Klebsiella pneumoniae blaNDM-1 carrying a class 1 integron causing a hospital outbreak in a Mexican attention center

Introduction: Infections acquired in hospitals are the cause of high morbidity and mortality and with the emergence of resistant bacteria, the problem is greater. The aim of this work was to determine the genetic characteristics and timeline of Klebsiella pneumoniae blaNDM-1 carrying a class 1 integron involved in an intrahospital outbreak. Methodology: Investigation was made from the first detection of K. pneumoniae blaNDM-1, strain “466”, and the last clone “423”. 16S rRNA gene analysis showed that 466 strain and clones were related to K. pneumoniae. Extended-spectrum β-lactamases (ESBL) was detected according to the Clinical and Laboratory Standards Institute (CLSI) and real time-PCR. Typing of K. pneumoniae blaNDM-1 strains was carried by ERIC-PCR and sequencing the variable region of the integrons were performed. Results: A cluster of six resistant isolates of K. pneumoniae blaNDM-1 was detected in intensive care unit (ICU), internal medicine (IM) and orthopedics (OT). Timeline revealed that the first bacterial identification was in ICU and the last clone in OT service. The array genetic of variable region was “IntI/aadA5-drfA17/qacEΔ1-Sul1”. Conclusions: The evidences highlight the importance of the epidemiological surveillance of Extended-spectrum β-lactamases (ESBL) strains, as well as the need for molecular epidemiological studies to identify the routes of transmission and the contamination sources within health personnel.


Introduction
Bacterial infections acquired in hospitals are the cause of high morbidity and mortality rates worldwide and with the emergence of antibiotic-resistant bacteria, the problem is greater, and it represents a challenge for health personnel and immunosuppressed patients [1][2][3]. It has been reported that inanimate surfaces are the main sources of contamination [4,5]. Moreover, cross contamination between patients and medical personnel, as well as patients releasing bacteria to the environment are determinant in the transmission of pathogens [6,7]. Reports indicate that 20 to 40% of pathogens causing hospital infections come from cross contamination by health personnel [8,9]. Gram-negative bacilli, especially species of the genus Klebsiella, are among the main bacteria that cause outbreaks in the intensive care units (ICU) [10][11][12][13]. With the recent emergence of multi-resistant strains due to extended-spectrum βlactamases (ESBL), the persistence of pathogens in the hospital environment plays an important role in morbidity and morbility [14].
Carbapenemases are increasingly detected in Klebsiella pneumoniae and other members of Enterobacteriaceae family, such as class A carbapenemase (KPC)-type, class B metallo-βlactamases (MBLs mainly bla IMP , bla VIM and bla NDM -type enzymes) and class D bla OXA-48 -like enzymes [15][16][17]. Another type of important genetic elements that are associated with antibiotic resistance are class 1 integrons. These structures are versatile for the capture of resistance cassettes, therefore, their identification and characterization in outbreaks associated with antibiotic resistant bacteria is of great relevance and could be a potential supplementary tool in the characterization of nosocomial outbreaks [18][19][20]. Additionally, the molecular typing has been widely used in the study of clonal dissemination of pathogens in the hospital environment to investigate nosocomial outbreaks. In the present study, we present the characterization of an outbreak in the "Hospital Juárez de México", caused by K. pneumoniae bla NDM-1 strain 466 that colonized six patients. This strain was able to produce a carbapenemase type bla NDM-1 and carry a class 1 integron with the genetic arrangement (IntI/aadA5-drfA17/qacEΔ1-Sul1) located in genome and plasmid respectively. Implications of the propagation of multiresistant strains in hospitalized patients and their attempt to eradicate them are discussed.

Outbreak investigation
This work describes the methodology in the characterization of an outbreak due to K. pneumoniae blaNDM-1 and carrier of a class 1 integron in the "Hospital Juárez de México". The investigation of the outbreak included the first detection of K. pneumoniae bla NDM-1 , strain "466" (June 21 st , 2016; beginning of the outbreak) and the identification of the last clone "423" (September 14th in the same year). On December 31st, the outbreak was declared as finished. Phenotypic and molecular assays were performed on all positive isolates of K. pneumoniae together with prospective epidemiological surveillance and data collection. Additionally, information (file number, diagnosis, antimicrobial treatment, gender/age of the patient, hospital service and source of isolation) was analyzed.
Bacterial cultures were obtained and characterized from routine diagnosis and demographic analyses were performed retrospectively, so neither Institutional Review Board (IRB) approval nor informed consent from patients was required.

