Carbapenem-resistant Enterobacterales (CREs) have been listed as an urgent threat to public health. CRE has emerged as a challenge in healthcare settings due to the limited treatment options (CDC, 2019a). CREs are Enterobacterales that confer resistance to at least one of the carbapenem antibiotics or produce a carbapenemase enzyme (CDC, 2019b, 2015). Among CRE infections, carbapenem-resistant Klebsiella pneumoniae (CRKP) is the most critical pathogen which was listed in the World Health Organization (WHO) priority list of antibiotic-resistant bacteria (Tacconelli et al., 2018). Studies pertaining to CRKP in the past decades have mainly focused on carbapenemase-producing CRKP (C-CRKP), predominantly due to the rapid global dissemination of carbapenemases (Lee et al., 2016). Nonetheless, non-carbapenemase-producing carbapenem-resistant Klebsiella pneumoniae (NC-CRKP) is of huge clinical importance yet underestimated due to a lack of relevant research.

NC-CRKP does not produce carbapenemase but can exhibit resistance to carbapenems through a combination of chromosomal mutations (e.g., porin gene mutation, overproduction of efflux pump, and/or alterations in cell structure such as penicillin-binding protein) and acquired non-carbapenemase resistance mechanisms (acquisition or upregulation of a β-lactamase) (CDC, 2019b, 2015; Westblade, 2018). Although studies suggested that C-CRKP was more virulent than NC-CRKP (Goodman et al., 2016; Tamma et al., 2017; Richards et al., 2017), the mortality rate of NC-CRKP was high (Su et al., 2018) and almost similar to C-CRKP (Orsi et al., 2013). Therefore, the significance of the disease burden posed by NC-CRKP should not be overlooked.

In Malaysia, studies on NC-CRKP are scarce, but the limited evidence garnered from these studies has demonstrated great diversity and dynamicity of local CRKP populations (Lau et al., 2021; Kong et al., 2022; Lee et al., 2022; Zaidah et al., 2017; Zawawi et al., 2021; Yeow, 2020; Heng et al., 2019; Mohamed et al., 2018; Low et al., 2017). An increasing trend of NC-CRKP was observed in our hospital setting. The number of cases of NC-CRKP increased from one in 2013 to 17 cases in 2018, followed by a slight decline to 12 cases in 2019 (Lee et al., 2022). NC-CRKP may potentially emerge through de novo mutations or genetic reassortment within carbapenem-sensitive Enterobacterales under antimicrobial selective pressure (Marimuthu et al., 2019). Although the prevalence of NC-CRKP remained lower than C-CRKP in Malaysia, the increasing trend should not be ignored since the treatment for NC-CRKP may be different from C-CRKP infections as NC-CRKP employ different mechanisms for carbapenem resistance. In addition, the lack of epidemiological data and an unknown mechanism may pose significant challenges in treating NC-CRKP.

In our recent study, 54 NC-CRKP isolated from 2013 to 2019 in Malaysia have been studied (Lee et al., 2022). The study identified the loss of porins (46.3%, 25/54) and the presence of AmpC β-lactamase gene (bla _DHA), along with other non-carbapenemase β-lactamase genes (bla _TEM, bla _SHV, bla _CTX-M, and bla _OXA-1) in the NC-CRKP isolates. In addition, a minimum inhibitory concentration (MIC) reduction assay using 26.3 µg/ml phenylalanine-arginine β-naphthylamide (PAβN), a Resistance Nodulation Division (RND) efflux pump inhibitor (Dupont et al., 2016), was performed to address conflicting findings from previous studies regarding whether PAβN-inhibited efflux pumps reduce carbapenem MIC value by 1 to 4 times (Khalid and Ghaima, 2022) or elevate resistance to carbapenem antibiotics (Saw et al., 2016). While all NC-CRKP isolates demonstrated the same MIC value for ciprofloxacin, four isolates exhibited increased susceptibility to ciprofloxacin but also showed increased resistance to carbapenem (Lee et al., 2022). In this study, we sought to investigate the genomic features, carbapenem resistance mechanisms, and virulence determinants of these four NC-CRKP isolates, collected from a tertiary teaching hospital in 2018, using whole-genome sequencing (WGS) to shed light on the mechanisms of carbapenem resistance in NC-CRKP isolates attributable to potential chromosomal mutations.

