High plasmid variability, and the presence of IncFIB, IncQ, IncA/C, IncHI1B, and IncL/M in clinical isolates of Klebsiella pneumoniae with bla KPC and bla NDM from patients at a public hospital in Brazil.

Abstract INTRODUCTION Antibiotic resistance in carbapenemase-producing Klebsiella pneumoniae is acquired and disseminated mainly by plasmids. Therefore, we aimed to investigate the occurrence of carbapenemase genes, analyze the genetic diversity by ERIC-PCR, and examine the most common plasmid incompatibility groups (Incs) in clinical isolates of K. pneumoniae from colonization and infection in patients from a hospital in Brazil. METHODS Twenty-seven isolates of carbapenem-resistant K. pneumoniae were selected and screened for the presence of carbapenemase genes and Incs by PCR, followed by amplicon sequencing. RESULTS The bla KPC and bla NDM genes were detected in 24 (88.8 %) and 16 (59.2 %) of the isolates, respectively. Thirteen isolates (48.1 %) were positive for both genes. The IncFIB (92.6 %) and IncQ (88.8 %) were the most frequent plasmids, followed by IncA/C, IncHI1B, and IncL/M, indicating that plasmid variability existed in these isolates. To our knowledge, this is the first report of IncHI1B in Brazil. We found eight isolates with clonal relationship distributed in different sectors of the hospital. CONCLUSIONS The accumulation of resistance determinants, the variability of plasmid Incs, and the clonal dissemination detected in K. pneumoniae isolates demonstrate their potential for infection, colonization, and the dissemination of different resistance genes and plasmids.


INTRODUCTION
Klebsiella pneumoniae is clinically important because it is involved in a variety of healthcare-association infections (HAI) such as in the urinary and respiratory tracts, wounds, endocarditis, and sepsis 1,2,3 . Intestinal colonization is one of the main factors favoring infection by K. pneumoniae, as colonized carriers can serve as important reservoirs for the spread of bacteria in the hospital environment. Further, K. pneumoniae can harbor resistance genes and spread them through conjugative plasmids and transposons 4 .

Bacterial isolates
Twenty-seven isolates of K. pneumoniae selected for being resistant to one or more carbapenems were isolated from different patients and sites of infection or colonization. The patients were admitted to a public hospital in the city of Recife-PE, Brazil, between 2017 and 2018. The isolates were kept as frozen stocks at -80ºC in 15 % glycerol.

DNA extraction and PCR conditions for resistance genes
Genomic DNA was extracted using a commercial kit as per the manufacturer's instructions (Wizard Genomic DNA Purification kit, Promega). After extraction, the DNA was quantified using the NanoDrop 2000c UV-Vis spectrophotometer. For PCR amplification of the bla KPC , bla NDM , bla GES , bla VIM , and bla IMP genes, the primers described in Table 1 were used. The amplification reactions were prepared in a total volume of 25 μL per tube, comprising a final concentration of 25 mM MgCl 2 , 8 mM dNTPs, 1U Taq DNA Polymerase (Promega), 10 μM of each primer, 5×buffer, and 1 ng of DNA.
The following thermal cycling conditions were used for bla KPC amplification: initial denaturation for 5 min at 95ºC, followed by 30 cycles of 1 min at 95ºC for denaturation, 1 min at 63ºC for primer annealing, and 1 min at 72ºC for the extension step. Subsequently, a final elongation step of 10 min was performed at 72ºC. For amplification of the bla GES gene, the following conditions were used: 3 min at 93°C, followed by 40 cycles of 1 min at 93°C, 1 min at 55°C and 1 min at 72°C, and a final extension for 7 min at 72°C. For the amplification of the bla NDM gene, the conditions used were 10 min at 94°C, followed by 36 cycles of 30 s at 94°C, 40 s at 52°C, 50 s at 72°C, and a final extension of 5 min at 72°C. For the bla VIM and bla IMP genes, we used PCR cycling of 5 min at 95ºC, followed by 30 cycles of 1 min at 95ºC, 1 min at 60 o C and 62 o C for the bla IMP and bla VIM genes, respectively, and an extension step for 1 min at 68ºC. Subsequently, a 5-minute final elongation step was performed at 68ºC.

