Klebsiella pneumoniae (K. pneumoniae) is a Gram-negative rod bacterium commonly found in the human respiratory and gastrointestinal tracts (Gu et al., 2018). It is a major pathogen responsible for both hospital-acquired and community-acquired infections. It particularly affects vulnerable populations such as newborns, the elderly, and immunocompromised individuals (Nagaraj et al., 2021). In recent years, Carbapenem-resistant Enterobacteriaceae (CRE) have become a global concern, and the situation in China has worsened. Particularly, Carbapenem-resistant K. pneumoniae (CRKP) has become a significant concern in recent years due to its increasing incidence in hospital-acquired infections (Yan et al., 2019). CRKP is known for its resistance to various antibiotics, including carbapenems and colistin, making treatment challenging. Carbapenems are considered the last resort for treating infections caused by third or fourth-generation cephalosporin-resistant K. pneumoniae (Guh et al., 2014; Gu et al., 2019) These antibiotics are a class of β-lactam antibiotics with a broad spectrum and strong antibacterial activity. However, the rise in carbapenem usage has led to the emergence of carbapenem resistance. The primary mechanism of resistance in CRE is the production of carbapenemases, enzymes capable of degrading cephalosporins, monobactams, carbapenems, and even β-lactamase inhibitors (Nordmann et al., 2009; Munoz-Price et al., 2013). Additionally, other mechanisms include inactivation of outer membrane porins, overexpression of efflux pumps, and increased expression of drug-degrading enzymes (Grundmann et al., 2010; Ho et al., 2016).

The production of carbapenemases, including K. pneumoniae carbapenemase (KPC), NDM, and OXA-48, is a crucial mechanism in CRKP (Guo et al., 2022). Since the first reported case of carbapenemase-producing K. pneumoniae in the United States in 1996 (Munoz-Price et al., 2013), KPCs has become the most prevalent enzyme. In the KPCs (KPC-1, KPC-2, KPC-3, et al.) family, KPC-2 is the most common enzyme (Hu et al., 2020). The bla_KPC-2 gene is located in different mobile genetic elements, such as Tn1721 and IS26 loci, and mainly exists in IncFII plasmids (Liu et al., 2021). In China, approximately 60% of CRKP isolates carry the bla_KPC gene (Guo et al., 2022). The dissemination of the bla_KPC gene involves various mechanisms, ranging from clonal spread to horizontal transfer through plasmids and transposons. This has led to the rapid global spread of carbapenemase-producing K. pneumoniae, with reported cases in dozens of countries and regions across the Americas, Europe, and Asia. In 2004, the first case of KPC-producing K. pneumoniae was reported in Zhejiang Province, China, and since then, KPC-producing isolates have rapidly spread in diverse regions of China (Centers for Disease Control and Prevention, 2013; Munoz-Price et al., 2013).

The spread of KPC-positive K. pneumoniae is mainly associated with different sequence types (STs). The dominant clone of KPC-positive K. pneumoniae is ST11, closely related to ST258. ST258 is most common in North America, Latin America, and Europe (Gu et al., 2018), while ST11 is most common in Asia, and it often causes clonal dissemination and outbreaks in healthcare institutions in China (Andrade et al., 2014). ST11 is a single-locus variant of ST258 (tonB), and they both belong to the clone complex CC258 (Gu et al., 2019; Liao et al., 2020; Guo et al., 2022). Genomic analysis shows that ST11 isolates have high heterogeneity and can be divided into two or three genetic clades based on capsule types, with KL47 and KL64 being the main types and causing outbreaks in many regions (Dong et al., 2018). Due to their strong survival ability and the emergence of multidrug resistance (MDR), the emergence and transmission of MDR K. pneumoniae have become important public health issues worldwide.

This study presents a comparative analysis of the resistant phenotypes and genomic characteristics exhibited by CRKP and Non-CRKP isolates in a Beijing hospital, with particular emphasis on elucidating the distinctive features of high-risk clonal isolates carrying the bla_KPC-2 gene, as well as investigating the structure of mobile genetic elements that facilitate their dissemination across various hospital departments. This comprehensive analysis enhances our understanding of the resistance and transmission dynamics of CRKP within healthcare settings. It enables the formulation of specific strategies for prevention and treatment, coupled with improved surveillance, to more effectively combat CRKP induced hospital-acquired infections.

