Prevalence and Antimicrobial Susceptibility of Vancomycin Resistant Staphylococci in an Egyptian University Hospital

Staphylococci can cause many nosocomial and community acquired infections with high rates of morbidity and mortality. The large scale of spread of resistance to vancomycin and other antibiotics has been perceived as a fearsome threat to the already challenging therapy of staphylococci. A total of 982 clinical samples were obtained from different departments of Tanta university hospital. Microscopical examination and standard biochemical tests revealed that 437 isolates were staphylococci. Resistance to vancomycin was determined using disk diffusion and agar dilution methods which revealed that 89 were vancomycin resistant staphylococci (VRS). The susceptibility of VRS isolates to 15 different antimicrobial agents was performed using agar dilution method. All VRS isolates were multidrug resistant (MDR). Polymerase chain reaction (PCR) studies were performed on selected staphylococci isolates with different vancomycin MICs ranging from 2 to 512 μg/ml for detection of vancomycin resistance genes (van genes). The vanA gene was detected in VRS isolates only with vancomycin MICs from 32 to 512 μg/ml. Neither vanB nor vanC gene was detected. β-Lactamase production by VRS isolates was investigated. A total of 88.7% of isolates produced β-lactamase enzyme. Disk approximation test was performed on all VRS isolates that were resistant to erythromycin and sensitive to clindamycin. Out of the 29 isolates on which the test was performed, 24 (82.8%) tested positive. Ciprofloxacin resistant VRS isolates were selected to study the presence of efflux mechanism of resistance. All the tested isolates were efflux positive. The VRS isolates in hospitals as well as in community are alarming to the clinicians and multi drug resistance in these isolates is very dangerous. In this study we tried to investigate the prevalence of VRS isolates in Tanta and to determine their antibiotic susceptibility.


Introduction
Staphylococci have been recognized as an important cause of human diseases for more than 100 years [1]. It is recognized as a cause of a lot of infections which range from minor skin infections to devastating septicemia and endocarditis [2]. Moreover, antibiotic resistance has complicated the treatment procedures. The resistance causes more costs and it may result in failure of treatment. The first clinical isolate of vancomycin resistant Staphylococcus aureus (VRSA) was isolated in japan in 1997. Since then, vancomycin resistant staphylococci (VRS) have spread with very high rapidity and are now detected in hospitals in most countries [3]. Resistance among VRS isolates to multiple classes of antimicrobials is a major threat to patient care. This has stimulated more efforts to understand the mechanisms of antimicrobial resistance. It was found that this problem has been attributed in part to inappropriate use of antimicrobials and failure among healthcare providers to comply with infection control precautions [4]. Significant controversy still exists regarding the current and the future role of vancomycin in the treatment of serious staphylococcal infections. It was of interest to investigate VRS isolates recovered from Tanta area, antimicrobial susceptibility testing, polymerase chain reaction (PCR) for analysis of van genes, βlactamase production, inducible clindamycin resistance, efflux mechanism of resistance was analyzed in this study. Although there is a considerable progress in the understanding of vancomycin resistance in staphylococci, more research should be done to find the ideal treatment and control of multi drug resistant VRS infections.

Identification of staphylococci
The specimen of each patient was cultured into nutrient broth and incubated overnight at 37°C. Then, 10 µl of the broth was plated on agar plates containing horse blood 5% v/v and mannitol salt agar (MSA) media then incubated at 37°C for 24 h. Identification of the developed colonies was depended on morphology and biochemical tests according to standard criteria [5].

Screening of VRS isolates
Kirby-Bauer disk diffusion method was carried out using vancomycin disk (30 µg) (Oxoid). The testing conditions and interpretation of results was done according to CLSI [6] criteria. Briefly, a swab was dipped into the bacterial suspension and spread over the dried agar plate. Three to five min after the surface of agar plate has been inoculated; the vancomycin disk was applied and gently pressed onto the agar using sterile forceps. The plates were inverted and incubated at 37°C for 24 hrs. Zone of growth inhibition around the disc was measured in 2 different directions.

