Introduction

Resistance to β-lactam antibiotics dates back to the first years of discovery of resistance to the first antibiotic, penicillin. The first β-lactamase was observed in Escherichia coli bacteria which hydrolyzed penicillin (1). The emergence of resistance to β-lactam antibiotics is related to beta-lactamases coding by plasmids which are rapidly spread from strain to strain among most clinical isolates (2). Until now, over 200 different types of extended-spectrum β-lactamase (ESBL) have been identified worldwide, the majority of which are found in Enterobacteriaceae family (3). Most ESBLs are derivatives of TEM (Temoneira) or/and SHV (sulphydryl variable) enzymes (4, 5). In recent years, a new family of plasmid-mediated ESBLs, called CTX-M (cefotaxime-hydrolyzing β-lactamase), has been

arisen that preferentially hydrolyzed cefotaxime (3). CTX-M was reported in 1989 for the first time in Germany (6) and is often found in E. coli                         and Klebsiella pneumoniae as well as in other Enterobacteriaceae (3). PER-1 (Pseudomonas extended-resistant) was first detectedin Pseudomonas aeruginosa(7) and later in Salmonella entericaand Acinetobacter isolates. PER-2, which shows 86% homology to PER-1, hasbeen reported in S. enterica, E.coli, K. pneumoniae, Proteus mirabilis, and Vibrio choleraO1 El Tor (8, 9). Currently, microbial resistance through ESBL has been recognized globally and now ESBLs are a problem throughout the world (3). Although some studies have been carried out to detect ESBL-producing bacteria in Mashhad (10, 11), there are not enough data on molecular properties of bla_CTX-Mand bla_PER

genes among Enterobacteriaceae bacteria in this area. Therefore, the present study was undertaken in two teaching hospitals of Mashhad University to determine genetic properties ofESBL-producing Enterobacteriaceae isolates related to bla_CTX-Mand bla_PER genes.

 Materials and Methods

Bacterial isolates and antibiotic susceptibility testing

During six months, from January to June 2012, 100 Enterobacteriaceae were collected from urinary samples of inpatients and outpatients referred to Qaem Hospital and 17-Shahrivar Hospital Laboratories in Mashhad. All isolates were identified by differential biochemical standard tests. Antibiotic susceptibility testing was applied by disk diffusion method for ampicillin (AP, 10 µg), cephalothin (KF, 30 µg), nalidixic acid (NA, 30 µg), gentamicin (Gm, 10 µg), amikacin (AK, 30 µg), cefotaxime (CTX, 30 µg), ceftazidime (CAZ, 30 µg), nitrofurantoin (NI, 300 µg), co-trimoxazole (TS, 25 µg), and imipenem (IMI, 10 µg) disks (MAST, UK) on Mueller-Hinton agar (Microbiology, Germany) according to Clinical and Laboratory Standards Institute (CLSI) guidelines (12).

 Phenotypic ESBL detection

Screening for ESBL producers was carried out by double disk approximation test and ESBL confirmative test according to CLSI standards. Double disk approximation test was employed by augmentin (20 + 10 µg), cefotaxime and ceftazidime disks, and ESBL confirmation test was tried out by ceftazidime/clavulanate (10 + 30 µg) combination disk and ceftazidime (13, 14). A positive test defined as ³5 mm increase in zone diameter in comparison to a disk without clavulanic acid (13, 14). Genus and species of ESBL-producing bacteria were confirmed by Microgen kit and Microgen identification system software (Microgen Bioproducts GNA- ID UK).

 PCR amplification of bla_CTX-M and bla_PER genes

Pure bacteria was inoculated in Lauria Bertani broth (Biomark, India) containing 100 µg/ml ampicillin (Sigma, Germany) and incubated at 37°C, shaking 185 rpm for 16-18 hr. Plasmid extraction was done by Perfect Prep-Spin Mini Kit-5 Prime-USA according to procedure instructions. A volume of 5 µl of extracted plasmid (contained 18-20 ng/µl of DNA) was used to perform PCR. Identification of bla_CTX-M gene was conducted with a pair of primer CTX-MU1 (5¢- ATG TGC ACC AGT AAR GT-3¢) and CTX-MU2 (5¢-TGG GTR AAR TAR GTS ACC AGA-3¢) which amplified a 593 bp fragment (15). The specific primers which were used for amplification of bla_PER were PER-forward (5¢-AATTTGGGCTTAGGGCAGAA-3¢) and PER-reverse (5¢-ATGAATGTCATTATAAAAGC-3¢) to amplify a 924 bp fragment (16). PCR was performed in 30 µl mixture of 3 µl 10X buffer, 1 µl of 10 mM MgCl₂, 0.25 µl of 5 U/µl Taq DNA polymerase (Fermentas-Lituania), 0.5 µl of 10 mM of each deoxynucleaotide triphosphate, 1 µl of 10 µM of each primer and 5 µl of plasmid extracts in a thermal cycler (Kyratec- Korea). The amplification was performed according to conditions which are shown in Table 1.

