Emerging of antimicrobial resistance in staphylococci isolated from clinical and food samples in Algeria

Objective The antimicrobial resistance of staphylococci rose worldwide. In total, 96 Staphylococcus isolates from food and clinical samples were collected from two provinces in Algeria. The antimicrobial susceptibility testing was performed and resistance-associated genes were detected. Results Fifty-one strains were isolated from food samples and differentiated into 33 Staphylococcus aureus and 18 coagulase-negative staphylococci. Forty-five staphylococci were collected from hospital and community-acquired infection cases. All S. aureus isolated from food were resistant to penicillin and 45.5% were resistant to tetracycline. The resistance rates of 45 clinical Staphylococcus isolates were 86.7%, 48.9%, 37.8% and 20.0% to penicillin, tetracycline, erythromycin and kanamycin, respectively. Nine isolates were confirmed as MRSA from food and clinical isolates. One S. aureus originated from food was confirmed as vancomycin-resistant. Multidrug-resistance was observed among 25.5% and 53.3% of food and clinical staphylococci, respectively. The tetM/K, blaZ, aacA-aphD, ermC and mecA genes were detected in food and clinical isolates. ermA gene was not found. This study provided insight into the status of antimicrobial resistance of staphylococci isolated from food and clinical samples in Algeria. Further investigations and surveillance programmes are mandatory.

The mechanism of antimicrobials resistance in staphylococci was due to the acquisition of mobile genetic elements like plasmids and/or transposons [7][8][9][10]. Penicillin-resistance was due to the production of betalactamases which is encoded by the blaZ gene located chromosomally or on plasmids [11,12]. Methicillinresistant Staphylococcus aureus (MRSA) where resistance is encoded by mecA gene is an important cause of human nosocomial infections worldwide [13,14]. In S. aureus and CoNS tetracycline resistance is mediated by ribosomal encoded tetM gene and/or by tetK encoded efflux protein [15][16][17]. Ribosomal target modification, mediated by the presence of ermA, ermB and ermC in S. aureus and CoNS is associated with resistance to macrolides, lincosamides and type B streptogramins [17,18]. For aminoglycoside resistance aacA-aphD genes are conferring cross-resistance to aminoglycosides such as gentamicin, tobramycin, kanamycin and amikacin [19][20][21].
Food is considered as an important vehicle for the spread of antibiotic-resistant bacteria [6,22]. In meatproducing animals such as cattle, poultry and swine, antibiotics are mainly used for the treatment and prevention of several bacterial diseases which may lead to arise of antimicrobial resistance in various bacteria like Campylobacter, Escherichia coli, Salmonella and Staphylococcus [23].
In Algeria, antimicrobial susceptibility of S. aureus was commonly investigated. [24][25][26]. Less attention was given to S. aureus and CoNS in community-acquired infections or those isolated from food samples [27].
The aim of this study was to determine phenotypic antimicrobial resistance and associated genes of staphylococci isolates from clinical and food samples in Algeria.
Forty-five S. aureus isolates from clinical samples like pus, sperm, urine, vaginal discharge, wounds, catheter tips and secretions were kindly provided from Mohamed Boudiaf hospital, Médéa.

Bacterial isolation and identification
Microbiological analysis of food samples was done according to NF EN ISO 6888-1/A1 standard procedure according to the IOS [28] and for clinical samples was done by the method described previously [29]. The identification of coagulase-positive S. aureus (CoPS) was made by biochemical tests (Rabbit plasma, Oxoid, Dardilly, France). Identity of S. aureus was confirmed by agglutination test (Bio-Rad, Marnes-la-Coquette, France).

Susceptibility test of antimicrobial agents
The antimicrobial susceptibility testing to 12 antimicrobial agents was performed using the disc diffusion test according to CLSI recommendations [30] and the guidelines established by the Antibiogram Committee of the French Microbiology Society [31]. The antibiotic discs (Bio-Rad; Oxoid) were used according to instructions of the manufacturer's (Table 1). Staphylococcus aureus ATCC 25923 was used as quality control.

Detection of resistance-associated genes by PCR
Genomic DNA was extracted using phenol/chloroform extraction method [32].
The DNA fragments were visualized using an UV transilluminator (EC3, UVP BioImaging Systems, Cambridge, UK) and read using AlphaEaseFC software (Genetic Technologies Ltd., Fitzroy, Australia).