Antimicrobial sensitivity testing
MICs of 15 different antimicrobial agents including: ampicillin, oxacillin, ampicillin/sulbactam, cefotaxime, gentamicin, amikacin, rifampicin, chloramphenicol, sulphonamide/trimethoprime, erythromycin, azithromycin, clindamycin, ciprofloxacin, linezolid and vancomycin (Oxoid) against vancomycin resistant staphylococci isolates were determined by agar dilution method according to the procedure described by CLSI [6] as follow, aliquot of the prepared suspension of each isolate was transferred by automatic pipette to a certain well of a sterile multiinoculator in the seed plates (containing different concentrations of antimicrobial agents) according to a configuration of a record key. After (18)(19)(20)(21)(22)(23)(24) hrs of incubation at 37°C, the presence of growth on the plate containing the lowest concentration of the given antimicrobial agent indicates the minimum inhibitory concentration (MIC) value of this antimicrobial against the tested isolate.
A 25 µl reaction volume consisting of 12.5 µl master mix, 2.5 µl primer mix (0.2 µM of each primer) and 3 µl of DNA template and 7 µl of RNase free water was prepared. DNA samples were subjected to thermocycling conditions with initial inactivation step (95°C,15 min) with three step cycling condition of denaturation (94°C, 30 sec), annealing (60°C, 90 sec) and extension (72°C, 90 sec) for 35 cycles with final extension (72°C,10 min) and soak at 4°C. Then 5 µl of amplified products were mixed with 2 µl of ethidium bromide (Fermentas) and loaded on a 1.8% agarose gel (Amresco) along with the 1 kb DNA Ladder (Fermentas) and electrophoresis at 100 volt for 60 min and visualized under UV transilluminator (Bio-Doc analyzer, Biometra).

Detection of β-lactamase production
It was carried out using iodometric overlay method as described by Abo-Kamar and Shohayeb [9]. VRS isolates were grown first onto plates containing 2 µg/ml penicillin G, then they were tooth picked onto the surface of nutrient agar plates. After overnight incubation at 37°C, the plates were over laid with 1% molten agarose containing 0.2% soluble starch, 1% penicillin G. They were incubated for 15 min at room temperature and iodine solution was poured onto the agar plates to cover it uniformly, after 10 sec, the residual iodine solution was damped out and the plates were incubated at room temperature until discoloration zones appeared around β-lactamase producing colonies within 5 min.

Detection of inducible clindamycin resistance among VRS isolates using the disk approximation test (D-test)
It was carried out according to the method reported by Fiebelkorn et al. [10]. A 15-µg erythromycin disk was placed 20 mm apart from a disk containing 2 µg clindamycin. The procedures of disk diffusion testing were performed according to CLSI [6]. Positive result was recorded if zone of inhibition of clindamycin was flattened in the area adjacent to erythromycin disk giving a D shape to the zone (D-zone effect).

Studying efflux mechanism in ciprofloxacin resistant VRS isolates
Efflux of ethidium bromide was measured fluorimetrically as described by Mullin et al. [11]. with some modifications. In nutrient broth containing 20 µg/ml reserpine (efflux inhibitor) and 0.5% glucose, overnight culture of bacterial cells was loaded with ethidium bromide (10 µg/ml) for 20 min. Then, cells were centrifuged, and the pellet was resuspended to an OD600 of 0.2 in fresh growth medium alone or in presence of reserpine for 10 min. Release of ethidium bromide from the cells was monitored by the decrease in fluorescence of cell suspension at excitation λ of 600 nm and an emission λ of 530 nm.

Results
A total of 982 clinical samples were collected from different departments of Tanta university hospital during the period from October 2011 to August 2012. All samples were cultured on nutrient agar and examined microscopically. Gram +ve bacteria were subjected to biochemical identification which revealed that 437 isolates were staphylococci. A total of 401 (91.8%) of these isolates were S. aureus and the remaining 36 (8.2%) were coagulase negative staphylococci

Preliminary screening of vancomycin resistance among Staphylococci isolates
All the recovered (437) staphylococci isolates were subjected to disk agar diffusion method for preliminary screening of vancomycin resistant staphylococci (VRS) and 89 (20.3%) VRS isolates were detected. Distribution of vancomycin resistant S. aureus (20.1%) and vancomycin resistant CoNS (0.2%) among the recovered staphylococci isolates is shown in Figure 1.

Antimicrobial sensitivity testing
The susceptibility of 89 VRS isolates to 15 different antimicrobial agents was performed using agar dilution method as shown in Figure  2. The antimicrobial resistance pattern is shown in Table 1 in supplemental information. In this study all VRS isolates exhibited MDR character according to Magiorakos et al. [12].

Detection of vancomycin resistance determinants (van genes) by PCR
The selected staphylococci isolate for PCR studies were representatives from each vancomycin MICs ranging from 2 to 512 µg/ml. Traditional PCR was performed on total DNA extract of each selected isolate for detection of vanA, vanB and vanC genes. The amplified products were electrophoresed, and the resultant gels were stained by ethidium bromide and the amplicons were visualized using UV-transilluminator. Bands with approximate size of 1032 bp for vanA gene, 647 bp for vanB gene and 815 bp for vanC gene were detected as illustrated in Figures 3. vanA gene was detected in 5 out of the 9 isolates with MICs ranging from 32 to 512 µg/ml, where vanB and vanC genes were absent in all the tested isolates.

