Proteus mirabilis is an opportunistic pathogen that can cause meningitis, septicemia and urinary tract infections. P. mirabilis has now become one of the health care associated infection pathogens in China. 15 carbapenem-resistant P. mirabilis isolates were investigated in a Chinese hospital during a 3 years period from 2012 to 2015. Antimicrobial susceptibility testing, Hodge test and plasmid transformation were performed. 15 isolates were all resistant to imipenem, meropenem and ertapenem. Hodge testing was positive for 7/15 strains. The blaKPC-2 gene was detected in 11 carbapenem resistant P. mirabilis isolates and mainly located in the plasmid. In addition, carbapenem-resistant P. mirabilis had a relatively low rate of amikacin resistance. Carbapenem resistance P. mirabilis isolates was mainly mediated by blaKPC-2 gene in our hospital. It is essential to prevent the spread of such factors conferring high level carbapenem resistance.

Carbapenemase resistance; KPC-2; Proteus mirabilis

Proteus mirabilis, the most common and ubiquitous in nature Proteus species, is a saprophytic bacterium under certain conditions. P. mirabilis can cause primary and secondary infections in the digestive, respiratory, urinary and respiratory tracts and in wounds and burns, as well as causing infections in other body sites. P. mirabilis is implicated in meningitis, peritonitis, rheumatoid arthritis, food poisoning and sepsis, and when sepsis is induced by it the fatality rate is high [1-3]. In recent years, with increasing use of broad spectrum and extended-spectrum β-lactam antibiotics in clinical settings, the incidence of drug resistance in Proteus spp. has increased [4,5], resulting in great difficulties in treating the diseases caused it. Carbapenems are antibiotics with good antibacterial activities against P. mirabilis and other Enterobacteriaceae. However, in recent years, increasing numbers of studies on carbapenems resistance strains have been published [4,5]. Research shows that the resistance mechanisms of Gram-negative bacteria to carbapenems is governed by four aspects: a decline in the affinity of periplasmic binding proteins to carbapenems, the active pump system, and excessive expression of the AmpC enzyme (AmpC β-lactamase) combined with outer membrane protein loss and carbapenemase production [6-8]. More and more carbapenem-resistance genes (e.g., blaKPC, blaNDM-1, blaGES, blaSME, blaIMI-1/NmcA, blaSHV-38 and blaVIM) have been detected in clinical Enterobacterial isolates since the Klebsiella pneumoniae Carbapenemase (KPC) enzyme was first reported in 2001 [9-13]. In 2009, Clinical and Laboratory Standards (CLSI) stated that bacterial carbapenemases may cause drug resistance to third generation cephalosporins in Enterobacteriaceae and reduce the sensitivity of carbapenem antibiotics. The CLSI suggested that the modified Hodge test should be used to identify positive strains with supplementary PCR amplification to confirm whether the isolate produces the KPC enzyme and the type of blaKPC enzyme it produces [14].

For infections caused by P. mirabilis to be treated effectively, it is important that a suitable detection method is available and that carbapenemase enzyme production can also be determined in a timely manner. Better understanding of the evolution of drug resistance in P. mirabilis will ensure that the best use of antibiotics is made and will help to reduce the evolutionary selection of antibiotic resistant isolates.

P. mirabilis isolates with carbapenem resistance (Table 1) were recovered from the First People’s Hospital of Ningbo, Zhejiang Province, China from January 2012 to December 2015. They included two isolates from 2012, two isolates from 2013, 3 isolates from 2014, and 8 isolates from 2015. Bacteria identification was performed with a Vitek System (bioMe ´rieux, Hazelwood, MO, USA). The control strains, Escherichia coli ATCC 25922 and K. pneumonia ATCC 700603, were purchased from the Shanghai Fu-xiang Biotechnology Co., Ltd (Shanghai, China). E. coli J53 was purchased from Takara (Dalian Inc., China).

The Minimum Inhibitory Concentrations (MICs) of Imipenem (IPM), Meropenem (MEM), Ertapenem (ETP), Ciprofloxacin (CIP), Amikacin (AK) and Minocycline (MIN) were determined by the agar dilution method according to CLSI recommendations [14]. E. coli ATCC 25922 and K. pneumonia ATCC 700603 were used as the quality controls.

Testing was performed according to the CLSI standards [14] as follows.

1. A 0.5 McFarland standard suspension was prepared (using the direct colony suspension method) for E. coli ATCC 25922 (the indicator organism) in saline, and then diluted 1:10 in saline. A MHA plate was inoculated according to the routine disk diffusion procedure. The plate was allowed to dry for 3-10 min. The appropriate number of ETP disks was used on the plate.
2. A 10-μL loop was used to pick 3-5 colonies of the test organism (previously grown overnight on a blood agar plate) and they were inoculated in a straight line out from the edge of the disk. The streak was 20-25 mm in length.
3. Bacterial growth was observed after incubation at 36°C for 16-20 h, and the samples were considered carbapenemase-positive when growth appeared in the ETP bacteriostatic circle.

Genomic DNA from each P. mirabilis isolate, which was obtained using an Axyprep bacterial genomic DNA miniprep kit (Axygen Scientific, Union City, CA, USA), was used as the template for the PCRs. The primers used to amplify blaKPC, blaVIM, blaNDM-1, blaGES, blaSME, blaIMI-1/NmcA and blaSHV-38 have been described previously and are shown in table 2. PCR products were sequenced using an ABI 3730 sequencer (Applied Biosystems, Foster City, CA, USA), and the sequences were compared with the reported sequences (the coincidence rate was 99% with the register number EU 784136) from GenBank (www.ncbi.nlm.nih.gov/BLASTn).

