Comparison between phenotypic and PCR for detection of OXA-23 type and metallo-beta-lactamases producer Acinetobacter spp.

BACKGROUND: Resistance to carbapenems is developing around the world and can cause many problems for treatment of patients. Production of metallo-beta-lactamase (MBL) is one of the main mechanism for this type of resistance. So, detection of MBL-producer microorganisms can prevent the spread of this type of resistance.
MATERIALS AND METHODS: In this study 94 Acinetobacter spp. were investigated. Resistance to imipenem was conducted after purification and identification. Combination disc (CD) and Double Disc Synergy Test (DDST) were performed for phenotypic detection of MBL and the molecular PCR method was done for vim-1, vim-2, imp-1 and OXA-23 genes.
RESULTS: According to TSI, SIM and oxidation-fermentation (OF) test and PCR assay 93 Acinetobacter baumannii and one strain Acinetobacter lwoffii were identified. 85% of them were resistant to imipenem. 34% of them have a positive combination disc test (CD) while Double Disc Synergy Test (DDST) was negative for all of them. The vim-1, vim-2 and imp-1 genes were not detected in PCR molecular method, however in 74% of strains with positive results in combination disc, were positive for the OXA-23 gene after PCR test. This study shows that the blaOXA-23 resistance determinant may become an emerging therapeutic problem. DISCUSSION: According to the results, it seems that combination disc does not have enough specificity for detection of MBL-producer Acinetobacter and using Double Disc Synergy Test (DDST) can be more convenient.

Introduction

The emergence of carbapenem-resistant Acinetobacter spp. in the last decade has become a serious problem in the health community [1], [2]. The ability of producing Carbapenemases enzymes such as metallo-beta-lactamase (MBL) are the most common resistant mechanism to carbapenem. The class B beta-lactamase (MBL) which is hydrolyzing carbapenem and other beta-lactams except Monobactam can cause resistance to this antibiotic family [3], [4], [5]. Transferring of mobile genetic elements like plasmids is the main reason for spreading of this resistance mechanism in gram negative bacteria [6], [7]. Several phenotypic methods for detection of MBL-producer isolates have been suggested by using chelating agents nevertheless the specificity and sensitivity of this method are diverse [8], [9]. The Combination Disc assay (CD) and Double Disc Synergism Test (DDST) by use of imipenem and EDTA as an MBL inhibitor are the most widespread phenotypic methods which are used in numerous studies [10], [11], [12]. Molecular method PCR with different primers have been used for confirmation of phenotypic assay since, blaVIM and blaIMP are the most common MBL, so detection of these gene use for evaluation of phenotypic methods [3]. Because resistance to carbapenems is an increasing problem among Acinetobacter baumannii.

The aim of this study is using the CD and DDST methods as two phenotypic methods and PCR for achieving blaVIM1, blaVIM2 and blaIMP1 as a common MBL gene in imipenem-resistant Acinetobacter baumannii isolated from burned patients.

Methods

Bacterial specimen

In this study 94 Acinetobacter spp. was isolated from burned patients that were hospitalized 3–10 days in Motahari hospital as a referral burn unit in Tehran, Iran. Isolated bacteria were purified in EMB and then collected for species confirmation by specific biochemical tests such as oxidase, motility, OF, etc. PCR was used as a confirmatory test for identification Acinetobacter baumannii strains by specific primers for OXA-51-like gene (Table 1 (Tab. 1)) [13], [14], [15], [16]. Acinetobacter baumannii ATCC 19606 was used as positive control. PCR program followed by: Initial denaturation 94°C for 5 min, denatuation 94°C for 45 seconds, annealing 58°C for 1 min, extension 72°C for 1 min, this program repeated for 30 cycles and the final extention 72°C for 5 min.

Table 1

Primer sequences and amplicon sizes

Antibiotic susceptibility test

The antibiotic susceptibility test was carried out according to CLSI 2011 recommendation and MAST company antibiotic discs. Strains by ≤13 mm zone size of inhibition were considered as imipenem-resistant. MIC was performed by macro dilution between the range of 0.5–128 µg/ml according to CLSI recommendation. MIC ≥16 µg/ml considered as an imipenem-resistant strain.

Phenotypic detection of metallo beta-lactamase

At first 0.5 M of EDTA reached by dissolving 186.1 grams in one liter distilled water and PH adjusted [8] by NaOH, then EDTA 750 µg/disc and 930 µg/disc were prepared. The inhibition zone of each disc was measured solitary. In the next step the DDST conducted by imipenem and EDTA distinctly for each disc with 750 and 930 µg/disc, which were placed on both sides of imipenem with a distance of 20 mm center to center, for eventual synergism effectiveness. Strains with increasing size in the imipenem inhibition zone towards EDTA are considered as MBL producers in DDST. In the CD assay usage of imipenem alone and imipenem plus two concentrations of EDTA, the strains with ≥7mm differences of zone inhibition between imipenem alone and imipenem plus EDTA in two concentrations are considered as MBL producer.

PCR molecular test for detection of MBL (blaVIM-1, blaVIM-2 and blaIMP-1)

Strains with at least one positive phenotypic test that explained above were examined for blaVIM-1, blaVIM-2 and blaIMP-1 genes by PCR [9]. In this study also detection of OXA-23 gene which is known as one of the common carbapenemase in Acinetobacter spp., was done by PCR for survey of false positive responds in CD assay. The Primers and PCR program were used in this study are shown in Table 1 (Tab. 1) and Table 2 (Tab. 2).