Statistical analysis
Possible relationships between the presence of resistance genes among S. aureus and CoNS isolated from food and clinical samples were statistically analysed using Fisher's exact test. P-value of < 0.05 was considered as statistically significant difference.

Results
Ninety-six staphylococci were isolated, identified and classified as 51 staphylococci (33 S. aureus and 18 CoNS) from food samples, 27 staphylococci (21 S. aureus and 6 CoNS) from community-acquired clinical samples and 18 S. aureus from hospital-acquired clinical samples, respectively.
Out of 51 staphylococci isolated from food samples, 49 staphylococci (33 S. aureus and 15 CoNS) were resistant to at least one tested antibiotic (96.1%) ( Table 1). The resistance rates to penicillin and tetracycline were 94.1% and 49.0%, respectively. Two S. aureus (6.1%) and three CoNS (16.7%) were phenotypically resistant to methicillin by disc diffusion test while MIC determination by E-test confirmed only one isolate as MRSA and another as methicillin-resistant CoNS.
Vancomycin resistance was detected in one S. aureus (3.0%). The resistance rates of CoNS to clindamycin and erythromycin were 33.3% and 27.8%, respectively. All staphylococci isolated from food samples were susceptible to gentamicin.
The antimicrobial resistance rates were remarkably higher in clinical isolates than in staphylococci isolated from food samples ( Table 1). The resistance rates of S. aureus and CoNS isolated from clinical samples to penicillin were 92.3% and 50.0%, respectively while those to tetracycline were 48.7% and 50.0% for S. aureus and CoNS, respectively. The prevalence of MRSA obtained by disc diffusion test was 38.9% and 52.4% for hospital and community-acquired isolates, respectively. However, the detected MRSA rate by E-test was 11.1% and 23.8%. Staphylococcus aureus showed considerable resistance to erythromycin for both hospital-acquired (27.8%) and community-acquired isolates (52.4%). Staphylococcus aureus isolates from hospital were susceptible to gentamicin, however, those isolated from community infection showed 14.3% resistance to gentamicin. All clinical staphylococci isolates were susceptible to vancomycin.   The antibiotic resistance-associated genes were demonstrated in Table 2. In food isolates tetM was the most prevalent gene detected in S. aureus and CoNS with 66.7% and 88.9%, respectively. Both tetM and tetK genes were detected in 11 S. aureus (33.3%) and 4 CoNS (22.2%). In food samples, mecA gene was observed in 15 S. aureus (46.9%) and 10 CoNS (88.9%) isolates, respectively. Six S. aureus (18.2%) and 3 CoNS (16.7%) isolates harboured aacA-aphD gene associated with gentamicinresistance. Fifteen S. aureus (46.9%) and 10 CoNS (88.9%) isolates possessed blaZ gene encoding resistance to penicillin. No significant difference for the detection of antibiotics resistance genes was observed between S. aureus and CoNS isolated from food samples (P > 0.05).
Regarding staphylococci from clinical samples, mecA was the most detectable resistance-associated gene in S. aureus originated from the hospital with 17 isolates (94.4%) and in the community with 19 isolates (90.5%). The tetM gene was found in 10 hospital isolates (55.6%) and 17 community-acquired isolates (81.0%). The tetK and tekM genes were detected together in 10 S. aureus isolates (25.6%). Six S. aureus from the hospital (16.7%) harboured ermC gene. The aacA-aphD gene was found by PCR in 12 S. aureus isolates (3 from hospital and 9 were community-acquired). The blaZ gene detection rate was significantly higher in hospital-acquired S. aureus with 15 isolates (81.0%) than in community-acquired isolates with only 6 (55.6%) (P = 0.025). No statistical differences were observed for the detection of remaining antibiotic genes.
The correlation between phenotypic resistance and detection of resistance-associated genes for food and clinical isolates was demonstrated in Table 3. In S. aureus from food samples correspondence between phenotypic resistance and associated genes was 45.5% and 50.0% concerning penicillin and methicillin, respectively. For tetracycline resistant isolates 46.7% and 53.3% harboured tetK and tetM, respectively. Both genes were detected in  33.3%. In clinical isolates the correlation between resistance and mecA gene was very high with 100% and 92.8% in hospital-acquired and community-acquired S. aureus.