The emergence of NC-CRKP has posed a significant challenge in healthcare settings (Hovan et al., 2021). The absence of major carbapenemase genes, such as bla _NDM, bla _OXA-48, bla _IMP, bla _VIM, and bla _KPC, in NC-CRKP emphasized the importance of exploring the other resistance mechanisms acquired by non-carbapenemase-producing carbapenem-resistant isolates. Efflux pumps or porins in combination with β-lactamase genes, have been identified as potential contributors to carbapenem resistance in Klebsiella pneumoniae (Nicolas-Chanoine et al., 2018). Nicolas-Chanoine et al. (2018) reported that bla _DHA-1-producing isolates conferred resistance to ertapenem when there was an increase in efflux or porin loss, and to imipenem when both mechanisms were associated. In our study, both isolates 1804-1 (mono-resistant to ertapenem) and 1805-12 (resistant to all three carbapenems) harbouring bla _DHA-1 did not have porin loss but exhibited a missense mutation in the ompK36 porin and a frameshift insertion mutation in efflux protein. Furthermore, Nicolas-Chanoine et al. (2018) also found that bla _CTX-M-15-producing isolates developed resistance to ertapenem in the absence of porins. In our study, isolates 1804-1 and 1805-11 harbouring bla _CTX-M-15 also exhibited a frameshift insertion mutation in efflux protein. Only isolate 1805-11 (resistant to all three carbapenems) exhibited ompK36 porin loss, whereas isolate 1804-1 (mono-resistant to ertapenem) exhibited a missense mutation in ompK36 porin but no porin loss. Additionally, previous studies have presented conflicting results on whether PAβN-inhibited efflux pumps reduce carbapenem MIC value by 1 to 4 times (Khalid and Ghaima, 2022) or elevate resistance to carbapenem antibiotics (Saw et al., 2016). To address this controversy, a MIC reduction assay was conducted using the RND efflux pump inhibitor PAβN on 54 NC-CRKP isolates (Lee et al., 2022). Our results concurred with Saw et al. (2016), indicating that the PAβN-inhibited efflux pump did not contribute to carbapenem resistance although increased susceptibility to ciprofloxacin.

Throughout the years, ST101 has been identified as the predominant sequence type among C-CRKP in Malaysia (Low et al., 2017). This is in concordance with other countries such as Italy (Mammina et al., 2012) and Spain (Fuster et al., 2020). However, none of the isolates in this study were of ST101. This suggested the genetic heterogeneity among the NC-CRKP isolates.

Besides the loss of porins as previously reported by Lee et al. (2022), this study further revealed the silent mutations of ompK35 porin in all isolates and the missense mutations in ompK36 porin. The loss or alteration of ompK36 porin has been associated with carbapenem resistance (Kaczmarek et al., 2006; Clancy et al., 2013).

This study has raised a few concerns. Firstly, the identification of IncR plasmids suggested the potential for horizontal gene transfer and dissemination of resistance determinants. The IncR plasmids detected in this study were identical to the plasmid pK245 (GenBank accession number DQ449578) that confers quinolone resistance and ESBL activity (Chen et al., 2006). IncR plasmids carry various antimicrobial resistance genes that mediate resistance to β-lactam, aminoglycoside, phenicol, tetracycline, sulfonamide, and trimethoprim (Schwanbeck et al., 2021).

Secondly, the presence of plasmid-encoded disinfectant resistance gene qacE identified in the isolate 1805-11 could reduce the disinfectant susceptibility and confer quaternary ammonium compound protection in K. pneumoniae isolates (Liu et al., 2024; Paulsen et al., 1993). A wide distribution of disinfectant resistance genes resulting from the widespread use of disinfectants could impose selective pressure on antimicrobial resistance isolates (Liu et al., 2024; Abuzaid et al., 2012). This was in accordance with a previous study that suggested that NC-CRKP was confined to individuals and settings with very high levels of antimicrobial selection pressure (Richards et al., 2017).

Thirdly, the examination of virulence determinants according to the VF class highlighted potential pathogenicity in NC-CRKP, including adherence, antiphagocytosis, efflux pumps, iron uptake mechanisms, regulation, secretion system, serum resistance, and autotransporter. The presence of these VF suggested the ability of NC-CRKP to evade host defenses and establish infections. In this study, all NC-CRKP isolates were resistant to ciprofloxacin and carried the AcrAB efflux pump. As AcrAB multidrug efflux system is crucial for expelling harmful substances from the bacterial cellular environment, it may also influence the isolates’ resistance to antimicrobial agents and other toxic compounds (Padilla et al., 2010).