Due to the overuse of antibiotics, the global spread of MDR and Extensively Drug-Resistant (XDR) isolates is increasing. CRKP, particularly those producing carbapenemases and extended-spectrum β-lactamases, are emerging, accounting for 70–90% of clinical carbapenem-resistant Enterobacteriaceae (CRE) infections in the European Union (EU) and China, leading to high morbidity and mortality (Zhang et al., 2017). Since the first description of a novel K. pneumoniae carbapenemase (KPC), named KPC-1, in 2001, carbapenem resistance in the genus has been rapidly increasing (Yigit et al., 2001). KPC enzymes have become endemic in the northeastern United States and the Mid-Atlantic (Cole et al., 2009), and have been reported in Greece (Cuzon et al., 2008), France (Naas et al., 2005), and China (Wei et al., 2007). In China, approximately 60% of CRKP isolates produce KPC enzymes, predominantly the KPC-2 variant (Hu et al., 2020).

This study compared the antibiotic resistance phenotypes and genomic characteristics of CRKP and Non-CRKP in a Beijing hospital, focusing on high-risk clonal isolates of CRKP carrying the bla_KPC-2 gene and the structure of mobile genetic elements facilitating their spread in hospital departments. Of the 40 K. pneumoniae isolates studied, 31 were identified as CRKP, showing resistance to all tested β-lactam antibiotics and most exhibiting high resistance rates (>70%) to quinolones, consistent with previous studies in China (Wang et al., 2016). KpnR13 was the exclusive isolate in CRKP sensitive to cefepime and amikacin, while KpnR22 was the only one without resistance to imipenem. The increasing impact of KPC-type carbapenemases poses a serious public health threat, necessitating effective measures to control their spread and further development of resistance. In terms of tetracycline antibiotics, while considering the potential for tetracyclines to chelate with metal ions in the culture medium, which can lead to false results or pseudoresistance, our study implemented stringent standardized Etest procedures and carefully selected Mueller Hinton agar as the culture medium. Additionally, E. coli ATCC 25922 was utilized as the quality control strain. These measures were taken to ensure the reliability and reproducibility of our test results, thereby minimizing the potential interference from such interactions. Our isolates not only show resistance to minocycline and doxycycline but also exhibit significant resistance to tigecycline, which is noteworthy. Tigecycline serves as a key medication for treating infections caused by carbapenem-resistant Gram-negative bacteria. It is commonly used in combination with colistin, forming the fundamental treatment approach against CRKP infections (Taniguchi et al., 2017). However, in our study, 12 of the 40 isolates demonstrated resistance to tigecycline with a resistance rate of 30%, including one Non-CRKP isolate. This finding contrasts with previous studies, as historically reported CRKP isolates have shown 100% sensitivity to tigecycline (Gu et al., 2019). Therefore, our findings suggest that the resistance of CRKP is continuously escalating, posing a severe challenge to medical practice that relies on tigecycline as a critical treatment option. The identified mechanisms of tigecycline resistance include overexpression of the resistance-nodulation-division (RND) efflux pump AcrAB encoded by the acrAB genes and the plasmid-mediated RND efflux pump (tmexCD1-toprJ1) (Liao et al., 2020). However, none of these mechanisms were found in our isolates, indicating the possible presence of other unknown resistance mechanisms. Further experimental studies are needed to uncover how these isolates acquire resistance to tigecycline. It is worth mentioning that the resistance rate of Non-CRKP isolates to minocycline and doxycycline was 77.8%, which is higher than that observed in CRKP isolates. This suggests the possibility of distinct resistance mechanisms in Non-CRKP isolates compared to CRKP, indicating a need for further investigation to understand the complexities underlying these resistance patterns. The increasing trend of these resistances poses a greater challenge for the treatment of K. pneumoniae infections, especially for combination drug therapy used to treat severe infections caused by multidrug-resistant Gram-negative pathogens.

