Carbapenemases often jeopardize chemotherapy for infectious diseases caused by common nosocomial pathogens. Currently, Klebsiella pneumoniae carbapenemases (KPC) belonging to the Ambler class A carbapenemases are the most clinically worrisome enzymes because of their global spreads and a broad-spectrum of substrates including most of the β-lactam drugs, except cephamycins (Munoz-Price et al., 2013; Pitout et al., 2015). KPC producers were first identified for K. pneumoniae from the United States in 1996 (Yigit et al., 2001) and, so far, a total of 23 KPC variants have been identified (http://www.lahey.org/Studies). KPC producers, particularly among Enterobacteriaceae isolates, are now being extensively reported worldwide. In some regions of the world, such as North America (some regions of the USA), Latin America (Colombia, Brazil, and Argentina), Europe (Greece and Italy), the Middle East (Israel), and Asia (some regions of China), KPC dissemination has reached endemic levels (Albiger et al., 2015; Mehta et al., 2015).

The rapid dissemination of the bla_KPC gene is known to be derived from the clonal spread of K. pneumoniae sequence type 258 (ST258) (Munoz-Price et al., 2013; Pitout et al., 2015). Besides, horizontal gene transfer drives KPC dissemination via a Tn3-type active transposon Tn4401 enclosing the bla_KPC gene. The transposon is bracketed by 39-bp imperfect inverted repeat sequences capable of mobilizing the bla_KPC gene at high frequency and it generates a 5-bp target duplication by insertion without target site specificity (Cuzon et al., 2011). At present, nine isotypes of Tn4401 (Tn4401a to Tn4401i) differing by the upstream sequences of the bla_KPC gene were identified. The bla_KPC gene promoters varied by the Tn4401 isotypes control the level of gene expression resulting in varied levels of carbapenem resistance (Naas et al., 2012; Cheruvanky et al., 2017).

The National Laboratory Surveillance System (NLSS) for CPEs started in November 2010 by the Korea Center for Disease Prevention and Control (KCDC) with 100 sentinel general hospitals (unpublished data). Through the surveillance study, an alarming transition of KPC epidemiological stages to inter-regional spread was observed between 2014 and 2015. The population of KPC producers and the inter-regional movement of core clones needed to be further investigated and, in this study, the molecular epidemiology was assessed with actively collected KPC producers located throughout the peninsula.

KPC enzymes are able to hydrolyze penicillins, cephalosporins, monobactams, carbapenems, and even β-lactamase inhibitors. Therefore, the wide spread of the producers is one of the worrisome global issues encountering antimicrobial resistance. In accordance of occasional reports (Rhee et al., 2010; Hong et al., 2013; Yoo et al., 2013) and data from NLSS (unpublished data), only infrequent occurrence of the KPC producers were identified in South Korea until 2013. However, after a sudden expansion of KPC producers in 2014, rapid nation-wide spread supported by single-hospital outbreaks (Kim et al., 2017), frequently associated with inter-hospital or inter-regional spread (Jeong et al., 2016) led the endemic stage of KPC in 2015. Actually, the KPC producer in 2013 has never been identified again and the four K. pneumoniae subclones, ST307/Tn4401a[bla_KPC-2], ST307/Tn4401b[bla_KPC-4], ST392/Tn4401b[bla_KPC-4], and ST11/ND[bla_KPC-2] in 2014 triggered the inter-regional dissemination of KPC producers.

In agreement with the Healthcare Delivery System in South Korea, convalescent hospitals are in charge post-acute care and general hospitals are responsible for acute care, leading a habitual inter-hospital patient transfer. In general, the infection control system in convalescent hospitals is poorly operated compared to that in general hospitals and their role as a reservoir of CPE has been highly suspected. Indeed, the core KPC-producing subclone K. pneumoniae ST307/Tn4401a[bla_KPC-2] was identified in all three types of healthcare facilities: 153 of the 362 KPC producers from general hospitals, one of the three isolates from the two convalescent hospitals and one of the two isolates from clinics. It suggested that the KPC has disseminated during the inter-hospital patient transfer and expanded up to the community level. Further study with more isolates from the two types of healthcare facility is needed.