Incs PCR
To detect the plasmid Incs, the primers for IncA/C, IncL/M, IncN, IncHI2, IncFIB, IncHI1B, and IncQ were selected and used 27,28,29 , as these plasmid incompatibility groups are more frequently described in the literature for K. pneumoniae. The cycling conditions used were 5 min at 94°C, followed by 35 cycles of 1 min denaturation at 94°C, 1 min of annealing with the specific temperature of each initiator used in the reaction ( Table 1), and 1 min of extension at 72°C. The final extension was performed for 10 min at a temperature of 72°C.

Electrophoresis and sequencing of resistance genes
The PCR products were analyzed via electrophoresis using a 1 % agarose gel in TBE buffer (0.089 M Tris-Borate and 0.002 M EDTA) at a constant voltage of 100 V. The gels were visualized under ultraviolet light using a transilluminator (Bio Rad) and photographed with a photo-documentation system (Photocap, Vilber Lourmat). The amplicons for each gene were purified using the SV Total DNA Isolation System (Promega) and the DNA was sequenced by the method of Sanger et al. (1997). The nucleotide sequences were analyzed using the BLAST program (http://www.ncbi.nlm.nih.gov/).

Enterobacterial Repetitive Intergenic Consensus Polymerase Chain Reaction (ERIC-PCR)
To assess the clonal relationship between the isolates, ERIC-PCR was performed as described by Duan Table 1). The DARWIN 6.0 software was used to generate the dendrogram.

Antimicrobial resistance profile
Although isolates resistant to at least one carbapenem were selected, we observed that most isolates were resistant to three carbapenems tested (i.e., ertapenem, imipenem and meropenem), except for isolate K5-A3 which was sensitive to imipenem and showed intermediate sensitivity to meropenem, and the K6-A3 isolate which had intermediate sensitivity to imipenem ( Table 2). The antimicrobials that showed the best activity against carbapenem-resistant K. pneumoniae isolates were amikacin and colistin, showing that 96.2% and 88.9% were sensitive to these antimicrobials, respectively.

Beta-lactamase genes
The bla KPC and bla NDM genes were detected in 24 (88.8 %) and 16 (59.2 %) isolates of K. pneumoniae, respectively ( Table 2), by amplifying the expected 800 bp and 621 bp genes for bla KPC and bla NDM respectively. The bla GES , bla VIM, and bla IMP genes were not detected. The bla KPC and bla NDM genes were detected in 13 (48.1 %) of the isolates analyzed. The bla KPC-2 and bla NDM-1 variants were confirmed by sequencing the PCR product from representative isolates.

Plasmid Incompatibility Groups (Incs)
The FIB (n=25; 92.6 %) and Q (n=24; 88.8 %) Incs were the most frequent in the K. pneumoniae isolates analyzed in this study, followed by the Incs A/C, HI1B, and L/M that were detected in 4 (14.8 %), 3 (11.1 %) and 2 (7.4 %) isolates, respectively. The Incs N and HI2 were not detected. Two isolates (K16-A3 and K20-A3) showed the presence of the IncFIB alone, and the isolate K17-A3 had only IncQ. The remaining 24 isolates had more than one Inc that were investigated in this study ( Table 2).