In our study, CRKP isolates exhibited high genomic consistency, particularly in specific combinations such as ST11-KL47-OL101. ST11, a high-risk clone, is the predominant sequence type in CRKP isolates and often causes clonal spread and outbreaks in healthcare settings (Liu et al., 2021; Guo et al., 2022). Certain characteristics of the ST11 clone, such as its ability to colonize, form biofilms, and resist phagocytosis (Zhou et al., 2015; Liang et al., 2017), contribute to its multidrug resistance. This explains why the ST11 type has been able to spread globally, while other ST types have not received equal attention. However, noteworthy in our study is that in our CRKP isolates, KpnR13 and KpnR22 are not ST11 but are ST3072 and ST3068, respectively, differing from the previously reported KPC-2 producing CRKP ST48 clone (Gu et al., 2019). This may reflect the evolution of CRKP clones under selective pressure, indicating diversity in the ST types of CRKP isolates carrying KPC-2, although ST11 remains the dominant MLST type. The emergence of these new Non-ST11 sequence types poses a significant challenge to public health experts and clinicians. Studies indicate that KL47 in CRKP is primarily associated with ST11 (Nishida and Ono, 2020), consistent with our findings. Although this clone is not a highly virulent isolate, it has the potential to become one, for example, by acquiring virulence plasmids. In China, KL64 is the most prevalent capsular type among Carbapenem-resistant hypervirulent K. pneumoniae (CR-hvKP) ST11 isolates, due to these isolates containing a higher number of plasmid-encoded virulence genes (Zhang et al., 2020; Wu et al., 2021). Studies have indicated that KL64 evolved from KL47, with KL64 replacing the previously prevalent KL47 (Chen et al., 2023). In our CRKP, we did not find KL64. Research indicates that a variety of virulence factors, such as adhesins and siderophores, are closely linked to urinary and respiratory infections. Notably, yersiniabactin, associated with the siderophore transport system, has been identified as a significant virulence factor in pulmonary infections (Ballén et al., 2021; Hu et al., 2023). In our study, we discovered that the CRKP KL47 strain predominantly harbors yersiniabactin related to iron ion transport and the siderophore system (irp1, ybts, et al.), as well as enterobactin (entA, entB, et al.). Additionally, most strains also carry genes ranging from yagV/ecpE to yagZ/ecpA, which encode the E. coli common pilus (ECP). These genes play a pivotal role in the adhesion process between bacteria and host cells. The ompA gene, encoding the outer membrane protein A, emerges as another key virulence factor that enhances the ability of K. pneumoniae to counteract the host immune defense. Leveraging these virulence factors, K. pneumoniae may effectively evade the host immune system and play a crucial role in both urinary tract and respiratory infections. The combination of CRKP ST11-KL47-OL101 found in our study was reported in this research, which describes a recombination event between this combination of CRKP and the donor CC1763 (excluding ST30) through regional coverage and homologous recombination in the ~154 kb region associated with lipopolysaccharide biosynthesis loci (OL), resulting in the formation of the ST11-KL64-O2v1 combination, which possesses a higher virulence potential (Chen et al., 2023). This underscores the increased need to establish and improve in-hospital monitoring systems and to strengthen surveillance to prevent the occurrence of the aforementioned recombination events and the emergence of highly virulent isolates.