Detection of β-lactamase production
All VRS isolates were resistant to ampicillin therefore; they were screened for production of β-lactamase enzymes using iodometric overlay method. β-lactamase producers were detected by the formation of clear zones visualized around their growth in dark blue background of the medium. It was found that 79 out of the 89 VRS isolates (88.7%) produced β-lactamase enzyme.

D-test
VRS isolates resistant to erythromycin and sensitive to clindamycin were selected for this test. Out of 29 isolates on which the test was performed, 24 (82.8%) tested positive which appeared as flattening of clindamycin zone of inhibition adjacent to erythromycin disk.

Efflux in ciprofloxacin resistant VRS isolates
The ciprofloxacin resistant VRS isolates were selected for this test. The MICs of ciprofloxacin and ethidium bromide against the selected isolates were determined. All tested isolates showed MIC values ranged between 16 and 512 µg/ml. On the other hand, the MIC for the sensitive control strain (S29231) was <0.5 µg/ml for both tested agents.
Reserpine (efflux inhibitor) was used to study its effect on MICs of ciprofloxacin and ethidium bromide. It was found that reserpine decreased the MICs by 7-10-fold changing it to the sensitive range as shown in Table 2 in supplemental information.   isolate with ciprofloxacin MIC=512 µg/ml. b) S144 isolate with ciprofloxacin MIC=256 µg/ml. c) S140 isolate with ciprofloxacin MIC=128 µg/ml. d) S155 isolate with ciprofloxacin MIC=64 µg/ml.