Molecular typing by ERIC-PCR
Of the 27 isolates of K. pneumoniae analyzed, 12 different genetic profiles were identified (Figures 1 and 2) by ERIC-PCR. Eleven isolates showed an 80 % similarity (E1a profile), of which eight isolates showed a 100 % similarity (E1 profile). Four isolates showed an identical resistance profile (presence of the bla KPC and bla NDM genes) and the presence of the same plasmid Incs (Q and FIB). The E2 profile grouped three isolates that were different for their resistance profile and the presence of Incs. The E4 profile grouped two isolates that showed an identical resistance profile (only the bla KPC gene was detected) and Incs (Q, FIB and HI1B). The E7 profile grouped three isolates, of which two isolates were identical for their resistance profile (bla KPC and bla NDM gene) and in the presence of Incs (Q and FIB), whereas one isolate in this profile had the bla KPC gene alone and the same plasmid Incs. The isolates that were not clonally related had a different resistance profile and the presence of distinct plasmid Incs.
Comparative analysis of the presence of genes for carbapenemases, plasmid Incs, and ERIC-PCR, based on colonization or infection as sources for isolation.
Of the 27 carbapenem-resistant K. pneumoniae isolates, 12 were from infected samples (blood or urine), 14 were from colonization samples (rectal swab), and 1 was from a cavity drain ( Table 2). The majority (58.3 %) of the isolates from the infected samples that had the bla NDM gene also harbored the bla KPC gene, except for the K16-A3 isolate that harbored the bla NDM gene alone. On analyzing the isolates from colonization samples, we observed that six isolates    were positive for the bla KPC and bla NDM genes. The Incs FIB, Q, A/C, and HI1B were detected in isolates from infection and colonization, and IncL/M was detected in only two isolates from colonization.

DISCUSSION
Most of the K. pneumoniae isolates evaluated in this study were resistant to all beta-lactams tested. These data justify the alert published by the CDC in 2013, which states that carbapenemaseproducing enterobacteria are a global threat due to the high rates of resistance to antimicrobials, and urgent and effective actions are required to control them 30 . Lorenzoni et al. (2017) 31 performed studies with carbapenem-resistant enterobacteria isolated from a hospital in the Rio Grande do Sul, Brazil, and detected high rates of sensitivity to colistin and amikacin in K. pneumoniae isolates. Further, the authors found that the bla KPC gene was detected in 80 % of the isolates, corroborating the data in this study, where we found an occurrence of 88.8 % for the bla KPC gene. These occurrence rates of the bla KPC gene identified in this study highlight the persistence of this gene in K. pneumoniae, and since its detection in 2006 from Recife, Brazil 13 , it remains the main carbapenemase associated with carbapenem-resistant K. pneumoniae samples in several Brazilian states, including Recife-PE 1,5,32 .
The occurrence of the bla NDM gene in 59.2 % of the isolates analyzed in this study is worrying and highlights the emergence of another carbapenemase that, in addition to KPC, can also hydrolyze carbapenems. Barberino et al. (2018) 33 were the first to report the presence of the bla NDM 17 detected the occurrence of bla NDM in 25 % of K. pneumoniae isolates, and also in isolates from Recife-PE. In our study, we detected a greater number of isolates with the bla NDM gene (n=16; 59.2 %), and these results indicated the rapid dissemination of this gene.
The rate of occurrence of the bla NDM and the bla KPC genes deserves to be highlighted because of the accumulation of these genetic mechanisms of resistance in the same bacterial species. In Brazil, the accumulation of resistance determinants in K. pneumoniae has been described by other authors. Nava et al. (2018) 7 detected the occurrence of the bla NDM , bla KPC , and bla TEM genes in clinical isolates of K. pneumoniae from a university hospital in Londrina-PR. The concomitant presence of the bla KPC and bla NDM genes in colonization isolates examined in this study is worrying and reinforces the need for surveillance cultures, as patients showing bacterial colonization are an important reservoir for the spread of resistance mechanisms within the hospital environment and are the main gateway to the development of infection.
We observed that K. pneumoniae isolates, despite being clonally related as indicated by an ERIC-PCR assay, presented different types of plasmids and different resistance genes. This may occur because the ERIC-PCR technique amplifies repetitive intergenic regions of the bacterial chromosome, yet does not necessarily amplify plasmid regions, where most of the resistance genes are located 36 .

