27853. These results shows a similar activity of PAO1 strain and ESBL-isolates and strain ATCC 27853 with that of ESBL+ strains. In P. aeruginosa isolates with MDR phenotype the activity of LasB and pyocyanin or pyoverdin was reported to be lower than the susceptible isolates (Cowell et al., 2003; Deptula and Gospodarek, 2010). LasA and LasB can decrease the level of toxins (ExoS, ExoU) and LasA/LasB plays a role in the regulation of the steady state level of ExoS and ExoT in the cell, modulating the effect of the bacteria to invade into mammalian cells (Cowell et al., 2003). Our data showed that LasA/LasB in the ESBL+ isolates (2.2±0.97) was significantly higher than the ESBL- isolates (1.85±1.05; p≤0.005). It is possible that ESBL+ isolates regulates the production of these virulence factors in order to become invasive or cytotoxic in response to the environmental conditions.
No significant difference between productions of either or all of the virulence factors tested and the type of specimens from which the bacteria were isolated were found in this study. These results are similar to results obtained for K. pneumoniae and Proteus spp. (Mishra et al., 2001; Sahly et al., 2004, 2008). Plasmid curing experiments were performed to change the phenotype of ESBL+ to ESBL-. ESBL negative isolates after curing had a reduced MIC against extended spectrum cephalosporins and became sensitive to these agents. However, Plasmid curing did not affects the virulence activity of the cured organisms significantly. These findings are in agreement with the results obtained with K. pneumoniae resistance to serum and suggests that the ESBL plasmids is unlikely to be involved in the expression of the virulence factors (Sahly et al., 2004).
An association between expression of certain virulence factor and different type of infections and MDR phenotype have been reported in P. aeruginosa (Bradbury et al., 2010; Deptula and Gospodarek, 2010) . In extra-intestinal Escherichia coli infections a significant association between ESBL phenotype and multiple virulence factors has been reported (Sharma et al., 2007). An association between high alginate production and higher susceptibility to antibacterial agents and lower β-lactamase production in P. aeruginosa isolates from cystic fibrosis patients has been documented (Ciofu, 2003). Expression of many virulence factors especially those involved in cell invasion such as colonial morphology, adhesion and aggregation in many gram negative bacteria including P. aeruginosa are regulated by phase variable mechanism (Deziel et al., 2001). Production of virulence factors could be also affected by the medium composition and environmental conditions. For example the production of rhamnolipid is reported to be higher in a medium containing an organic nitrogen source and high phosphate concentration (Mehdi et al., 2011). Although, this study was performed under similar growth conditions, testing the virulence factor under different conditions is recommended.
It is possible that P. aeruginosa clinical isolates producing ESBL regulate the expression of virulence factor by switching the production of certain factors on and off in special environment conditions in favor of their growth, survival and invasion into host tissue.
CONCLUSION
In conclusion, although the in vivo reaction of the bacteria or growth under different growth condition may affect the production of the virulence factors greatly, the in vitro experiments showed the more hydrophobic and stronger biofilm forming ESBL+ isolates, were significantly more resistant to antibacterial agents then the ESBL-isolates. The ESBL+ isolates showed a higher activity for LasA but the production of pyoverdin, LasB and MS was significantly lower in these isolates compared to ESBL- isolates, suggesting the complexity of virulence factors expression under in vitro experiments. Study on the genes involved in the production of these factors and expression of them under different environmental condition may help to understand the possible co regulatory mechanisms which might be involved in the expression and their correlation with ESBL phenotype.
ACKNOWLEDGMENTS
We thank Mozhdeh Razavi for her assistance about bacterial isolation and identification. This investigation was financially supported by Kerman University of Medical Sciences (grant no 88/24).