Bacterial identification
Thirty-six substrates were used for the identification of strains at the genus and species level through a phenotypic test by using the system Phoenix-100 (Becton Dickinson, USA), according to the manufacturer protocol. In order to confirm, sequence analysis of the 16S rRNA gene was performed as follow. 16S rRNA gene was amplified by using universal primers (27F and R1492) according to DeSantis et al. [21] (Table 1). Reactions were performed in a SEEAMP™ PCR system (Seegene, Seoul, Korea). The nearly complete 16S rRNA gene sequence was determined as described by Kwon et al. [22]. Sequencing was carried out by the Biology Institute, at the "Universidad Nacional Autónoma de México" by using an ABI PRISM ® 310 Genetic Analyzer sequencer (Applied Biosystems, CA, USA). Sequences were compared with the nucleotide sequence database (GenBank) by means of the Blast algorithm [23]. Phylogenetic analysis was carried out by maximumlikelihood after multiple alignments of data by Thompson et al. [24].

Antimicrobial resistance assay
Antimicrobial resistance was determined by Clinical and Laboratory Standards Institut (CLSI) [25].

Carbapenemase-coding genes detection by real time-PCR
Detection of bla KPC , bla OXA-48 and bla NDM-1 was performed by real time-PCR, using the BD MAX TM CRE ASSAY set in the system BD TM MAX (Becton Dickinson, NJ, USA). Additionally, a macroarray assay based in reverse hibridization (Sepsis Flow Chip TM Kit) was performed in order to confirm the presence of bla NDM-1 marker (Master Diagnóstica, Granada, Spain). Location of carbapenemase-coding gene was done by plasmid DNA isolation and used as template in PCR assays.

Molecular typing of Klebsiella pneumoniae bla NDM-1 466 strain and clones
Klebsiella pneumoniae bla NDM-1 466 and the five derived clones were analyzed by ERIC-PCR.

Conjugal transfer experiments
Donors K. pneumoniae bla NDM-1 and recipient E. coli J53-1 strains were adjusted to 10 8 CFU/mL, mixed (1:1) and spotted on the surface of 0.22 μm nitrocellulose filters, positioned on LB agar and incubated overnight at 37 °C. Bacteria were suspended in 100 μL and were diluted serially and spread on MacConkey agar plus 10 μg/mL nalidixic acid and 25 μg/mL streptomycin or 1 μg/mL meropenem for transconjugants and nalidixic acid for recipients. Conjugal transfer frequencies were calculated by dividing the CFU transconjugants by the CFU recipients.

Outbreak investigation
Six patients infected with K. pneumoniae bla NMD-1 were identified in a period of 4 months. This is equivalent to 0.137% of a total of 4,355 cultures of all Gram-negative bacteria during 2016. According to the timeline in the occurrence of cases associated with K. pneumoniae bla NDM-1 , the origin of the outbreak was associated with strain 466 (June, 2016) in the intensive care unit (ICU). K. pneumoniae bla NDM-1 (466 strain and 391, 332, 381, 208 and 423 clones) were detected in three hospital services at HJM: ICU (n = 4), internal medicine (IM) (n = 1) and orthopedics (OT) (n = 1). From this first finding, epidemiological and microbiological surveillance was performed in hospital services of the HJM. Only one patient of the female gender in the ICU survived. A summary of microbiological detection, patient characteristics and isolation origin of 466 strain and its clones are shown in Table 2. Finally, an outbreak investigation timeline was carried out from the first identification in ICU (Figure 1).

Phenotypic and genetic bacterial identification
Biochemical reactions revealed that 23 substrates out of 43 were assimilated by isolates with a 99% confidence assignment. The analysis by system assigned the genus and species to strain 466 and clones 391, 332,381, 208 and 423 as K. pneumoniae. Phylogenetic analysis and maximum-likelihood indicated that 466 strain and clones 391, 332, 381, 208 and 423 were closely related to K. pneumoniae subsp. pneumoniae strain Z-24 (KP135420.1) with 99.99% of similarity, correlating with the phenotypic assay.

Antimicrobial susceptibility assay
The results of antimicrobial susceptibility assay, revealed that 466 strain and clones 391, 332, 381, 208 and 423 showed resistance profile in both methods probed. All isolates were resistant to CTX, CAZ, FEP, FOX, CTT, CRO, ETP, MEM, IPM, and ATM. No inhibition halo was detected for CTX-CLA and CAZ-CLA.

Plasmid DNA isolation
Plasmid extraction assays in the isolates revealed in a plasmid of ≈ 15 Kb in 466 strain and clones 391, 332, 381, 208 and 423.

Typing of K. pneumoniae bla NDM-1 466 strain and clones
Profiles of the intergenic products revealed sizes of amplicons ranged from slightly more than ≈300 bp to about ≈1700 bp (Figure 3). Intergenic region diversity not allowed the differentiation of 6 isolates, therefore were clustered in one genetic group. According to above, the six strains were indistinguishable from each other (100% similarity).

Conjugal transfer experiments
Conjugal transfer frequencies were 2.27×10 -6 and identical plasmid profile was detected in all tranconjugants compared with strains donor. Also, PCR assays were performed in order to detect all elements of class 1 integron. According to above, class 1 integron was located in plasmid and it is transferable to recipient bacteria.