6/8
Clonal dissemination of the isolates was observed in different sectors of the hospital under study and among colonized and infected patients. Therefore, our results indicate that K. pneumoniae can potentially spread in the hospital environment.
The persistence of genes that confer resistance to carbapenems results due to the clonal dissemination of the isolates and via the dispersion of these genes through conjugative or mobile plasmids 8,9,10,11 . In this study, we detected five types of Incs which are described in the literature as being potentially responsible for the spread of bla KPC and bla NDM genes in K. pneumoniae isolates. Additionally, all Incs identified in this study already harbor resistance genes, including bla KPC and bla NDM 4,16,19,20,37,38 , and the Incs FIB and Q were the most frequently detected in this study. This study is the first report of the IncFIB in bacterial isolates from Recife-PE, Brazil. IncFIB is a conjugative plasmid that has been associated with the dissemination of the bla IMP gene in E. cloacae in Japan 39 , and was also reported in Africa in E. coli isolates carrying the bla TEM gene 40 . In Europe, it was responsible for the spread of the bla NDM-1 , bla SHV-12 , bla CTXM-15 , and bla OXA-1 genes in K. pneumoniae 11 .
The second most frequently detected Inc in K. pneumoniae isolates was IncQ, a plasmid that harbors carbapenem resistance genes and had been gaining prominence in some regions of Brazil. Nicoletti et al. (2015) 41 identified the bla KPC gene inserted into an IncQ plasmid in K. pneumoniae. Cerdeira et al. (2019) 16 detected two isolates of K. pneumoniae that were resistant to carbapenems and had the bla KPC gene inserted into IncQ plasmids. According to Smalla et al. (2000) 27 , IncQ is a small plasmid that can vary between 5.1-14.2 kb in size, and it can be found in several host bacterial cells. IncQ is not conjugative, but is mobilizable and promiscuous, and can be transferred from one bacterium to another by conjugative plasmids, which are present in the same bacterial cell. Given the variability in plasmids, including for conjugative plasmids, which were detected in this study in K. pneumoniae isolates, IncQ can probably be disseminated to other species. Additionally, it plays an important role in enhancing the dissemination of the bla KPC gene in Brazil 16 .
Incs A/ C and L/M, despite being detected in a smaller number of isolates in this study, have been described in previous studies as carrying resistance genes in different species of enterobacteria in Brazil, including K. pneumoniae 20 . Pereira et al. (2015) 37 detected the simultaneous presence of the bla KPC and bla NDM genes in E. hormaechei in Rio de Janeiro, with bla NDM inserted into IncA/C. Only one study has investigated Incs in K. pneumoniae isolates from Recife-PE, Brazil; however, the plasmids were not typed, and the small number of isolates analyzed (only four) may be a limitation of the study 19 .
We also detected the presence of the conjugative plasmid IncHI1B in this study, and to our knowledge, this is the first report of this Inc in Brazil. This plasmid has been reported in clinical isolates of E. cloacae, K. pneumoniae, E. coli, and C. freundii as carrying the bla NDM 38 performed a conjugation and transformation experiment in isolates from surveillance programs and identified the bla VIM gene in plasmids IncL/M, IncN2, IncHI1B, and IncFIB in K. pneumoniae isolates from Greece and Spain. These studies show the ability of IncHI1B to host genes for carbapenemases in different species.
We conclude that the accumulation of resistance determinants, the variability of plasmid Incs, and the clonal dissemination of these in K. pneumoniae isolates from infection and colonization, demonstrate the ability of this species to acquire genes for resistance and disseminate them via conjugative and mobilizable plasmids. The importance of the early phenotypic and genetic identification of resistance mechanisms in bacterial isolates from infection and colonization samples needs to be highlighted to prevent and halt the development of infection in patients hospitalized due to immuno-depression. ACKNOWLEDGMENTS We are grateful to the Bioinformatics and Evolutionary Biology Group -LABBE UFPE, in particular Dr. Heidi Lacerda and Dr. Valdir Queiroz Balbino for their promptness in sequencing the amplicons. We thank Dra. Josineide Ferreira Barros for the help in obtaining the analyzed isolates from the study, as well as in their automated identification.

LIMITATIONS OF THE STUDY
It was not possible to identify all known plasmid Incs, as well as to sequence the plasmids to determine the specific location of each resistance gene.