Pseudo. aeroginosa containing bla_PER, and E. coli containing bla_CTX-M received from Pasteur Institute, Iran, were used as positive controls. The PCR products were evaluated after electrophoresis on 1% gel and staining with ethidium bromide. A 100 bp ladder standard (Fermentas, Lithuania) was used as molecular weight ladder.

 Restriction analysis of PCR products and confirmation of the amplified products

Following PCR, the bla_CTX-M PCR products were digested with Taq1 (Fermentas, Lithuania) for 3 hr at 65°C. PCR products were extracted with Agarose Gel Extract Mini Kit-50 Prep (5 Prime, USA) according to procedure guidelines. For restriction enzyme digestion, 6 µl of each PCR products were mixed with 25 µl buffer, 1 µl distilled water and 2 µl Taq1 restriction enzyme. The restriction enzyme was selected by using CLC Main workbench 5 software.

 DNA sequencing analysis

The PCR products of two samples that showed the highest multi-resistance to tested antibiotics were sequenced on both strands by using CTX-MU-1 and CTX-MU-2 primers with an ABI 3730 XL automated DNA sequencer (Macrogen, Korea). The nucleotide sequences were analyzed using the Sequencher sequence software alignment (Version 4.10.1) and were compared to the identified β-lactamase: CTX-M-15 gene from K. pneumoniae plasmid pMRC151 in the GenBank nucleotide database under accession no. AY995205 available on the Internet at the National Center of Biotechnology Information website (http://www.ncbi.nlm.nih.gov).

 Table 2. Antibiotic susceptibility of isolated Enterobacteriaceae family bacteria to various antibiotics

Statistical analysis

Statistical analysis was carried out by using Statistica software. Chi-square test was used for determination of significance of association. The P-value ≤ 0.05 was considered significant.

 Results

One hundred bacterial isolates of the Enterobacteriaceae family were detected from urinary samples of patients. There were 67 E. coli, 19 K. pneumoniae, 5 K. oxytoca, 5 Enterobacter cloacae, 2 Proteus mirabilis, 1 Pro. vulgaris, and 1 Citrobacter diversus. Among them, E. coli was the most common isolated microorganism followed by K. pneumoniae.

Fifty seven urinary samples were from inpatients and 43 samples were from outpatients. Table 2 shows the results of disc diffusion susceptibility test for isolated bacteria. As Table 2 indicates, there is high prevalence of resistance and intermediate to ampicillin (88%) and cefotaxime (71%). Moreover, the most susceptibility was for imipenem (100%) and nitrofurantoin (96%) among the tested antibiotics.

Compared to K. pneumoniae, a higher percentage of E. coli isolates were resistant to nalidixic acid (81.8% vs. 9.1%), co-trimoxazole (6.1% vs. 33.4%),

Figure 1. PCR products of CTX-M type extended-spectrum             β-lactamases producing isolates on 1% agarose gel (Pos: positive control, Neg: negative control, M: marker 100 bp)

 cefotaxime (75% vs. 25%) and cephalothin (6.2% vs. 26.1%); while resistance to ceftazidime was higher in K. pneumoniae in comparison to E. coli (60% vs. 40%).

The antimicrobial susceptibility results of inpatients and outpatients are shown in Table 3. Antimicrobial resistance among hospitalized patients was higher than that of outpatients.

ESBL production was observed in 27% of isolated bacteria by phenotypic confirmatory test. Twelve (27.9%) of 43 isolated bacteria from inpatients and 15 (26.3%) of 57 bacteria from outpatients were ESBL-producing. Among isolated Enterobacteriaceae, the frequency of ESBL production for K. pneumoniae and E. coli was 47.4% and 17.9%, respectively. A high percentage of ESBL producers were resistant to co-trimoxazole, gentamicin and nalidixic acid (Table 4). Resistance to gentamicin and co-trimoxazole was associated with ESBL production (P-value < 0.01).

Among seven bacteria with multi-resistance to gentamicin, co-trimoxazole and nalidixic acid, 5 (71.4%) were ESBL producers, 4 were K. pneumoniae, and three were E. coli. 

In the present study, 26 isolates harbored the bla_CTX-Mgene and none of them contained bla_PER. Figure 1 shows PCR products on 1% agarose gel.

The restriction enzyme digestion analysis