Using Qiagen plasmid MIDI kits (Qiagen, Japan), the blaKPC gene containing plasmids from P. mirabilis were extracted. The concentration and purity of the extraction plasmids were measured using a nucleic acid determination instrument (Bio-Rad Laboratories, Inc. Hercules, USA). The DNA concentration was adjusted from 105 to 114 µg/mL, and the optical density (OD260/OD280) values were 1.74 to 1.83. A 5 µl aliquot was added to 50 µl of NaN₃-resistant E. coli J53, after mixing with an electric rotary shaker instrument (Sanyo, Japan); the electro oration voltage was 1500 V and the shock time was 5.6 s. After electrical transformation, 1.0 ml of the SOC culture liquid was added to the electrical slot; the bacterial solution was transferred to the test tube, and then incubated at 37°C for 1 h. Next, 100 µl of the bacterial solution was spread on an LB agar plate (OXOID, UK) with 200 µg/mL of NaN₃, 100 µg/mL of ampicillin, and 2.0 µg/mL of Meropenem (MEM) for screening. In addition, E. coli J53 was inoculated overnight at 37°C on the LB agar plates with and without the above three substances. We compared the MIC changes for IPM, MEM and ETP with the treated and untreated E. coli J53 through the electrical transformation results, after which the blaKPC gene was detected in the transformants using PCR.

We screened 15 P. mirabilis isolates from 2012 to 2015 for carbapenem resistance. These isolates were from 15 patients in seven different departments, and they included seven from an Intensive Care Unit (ICU), four from oncology and the elderly, and one each from renal medicine, dermatology, neurology and hematology departments. In addition, the isolates are from four different specimens, eight strains come from urine, three strains each from blood and sputum, and one strain from sanies. The distribution of the disease profile and age of the patients for these isolates is shown in table 1.

These P. mirabilis isolates are tested antimicrobial susceptibility and display resistance to carbapenem. The MICs for IPM and MEM both ranged from 4 µg/mL to greater than or equal to 128 µg/mL. The MICs for ETP ranged from 4 µg/mL to greater than or equal to 128 µg/mL. MIC₅₀ of these isolates to IPM, MEM and ETP are 64 µg/mL, 16 µg/mL and 128 µg/mL, respectively. MIC90 of these isolates to IPM, MEM and ETP are 128 µg/mL. Of the 15 P. mirabilis isolates, most isolates were resistant to CIP (13 [86.7%] isolates) and MIN (11 [73.3%] isolates), while most were susceptible to AK (11 [73.3%] isolates). The details are shown in table 3.

In this study, seven carbapenem resistance genes including blaKPC, blaVIM, blaNDM, blaGES, blaSME, blaIMI-1/NmcA and blaSHV-38 were identified in a fifth of the carbapenem resistant P. mirabilis strains using PCR. Eleven isolates carried the plasmid encoded blaKPC gene, while the others were PCR-negative for the carbapenem resistance genes (Table 3). Moreover, 11 P. mirabilis strains of carbapenem resistance were obtained successfully by plasmid transformation of E. coli J53. The blaKPC-2 gene in the E. coli J53 transformants were detected by PCR amplification. These results showed that blaKPC-2 gene in the 11 P. mirabilis strains were located in the plasmid.

At present, carbapenemase resistance Enterobacteria were mostly reported in K. pneumoniae, K. oxytoca, E. coli, Enterobacter, Serratia marcescens and Salmonella [1,10,12]. However, there have been few studies on carbapenem resistance in P. mirabilis strains. In fact, infections caused by carbapenem resistance P. mirabilis strains present a serious challenge due to their intrinsic resistance to colistin and tigecycline [2,3].

Our study showed 15 clinical P. mirabilis strains carrying blaKPC-2 carbapenemase resistance gene emerged in a Chinese hospital. Compared with other previous studies on P. mirabilis strains carrying blaKPC-2 gene [15], our strains displayed a higher MIC values of carbapenem resistance. MIC₅₀ of these isolates to IPM, MEM and ETP are 64 µg/mL, 16 µg/mL and 128 µg/mL, respectively. Noted, several strains displayed greater than or equal to 128 µg/mL to IPM, MEM and ETP. This may be related to carbapenem antibiotics that were widely used in clinics in recent years.

The modified Hodge test was performed to clear mechanism of carbapenem resistance, 7 P. mirabilis isolates were able to produce carbapenemases. Moreover, we also detected carbapenem resistance gene and confirmed that 11 of the 15 carbapenem resistant P. mirabilis isolates carried blaKPC-2 gene. Nevertheless, four isolates were negative for the carbapenem resistance gene. This suggests that additional mechanisms may be involved in carbapenem resistance in these strains. Among the possibilities for such a mechanism are porin alterations, which reduce the entry of carbapenems into the cell as has been shown mainly for imipenem resistant K. pneumoniae strains [16], or efflux pumps mechanisms, which are more common in nonfermenters such as Pseudomonas aeruginosa and Acinetobacter [17]. To further definite blaKPC-2 gene location, plasmid transformation experiment was performed. The plasmid carrying transformants from the 11 P. mirabilis strains successfully transferred the blaKPC-2 gene to the antibiotic susceptible E.coli J53 strains. In addition, AK is a good choice to carbapenem resistance P. mirabilis strains according to antibiotic susceptibility tests.

In conclusion, carbapenem resistance P. mirabilis isolates was mainly mediated by blaKPC-2 gene in our hospital. It is essential to prevent the spread of such factors conferring high-level carbapenem resistance. Rapid and accurate detection of bacteria carrying multidrug-resistance genes will provide the basis for monitoring such resistance and, hopefully, improve clinical treatment.

This work is supported by the Zhejiang Medicine and Health Scientific Research Fund, Project (No. 2014KYA196).