Discussion
All the recovered (437) staphylococci isolates were subjected to disk agar diffusion method for preliminary screening of vancomycin resistance. Out of these, 89 (20.36%) VRS isolates were detected. This finding was comparable to the percentage reported by Ghoniem [13] in Egypt which was 20.68% and lower than the percentages reported by many authors [14] 76.9% in South Africa; [15] 37% in Iran; [16] 30% in Saudia Arabia and [17] 26.67% in India. On the other hand, this finding was higher than that reported by Olivo et al. [18] in Brazil and Moses et al. [19] in Nigeria which were 9.3% and 5.4%, respectively. The relatively high rates of VRS isolates in this study could be partially explained by a selection pressure induced by an inadequate use of antimicrobials [20]. It was found that out of the tested 89 VRS isolates there were 88 S. aureus (98.88%) and only one isolate (1.12%) was S. epidermidis. [21] reported that 100% of VRS were S. aureus. However, vancomycin resistant S. epidermidis was reported by Sanyal and Pienheiro [22,23]. The glycopeptide antibiotic vancomycin is often the sole remaining antibiotic active against staphylococcal infections. As such, the acquisition of glycopeptide resistance by staphylococci has been anticipated with apprehension [24][25][26][27]. In our study, the results of disk diffusion test for determination of vancomycin resistance were in accordance with that of agar dilution as 89 VRS isolates were detected in either of them. In this study, 231(52.86%) isolates were VSS (MIC ≤ 4 μg/ml), 117 (26.77%) were VIS (MIC 8-16 μg/ml) and 89 (20.36%) were VRS isolates (MIC ≥ 32 μg/ml). [13] reported that 67% of Staphylococci isolates were VSS, 12% were VIS and 21% were VRS. On the other hand Tiwari and Sen [24] reported that 86% of staphylococci isolates were VSS, 9.3% were VIS and 4.7% were VRS. The emergence of vancomycin resistance in staphylococci isolates has wide implications for the clinicians as this will increase the morbidity and mortality rate due to VRS infections. Susceptibility testing to other antimicrobials was performed to identify which one may be expected to have useful activity against VRS to help in the management of the infections caused by these isolates. So, the susceptibility of VRS isolates to 15 antimicrobial agents was carried out in the present study. Interestingly, all these VRS isolates showed resistance to a minimum of 10 antimicrobials including vancomycin, a finding that is comparable to what was detected by Adigoke [14] which was 9 antimicrobials including vancomycin. Also, Thati et al. [24] found that all VRS isolates were resistant to at least six antimicrobials other than vancomycin. All VRS isolates in the present work were MDR isolates and this phenomenon of multiresistance among VRS isolates was also reported by several authors [13,[24][25][26][27]. In general, the problem of antimicrobial resistance has multiple implications with respect to outcomes. Persons infected with multiresistant VRS may face higher mortality rates [28]. This is partially due to the fact that infection with multiresistant VRS isolates is associated with the risk for initially inappropriate antimicrobial therapy, a major determinant of patient survival in serious infections [28,29]. Moreover, the high risk for death may reflect that multiresistant VRS have more virulence factors. Beyond mortality, resistant pathogens contribute to escalating hospital costs [28,29]. All isolates (100%) were resistant to ampicillin and cefotaxime followed by ampicillin/sulbactam (88.8%) and oxacillin (87.6%). Sulphamethoxazole/trimethoprime (cotrimoxazole) and linezolid showed the least incidences of resistance, where only 43.8% and 37% of the VRS isolates, respectively, were resistant to them. A finding which is also reported by many other authors in this field [13,25]. Therefore, these drugs could be used for therapy of these infections. However, until more experience exists, the best treatment strategy would appear to be to tailor therapy based on the susceptibility of the isolate in each case.
Detection of van genes in the tested isolates was performed using PCR technique. The selected isolates were representatives for each vancomycin MICs ranging from 2 to 512 µg/ml. Band of vanA gene (1032 bp) was detected in VRS isolates only and it wasn't found neither in VSS nor VIS isolates. On the other hand, bands of vanB (647 bp) and vanC (815 bp) genes were absent in all tested isolates. [24] reported that out of 7 VRS isolates, 6 (85.7%) contained vanA gene and didn't contain neither the vanB nor vanC genes and El-Daker [21] reported that 100% of VRS isolates possessed vanA gene only. On the other hand, Tiwari and Sen [30] reported van genes-negative VRS isolates.
Resistance mechanisms to some other antimicrobials were studied including β-lactamase production, detection of inducible clindamycin resistance mediated by erm gene using the disk approximation test (Dtest) and the efflux mechanism of resistance. Liang et al. [31] in his study on staphylococci, reported that one of the important mechanisms of resistance to β-lactams was the expression of βlactamases which hydrolyze β-lactam ring of penicillins and cephalosporins, thereby making these antimicrobials inactive [32]. In the present study, 88.7% of staphylococci isolates were β-lactamase producers using iodometric overlay method. This percentage was higher than that reported by Arslan and Ozkarde et al. [33] in Brazil and Gad et al. [34] in Egypt which were 74.4% and 55%, respectively.
Clindamycin resistance is either constitutive or inducible following exposure to a macrolide. Induction tests (D-test) utilize closely approximated erythromycin and clindamycin disks; the flattening of the clindamycin zone of inhibition adjacent to the erythromycin disk indicates inducible clindamycin resistance [10]. In the present study, detection of inducible clindamycin resistance using the disk approximation test (D-test) was performed on all VRS isolates that were resistant to erythromycin and sensitive to clindamycin. This test appears to be a reliable indicator of the presence of erm gene which encodes enzymes that confer inducible resistance to lincosamides (clindamycin) by methylation of the 23S rRNA [10]. Out of 29 isolates on which the test was performed, 24 (82.8%) tested positive. Higher result was reported by Thati et al. [24] who found that 100% of the isolates exhibited inducible clindamycin resistance. On the other hand, lower results were reported by Rice and Sexena et al. [35,36] where they found that only 18.6% and 25.8%, respectively of isolates had inducible clindamycin resistance.
Fluoroquinolone antibiotics are important class of antimicrobials that exhibit a broad spectrum of antibacterial activity [37]. The efflux of fluoroquinolones and the expression of multidrug transporters have been demonstrated for staphylococci, in which the major role in fluoroquinolone efflux is done by the membrane transporter NorA [38]. NorA promotes the active efflux of a wide variety of organic compounds like ethidium bromide with fluoroquinolone antibiotics being one of the best transporter substrates [38]. In fact, reserpine was discovered to reverse NorA-mediated drug resistance [39] so it was used as efflux inhibitor.
In this work, the efflux mechanism in 18 ciprofloxacin resistant VRS isolates was studied using fluorometric assay. Efflux mechanism was detected in all of these isolates where they showed accumulation of the fluorescent dye ethidium bromide inside the cells that remained relatively higher when the cells treated with the efflux inhibitor reserpine. Reports of Mullin et al. and Gadd et al. [11,34] involved fluorimetric method to study the efflux mechanism of resistance and the effect of reserpine on the accumulation of ethidium bromide in VRS isolates.

Conclusion
Finally, the findings of this study indicated that infection by VRS isolates has started to spread in Tanta area as up to 20.4% of staphylococci isolates were VRS. It also appears that there are few treatment options for VRS infections because they tend to be multi drug resistant against a large number of currently available antimicrobial agents. The VRS isolates exhibited resistance to antimicrobials through several ways indicating that different resistance mechanisms may interact to increase the level or spectrum of resistance of such organisms.