Discussion
The results of this study showed less prevalence of S. aureus resistant to oxacillin in food isolates (6.1%) compared to those reported previously 67.5% [35], 38.0% [36] and 17.3% [37]. The prevalence of oxacillin-resistance in clinical isolates was slightly similar to previous study in Algiers [38] but lower than another study in eastern Algeria [39].
Staphylococcus aureus from food and clinical samples were phenotypically oxacillin-resistant with 6.1% and 46.2%, respectively. From isolates 46.8% and 92.3% of them harboured the mecA gene, respectively. In contrast, 38.0% of Portuguese S. aureus strains from food samples were resistant to oxacillin but only 0.7% of them harboured mecA gene [36] and in Libya, mecA gene was not detected by PCR in MRSA isolated from hospitalized patients [40].
In Morocco, Kenya, Nigeria and Cameroun relatively high occurrence rates of MRSA were reported [41]. In Algeria, the majority of MRSA did not represent a high multidrug resistance rate [41,42].
All isolates from food were resistant to penicillin which was in agreement with previous reports [37,43,44] but significantly higher than reported in other studies [35,36,45].
The presence of the blaZ gene did not influence production of β-lactamase alone [50,51]. Here, not all penicillin-resistant S. aureus isolates from food (45.5%) and clinical samples (52.8%) possessed blaZ gene.
All S. aureus isolates from food samples were susceptible to gentamicin, which agrees with several reports [36,52,53]. In all clinical isolates, the resistance rate of S. aureus to gentamicin was 7.7% which was similar (7.0%) with a study performed in Algeria [25]. In Algeria, no resistance to gentamicin was found in staphylococci in milk [54,55]. One S. aureus isolate exhibiting phenotypic resistance to gentamicin harboured the aacA-aphD gene which is in contrast to previous studies [16,33]. In this study, aacA-aphD was detected more often in staphylococci isolates from clinical (31.1%) than those of food origin (17.6%).
Staphylococcus aureus isolates from food were slightly resistant to erythromycin (9.1%) which was higher than 5% reported previously in Portugal [36], but significantly lower than detected in China and Turkey [37,56].
The resistance of clinical S. aureus to erythromycin was higher than reported previously in Algerian hospitals [25]. The ermA gene encoded erythromycin-resistance could not be detected in both S. aureus and CoNS from clinical and food origin. In contrast, high rate of ermA and ermC genes in MRSA isolates from hospitalized patients was determined [57]. The correlation between phenotypic susceptibility testing and detection of ermC gene was 31.3% which was lower than obtained previously [16,33]. In Libya, ermA and ermC genes could not be detected in erythromycin-resistant S. aureus [40].
The resistance of S. aureus isolated from food samples to tetracycline was higher than reported formerly in China [56], in Italy [6], in Portugal [36] and in Turkey [45]. Tetracycline resistance of S. aureus isolated from clinical samples was considered lower than reported before [25,48]. There was no significant difference between tetracycline resistance in staphylococci isolated from food and clinical samples. In contrast, a higher resistance rate to tetracycline in food isolates compared to clinical isolates was described [52]. The tetK and tetM genes were detected in 53.3% and 60.0% in resistant S. aureus from food samples, respectively. In clinical S. aureus isolates, the correlation rate between phenotypic resistance and prevalence of tetM and tetK genes was 72.2% and 66.7%, respectively. The discrepancy between phenotypic resistance to tetracycline and resistance determinants may attributed to other mechanisms [15,58].
Vancomycin was the most effective antimicrobial agent against MRSA [41,59]. In this study, three S. aureus strains from food showed resistance to vancomycin. In Turkey, 21.7% of S. aureus isolated from food samples were resistant to vancomycin [37].
Most of CoNS isolates were resistant to penicillin, tetracycline and erythromycin. In addition, 16.7% of CoNS isolated from food samples were confirmed phenotypically as methicillin-resistant CoNS (MR-CoNS). The mecA gene was detected in 75.0% of isolated CoNS which was in agreement with previous study [33]. The ermC gene was detected only in one CoNS isolate (5.6%) which was in accordance with formerly reports [33,60]. In contrast, staphylococci isolated in Turkey showed high detection rate of ermC and ermA [61,62].
In conclusion, the high prevalence of resistance to penicillin, tetracycline and erythromycin was particularly alarming. The knowledge about antibiotic resistance of staphylococci originated from daily food from markets and of community-acquired S. aureus is not fully addressed in Algeria, yet. Information concerning resistance in CoNS is very limited in this country. Vancomycin-resistant isolates were found in food isolates not in clinical ones. This study considered a first impression on existing situation concerning antibiotic