From past research, it is known that the rapid spread of carbapenemase-encoding genes via plasmids, integrons, and transposons is the primary cause of their dissemination (Shen et al., 2016). In our study, we identified a plasmid pKpnR03_2 containing the bla_KPC-2 gene, which includes a structure of IS26-tnpR-ISKpn27-bla_KPC-2-ISKpn6, also reported in previous research. However, unlike prior reports, our pKpnR03_2 plasmid additionally carries the ars and fec gene clusters, which may affect the drug resistance and environmental adaptability of K. pneumoniae. The ars gene cluster is an arsenic resistance system that can be spread through chromosomes or plasmids, primarily responsible for bacterial detoxification of arsenic. The arsenic resistance efflux system uses a two-component ATPase (arsA and arsB) or a single polypeptide (arsB) to transport arsenite, while the enzyme encoded by the arsC gene converts intracellular arsenate [As(V)] to arsenite [As(III)], the substrate for the efflux system. This system enables bacteria to convert arsenate to arsenite and expel it from the cell, maintaining intracellular arsenic concentrations below safe levels in arsenic-polluted environments, giving the isolate a survival advantage in such conditions (Mukhopadhyay et al., 2002). The fec gene cluster is responsible for the uptake and transport of ferric citrate, with fecA acting as a receptor for ferric citrate on the outer membrane, initiating a signaling process involving fecR and fecI, leading to the transcription of the fecABCDE genes responsible for ferric citrate transport (Braun et al., 2022). Iron is a crucial element for many biochemical processes, essential for bacterial growth and reproduction (Nairz et al., 2014). Therefore, isolates carrying the fec gene cluster may have a competitive advantage in iron-limited environments. In our ST11-KL47-OL101 CRKP, most isolates possess the pKpnR03_2 plasmid structure, indicating that isolates with this plasmid structure may have the ability to spread in different environments and may exhibit increased drug resistance. However, further experimental research is required to precisely understand the impact of these gene clusters on the phenotypes of the isolates. On the other hand, in the two CRKP isolates KpnR13 and KpnR22, we did not find the same replicon type IncFll(pHN7A8) as in other CRKPs, nor the pKpnR03_2 plasmid structure. This finding suggests that IncFll(pHN7A8) is associated with the dissemination clone of CRKP ST11-KL47-OL101, and isolates lacking these structures may exhibit different dissemination and resistance characteristics.

In our study, the phylogenetic tree showed that 29 CRKP isolates (93.5%) clustered in the C3 clade, with a median SNP number difference of 32 among isolates, indicating relatively small genetic differences and close evolutionary distances. These CRKP isolates were found in various departments, especially common in the respiratory department, suggesting widespread cross-transmission of CRKP within hospital departments. In contrast, Non-CRKP were mainly found in the nephrology department, and the sources of CRKP and Non-CRKP samples were different: CRKP primarily came from urine, while Non-CRKP mainly from sputum. This finding may imply that different clonal isolates have varying colonization sites within the human body. However, due to the limited number of our isolates and missing data, this conclusion needs further research for validation. Although previous studies have commonly found the clonal spread of CRKP in neurology and intensive care units (ICU) (Wu et al., 2023), our research findings indicates that the primary distribution of CRKP is within respiratory medicine. This finding highlights the importance of the use of invasive devices, whether tracheal intubation or other forms, in the transmission of CRKP. In addition, CRKP-associated infections are not limited to specific parts, organs or tissues (Chen et al., 2022). In fact, they occur more frequently in patients with a variety of invasive devices, and in those with urinary tract infections without catheters, particularly in patients with compromised immune systems. In addition to this, long-term hospitalization and stays in intensive care units are also significant risk factors. The majority of patients in our study were elderly and had compromised immune systems, closely related to the risk factors for CRKP infection. Therefore, our findings further confirm the complexity of the interaction between CRKP infection and various clinical factors, highlighting the importance of tailored preventive measures for high-risk patients in different departments. Under the selective pressure of antimicrobials widely used in clinical treatment, the rapid spread of mobile genetic elements like plasmids or transposons may play a key role in resistance determinants. Given the limited options for selective clinical treatment of infections caused by these isolates in hospitals, developing effective infection control strategies is urgently needed. This includes emphasizing antibiotic stewardship programs, appropriately using targeted antibiotics such as carbapenems and tigecycline, and being vigilant about the emergence of highly virulent isolates. Furthermore, early monitoring of the risk factors mentioned above will help us better control hospital infections and curb the further spread of MDR isolates across various hospital departments.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

XG: Data curation, Formal analysis, Methodology, Visualization, Writing – original draft. YZ: Conceptualization, Resources, Validation, Writing – review & editing. HJ: Methodology, Writing – review & editing. KL: Methodology, Conceptualization, Writing – review & editing. KZ: Conceptualization, Methodology, Writing – review & editing. YY: Validation, Writing – review & editing. JX: Validation, Writing – review & editing. QW: Project administration, Writing – review & editing. XD: Project administration, Writing – review & editing. HW: Data curation, Writing – review & editing. YX: Conceptualization, Project administration, Supervision, Writing – review & editing. WL: Methodology, Validation, Writing – review & editing. ST: Validation, Software, Writing – review & editing. SQ: Conceptualization, Funding acquisition, Writing – review & editing. ZY: Resources, Writing – review & editing.