In this study, a few mutations identified could play important roles in resistance mechanisms. A frameshift deletion mutation was identified in the cation efflux system protein CusF (cusF) in isolate 1801-2. The cation efflux system protein CusF, also known as copper-binding periplasmic protein CusF, is present in the periplasm. It binds and transports periplasmic metal cations such as copper (Cu; Cuprum) or silver (Ag; Argentum) to the Cus (Cu sensing) system for efflux (Loftin et al., 2005; Kim et al., 2011; Franke et al., 2003; Lee and Choi, 2020). In previous studies, the periplasmic metallochaperone CusF demonstrated upregulated expression under the induction of silver cation (Imran et al., 2024) and copper cation in both aerobic and anaerobic conditions (Zulfiqar and Shakoori, 2012; Egler et al., 2005). The protein CusF utilizes cation and methionine interactions to bind Cu⁺ or Ag⁺ for delivery to the Cus system, as determined by the site-directed mutagenesis of methionine residue to isoleucine (substitution of methionine residue) or NMR chemical shift analysis (Loftin et al., 2005; Franke et al., 2003; Randall et al., 2015). We deduced that the frameshift deletion mutation in the cation efflux system protein CusF, which particularly affects the methionine residues, as shown in our study, could lead to a conformational change in protein CusF, thus leading to the accumulation of copper ions in bacteria. Previous studies reported that copper ions may form complexes with meropenem, leading to meropenem structure degradation and thereby contributing to carbapenem resistance (Božić et al., 2018; Božić Cvijan et al., 2023). Metal ions may diminish the efficiency of antimicrobial activity through respective efflux pumps by forming complexes, that eventually may lead to the degradation of antimicrobial agents. Further investigation can be performed to explore the potential interaction between metal ions and antimicrobial agents with their impact on the efficiency of antimicrobial therapy.

Previously, Perron et al. (2004) reported that a co-increase in resistance to heavy metals and imipenem, along with the suppression of OprD porin, was associated with the overexpression of a mutated czc efflux system in Pseudomonas aeruginosa (Perron et al., 2004). This czc (cobalt/zinc/cadmium) efflux system is similar to the nickel/cobalt efflux protein RcnA in terms of the function where both efflux protein/system can export cobalt cations out of the cytosol when excessive toxicity is detected. In the present study, frameshift insertion mutations in the nickel/cobalt efflux protein RcnA (A8C11_RS17640) were identified in NC-CRKP (isolates 1804-1, 1805-11 and 1805-12). This mutation may potentially contribute to the carbapenem resistance, which is similar to the mechanism found in Pseudomonas aeruginosa. Additionally, the inactivation of RcnA (yohM), a nickel and cobalt resistance gene in Escherichia coli, conferred sensitivity to nickel and cobalt (Rodrigue et al., 2005). The yohM gene possesses a remarkable histidine-rich region composed of histidines, aspartate, and glutamate residues, which may function as nickel storage (Rodrigue et al., 2005). Rodrigue et al. (2005) reported that a mutant transformed with a high number of copies of vector harbouring yohM greatly enhanced the wild-type resistant levels by 100-fold (Rodrigue et al., 2005). In our study, a frameshift insertion mutation in the nickel/cobalt efflux protein RcnA (A8C11_RS17640) introduced additional histidine (H), aspartate (D), and glutamate (E) residues. This modification likely enhances the binding of metal ions by the histidine-rich region, potentially leading to an enhanced resistance level. This could enhance resistance towards imipenem, as described in the mechanism found in Pseudomonas aeruginosa. Nevertheless, future studies are warranted to confirm our hypothesis.

Based on the analysis, the potential mechanisms of carbapenem resistance in the four isolates were postulated: 1) Isolate 1801-2 may confer carbapenem resistance through ompK36 porin loss and efflux pump mutation (cusF) combined with the acquisition of non-ESBL (bla _SHV-11 and bla _OXA-1) genes. 2) The resistance of carbapenem in isolate 1805-11 may be due to the chromosomal mutations such as ompK36 porin loss and efflux pump mutation (A8C11_RS17640) in combination with the acquisition of non-ESBL (bla _TEM-1) and ESBL (bla _CTX-M-15 and bla _SHV-28) genes. 3) A diverse combination of β-lactamase genes has been identified in isolate 1804-1 and could be coupled with ompK36 porin mutation and efflux pump mutation (A8C11_RS17640) for carbapenem resistance. 4) Isolate 1805-12 may confer carbapenem resistance through ompK36 porin mutation and efflux pump mutation (A8C11_RS17640) in combination with the acquisition of non-ESBL (bla _SHV-1) and AmpC β-lactamase (bla _DHA-1) genes. However, future studies are warranted in order to validate the genomic findings.

Based on the WGS data, we have identified potential contributing factors to carbapenem resistance mechanisms and elucidated the genomic characteristics as well as the virulence determinants of NC-CRKP isolates. Despite the small number of NC-CRKP isolates in this study, the relevance of our findings sheds light on this issue. To enhance the robustness of future studies, a larger sample size should be included to better assess the mechanisms of carbapenem resistance. Moreover, experimental confirmation through functional validation of the identified resistance genes and mutations is quintessential to confirm the roles of these genetic changes in resistance. Unraveling these potential mechanisms of carbapenem resistance and understanding the genetic heterogeneity within NC-CRKP isolates are imperative for devising effective strategies to control their dissemination and combat antimicrobial resistance in healthcare settings.