The well-known KPC-producing clone in a global level is K. pneumoniae ST258 (Pitout et al., 2015; Sheppard et al., 2016), however, it was not the case in South Korea. The foremost South Korean KPC producer was K. pneumoniae ST307 clone and the ST11 was followed by presenting half the prevalence of ST307. Indeed, the ST258 (gapA-infB-mdh-pgi-phoE-rpoB-tonB, 3-3-1-1-1-1-79) is a deduced variant of ST11 (3-3-1-1-1-1-43) manufactured by multiple events of a genetic rearrangement (Breurec et al., 2013). K. pneumoniae ST307, which has gained attention as a potentially high-risk KPC producer, has been a notorious clone producing CTX-M-15 (Villa et al., 2017) and the clone co-producing KPC and CTX-M-15 was recently identified in Italy (Geraci et al., 2015). The South Korean ST307 clone are assumed to be close to the Italian clone in an evidence of 71.9% KPC-producing K. pneumoniae ST307 co-produced CTX-M enzyme. The second-most disseminated K. pneumoniae ST11 is one of the main KPC-producing clones in China, and it frequently co-harbors the rmtB genes (Cheng et al., 2016). The 22 RmtB-co-producers were all K. pneumoniae ST11 representing 28.7% (22/77) the clone.

KPC-2 and KPC-3 are the prevailing global subtypes of KPC, with regional variation (Munoz-Price et al., 2013). Among the South Korean KPC producers, KPC-4 was the second most prevalent subtype, although the number of KPC-4 producers was only one fifteenth that of the KPC-2 producers. Since the bacterial host species varied, the Tn4401b-associated plasmid (along with sporadic occurrence) was likely responsible for its observed dissemination. KPC-3, which was produced by fewer bacterial hosts, was limited to one district. KPC-2, KPC-3, and KPC-4 each confer a similar level of carbapenem resistance (Mehta et al., 2015). However, among the KPC producers studied, KPC-2 had the highest MIC₅₀s for carbapenems, while KPC-3 and KPC-4 had 3- and 4-dilution-reduced MIC₅₀s, respectively (Figure 2). It is likely that multiple factors (such as small relative group size, or low gene expression in the Tn4401b isotype) accounted for this unexpected result.

Although group size varied (334 vs. 17), the KPC-producing K. pneumoniae isolates had 3-dilution higher MIC₅₀s for carbapenems compared to the KPC-producing E. coli isolates (Figure 2), pronouncing the importance of bacterial host background for revealing resistance. As well, the Tn4401 isotypes, differing the promoter of the bla_KPC gene, is closely associated with the level of gene expression resulting in varied carbapenem susceptibility (Naas et al., 2012; Cheruvanky et al., 2017). For example, Tn4401a and Tn4401d possessing two promoters upstream from the gene confer high carbapenem resistance whereas Tn4401c and Tn4401e, with one promoter, confer low levels of resistance and the Tn4401b, with three promoters, confers moderate resistance (Naas et al., 2012; Cheruvanky et al., 2017). Three isotypes of the transposon that are Tn4401a, Tn4401b, and Tn4401c were identified in our study. Tn4401a had the highest MIC₅₀s for carbapenems and the Tn4401b exhibited the lowest values. Tn4401c of similar MIC₅₀ values to those of Tn4401a could be explained by the bacterial host, as K. pneumoniae likely, somehow, has the most efficient KPC production. The relatively high meropenem MIC₅₀ by ND transposon encoding the bla_KPC-2 compared to the other isotypes of Tn4401 needed to be further studied.

Though the global spread of KPC-producing Enterobacteriaceae is becoming more common, local epidemiology patterns need to be actively assessed. An awareness of the details of regional epidemiology is fundamental for understanding specific situations. Most endemic cases have occurred via the clonal dissemination of notorious KPC-producing clones, and the South Korean situation was no exception. The disseminated K. pneumoniae clones in South Korea have been linked to the Italian ST307 and Chinese ST11 strains, confirming the importance of international and intercontinental travel for clonal spread. There remains a paucity of antimicrobial options to treat pathogens capable of KPC production, which raises mortality rates. Only a few new drug candidates are in development at present. Combination therapy using both β-lactams and β-lactamase inhibitors that are stable in the presence of KPC enzymes (such as avibactam) is available in other parts of the world, but is not yet accessible in South Korea. Because South Korea is facing endemic dissemination, the situation is critical. Only systematic surveillance will empower researchers to control and prevent the spread of KPC producers.

The research, which has no involvement of human subject but the clinical isolates, does meet the exempt category without approval from Ethics Committee on Human Research of the Health Ministry in South Korea and the study design has not been reviewed by the committee.

SJ and E-JY: contributed to idea conception, designed the study, conducted the experiments, interpreted the data, and wrote the manuscript; JK, DK, HL, JY, and KL: conducted the experiments and interpreted the data.