Discussion
The ESBL-producing Enterobacteriaceae causing hospital outbreaks continues to be a major public health problem mainly in developing countries, such as Mexico. Here, we describe the K. pneumoniae bla NDM-1 isolation in six patients from June to the end of September 2016. The presence of the molecular markers of carbapenemases has been previously identified in several pathogens belonging to the family Enterobacteriaceae, being K. pneumoniae one of the most important in the hospital environment [12,13,26,27].
According to the Mexican national system of epidemiological surveillance, the notification of the detection of carbapenemase-producing pathogens is mandatory along with the implementation of prospective epidemiological surveillance. In particular, the identification of the molecular marker "bla NMD-1 " is relevant because the β-lactamase encoded by this gene is capable of inactivating all the antibiotics of the carbapenemic family, except aztreonam [28]. In this sense, the hospital outbreaks related to K. pneumoniae blan NDM-1 are alarming, due to the fact that all broadspectrum antibiotics are not effective in antimicrobial therapy.
Molecular tools play an important role in the identification and dissemination of pathogens in the hospital environment. The analysis of the sequences of the 16S rRNA gene allowed classifying the pathogen detected in K. pneumoniae, with an identity of 99.99% with sequences described previously. It is interesting to know if the isolates reported in this work are associated with similar clones detected in other hospital outbreaks. In Mexico, K. pneumoniae harboring bla NDM-1 was firstly identified in an hospitalized pediatric patient [29] and more recently, it has been reported as a causative agent of intrahospital outbreaks [30][31][32][33], suggesting an emerging role of this multiresistant strain in nosocomial environments. For this reason, futures studies are needed to determine the possible genetic association of strain 466 with other K. pneumoniae bla NDM-1 identified in Mexico and other parts of the world by sequencing additional genes such as recA and gyrB and even the complete genome. Clonal relation assays reported in the present work are important to identify the relationship between isolates with similar phenotypic characteristics. The results of the clonal relationship provide evidence to strengthen the hospital epidemiological surveillance system in making decisions about infection control measures associated with health care. The evidence generated by ERIC-PCR showed a clonal dissemination of K. pneumoniae bla NDM-1 in six patients from three direct hospital departments. Even though the patients hospitalization periods did not overlap each other (according to the timeline), the analysis revealed that all the isolates were genetically related. Cross contamination by medical personnel or by environmental contamination is suggested. It has been shown that the ICU is one of the hospital areas with the highest incidence in the detection of hospital outbreaks [10,11]. Sources of contamination derived from cross-contamination by staff, and the use of equipment that is not subject to a high-level cleaning process, are the main sources of pathogens [4,5]. Plasmids encoding antimicrobial resistance are diverse in K. pneumoniae and can also carry other resistance genes, including carbapenemase genes [34]. Conjugational transfer experiments were performed to demonstrate that the bla NDM-1 marker was located in plasmid. Our results showed that the resistance marker bla NMD-1 was not able to transfer because this marker was located in the bacterial chromosome, correlating with previous studies where chromosomal bla NDM-1 has been observed, such as existence of integration events of carbapenemic markers, mediated by presence of transposons [35]. Other elements associated with antibiotic resistance in Gram-negative bacteria are the class 1 integrons [36]. These structures, having the capacity to capture resistance genes, assumed that the bla NDM-1 gene could possibly be found in an integron. Under this hypothesis, search of class 1 integrons in K. pneumoniae bla NDM-1 was done.
PCR assays and sequence analysis revealed an identical genetic arrangement in the six isolates (intI1/aadA5-dfrA17/qacEΔ1-sulI), located in plasmid. The genetic arrangements constituted by aadA and dfrA cassette variants have been previously reported in other members of the Enterobacteriaceae family that cause hospital outbreaks and that carry class 1 integrons [37]. One of the main causes of the spread of pathogens with resistance markers is the border transfer of infected patients [38]. Under this argument, the clinical record of the patient from where strain 466 was isolated was analyzed and did not contain information on the history of travel abroad. The control measures implemented after the detection of K. pneumoniae bla NDM-1 , included the exhaustive disinfection of the areas of patients infected with the subsequent search for the carbapenemase producing pathogen, and extreme isolation precautions of the patients to avoid possible cross-contamination.
In conclusion, the temporality of the six cases, the genetic relationship determined by molecular typing, as well as the phenotypic resistance profile and the genetic elements associated with resistance shared between the isolates allowed to define K. pneumoniae as the causing of a nosocomial outbreak that included fatal cases. This strain was able to produce a carbapenemase type bla NDM-1 and carry a class 1 integron with the arrangement (IntI/aadA5-drfA17/qacEΔ1-Sul1) located in genome and plasmid respectively.

Conclusions
The evidences presented in this work, highlight the importance of the constant epidemiological surveillance of strains producing EBLS at local level, as well as the need for epidemiological studies of national coverage to identify the possible routes of transmission and the possible sources of contamination within health personnel.