Table 2

The PCR program for blaVIM-1, blaVIM-2, blaIMP-1 and OXA-23

Polymerase chain reaction was performed for each sample with the following compounds. 1X concentration Specific PCR buffer, 0.4 mM of dNTPs mix, 0.7 mM of MgCl2, 1.6 M of each primer, one unit of Taq polymerase enzyme, 2 µl of DNA and sterile distilled water to get 25 µl as a final volume. Finally, the PCR products were evaluated on a 1% agarose gel.

Results

Ninety-three strains were identified as Acinetobacter baumannii by specific biochemical test and confirmed by PCR. According to CLSI 2011 guideline, 80 (85%) strains were resistant to imipenem. MIC verified these results (MIC 16 µg/ml: 20%, MIC 32 µg/ml: 26%, MIC 64 µg/ml: 46% and MIC 128 µg/ml: 8%). 31 (34%) of imipenem-resistant strains were positive in the CD test with 750 and 930 µg/disc concentration of EDTA. Simultaneously, 750 and 930 µg/disc EDTA alone made the inhibition zone up to 13 mm and 20 mm respectively (Figure 1 (Fig. 1)). None of them had synergistic effects between these two charges of EDTA alone and imipenem alone. In the molecular test there was not detected any VIM1, VIM2 and IMP1 genes in expected size after gel electrophoresis. Conversely, OXA-23 gene was observed in 25 out of 31 strains in positive CD test (Figure 2 (Fig. 2)).

Figure 1

The inhibition zone of EDTA disc

A. 930 µg/disc, B. 750 µg/disc

Figure 2

(From left to right) 1-Ladder 2-7 positive OXA-23 and 8-10 negative OXA-23

Discussion

The ability of producing metallo-beta-lactamase enzymes in gram negative bacteria is one of the resistance mechanisms, consequently the several methods have been suggested for phenotypic identification of metallo-beta-lactamase enzymes [2], [3], [7]. CD and Double Disc Synergism DDST methods with EDTA and imipenem are known as the most ordinary methods for phenotypic recognition of metallo-beta-lactamase [8], [9]. In the Eser and colleagues’ study which was done on 124 strains of Acinetobacter baumannii in Turkey in 2009, 80% imipenem-resistant strains were detected as CD positive, however the results of PCR molecular tests for detection of blaVIM and blaIMP1 were negative [11]. Results of recent studies are exactly similar to ours while the blaVIM1, blaVIM2 and blaIMP1 genes in CD positive strains didn’t distinguish, therefore our data shows that this technique is not specified for the detection of MBL. It is notable that the specifity of CD test was reported 98% in another study on Enterobacteriaceae family in Greece in 2008 [7], despite that the EDTA concentration was 2.5 times higher than ours. On the other hand in 2012 a Malaysian study was conducted on imipenem-resistant Pseudomonas aeruginosa. The results showed that the specificity of the CD test in 43% of the strains was confirmed by molecular method while the specificity of DDST 97% and E-Test 62% reported respectively [8]. In 2008, a Brazilian study on gram-negative bacilli to identify the MBL enzyme using different inhibitors was performed. Test results showed that the CD test could not classify Acinetobacter species [10]. In the present study, no synergistic effect was observed in DDST techniques and molecular tests for the presence of genes VIM1, VIM2 and IMP1 was negative. According to the study, it was shown that EDTA alone was tested for Acinetobacter strains has an inhibition zone from 13 mm for 750 µg/disc and up to 20 mm for 930 µg/disc). Therefore, an increase of 7 mm zone of inhibition between imipenem alone and imipenem with EDTA can not explain the synergistic effect of these two materials. In 2011 in Iran the DDS test for the detection of MBL in Pseudomonas spp. showed that phenotypic tests for MBL detection was confirmed by molecular testing [3]. In 2002 in Italy 28 Acinetobacter imipenem-resistant strains were studied that 16 strains with DDST were positive for MBL production and the results confirmed by PCR. Maybe using CD test by other MBL inhibitors can partially compensate the problems caused by the use of Imipenem and EDTA simultaneously. MBL mediated imipenem resistance in Acinetobacter baumannii is a cause for concern in the treatment of infected burns patients. The rate of imipenem resistance due to MBLs was increased dramatically [8], [17]. The clinical significance of OXA-23 isolates, which were widespread [18], [19] in burn unit, is of great importance, since clinicians are advised against the use of extended-spectrum cephalosporins, aztreonam, cephamycins, imipenem, and aminoglycosides. In 2010 in China, an outbeak of OXA-23 producer A. baumannii was reported [18]. In 2013 in Nigeria 60% of imipenem-resistant isolated A. baumannii positve for OXA-23 [20].

This observation emphasizes the importance of having effective control measures in Iranian burn hospitals, such as early detection of colonized patients, isolation procedures, and a judicious use of antibiotics.

Conclusions

According to the results of this study and other studies, it seems that the DDS test for MBL detection is more specific than the CD test. On the other hand, using different charges of EDTA and zone inhibition in CD test is another reason that proves this phenotypic method is unreliable in the detection of MBL producing strains. It appears that the different species of bacteria and geographical area have important role in determining the specificity of the CD, also the variety of used concentrations and variables methods which have been applied in any of the studies mentioned above, confirm the instability of conditions and impossibility of standardization of the CD by EDTA tests. In addition, because the 80% of CD test responses in imipenem-resistant strains were fake, the OXA-23 gene was detected instead of metallo-beta-lactamase.

Notes

Competing interests

The authors declare that they have no competing interests.

Funding

This study was supported by a grant (M/T 91-01-13-16573) from Tehran University of Medical Sciences, Tehran, Iran.