Klebsiella pneumoniae represents a paradigmatic example of rapid pathogenic evolution in the modern clinical era. Historically recognised as an opportunistic pathogen primarily causing nosocomial infections in immunocompromised hosts, this Gram-negative bacterium has undergone a remarkable transformation that fundamentally challenges traditional concepts of bacterial pathogenicity (Paczosa and Mecsas, 2016). Classical K. pneumoniae (cKP) isolates retain this profile, exhibiting high genetic diversity and typically low virulence (Kochan et al., 2023). The pivotal recognition of this evolutionary shift of this organism occurred in 1986 when Liu et al. documented seven cases of invasive K. pneumoniae infections manifesting as hepatic abscess and septic endophthalmitis in previously healthy community-dwelling individuals without underlying biliary tract disease. This seminal observation marked the emergence of what would later be characterised as hypervirulent K. pneumoniae (hvKp), distinguished by its capacity to cause severe invasive diseases in immunocompetent hosts (Liu et al., 1986).

Subsequent decades witnessed parallel K. pneumoniae evolution along two increasingly convergent pathways. The first involves hypervirulence development, characterised by enhanced capsular polysaccharide (CPS) production (mucoid phenotype), iron acquisition systems (aerobactin, yersiniabactin, salmochelin), and systemic infection establishment from initial colonisation sites (Shon and Russo, 2012; Gu et al., 2018). The second pathway encompasses extensive antimicrobial resistance (AMR) acquisition, particularly emergence of carbapenem-resistant K. pneumoniae (CRKP) which has been linked to mortality rates as high as 42.86% (Chen et al., 2022). Trait convergence has culminated in carbapenem-resistant hypervirulent K. pneumoniae (CR-hvKp), a true “superbug” combining exceptional pathogenic potential with broad AMR profiles (Gu et al., 2018; Roulston et al., 2018). Complexity intensifies through highly clonal populations and persistence of environmental reservoir (hospital surfaces, water systems, medical equipment), facilitating rapid healthcare dissemination (Paczosa and Mecsas, 2016; Farzana et al., 2023).

Given substantial pathogenic potential, rising AMR challenges, and epidemiological significance as both harmless coloniser and pathogen, K. pneumoniae has garnered extensive scientific attention. In recognition of these concerns, the 2024 WHO Bacterial Priority Pathogens List (BPPL) formally designated carbapenem-resistant K. pneumoniae as the top “critical priority” pathogen, assigning it the highest possible risk score (84%) amongst 24 evaluated bacteria (Sati et al., 2025). This designation reflects its high burden of disease, increasing resistance trends, treatment difficulty, and significant public-health impact globally. Given its substantial pathogenic potential, evolving resistance profiles, and epidemiological relevance, K. pneumoniae remains a major threat and therefore continues to warrant focused research and surveillance.

Central to understanding this rapid pathogenic evolution is recognising that the transformative capacity of K. pneumoniae can be driven by mobile genetic elements (MGEs). These genetic structures including plasmids, integrative and conjugative elements (ICEs), insertion sequences (ISs), transposons, and integrons, function as horizontal gene transfer vehicles, enabling rapid virulence and resistance trait dissemination across diverse bacterial populations and accelerating bacterial adaptation to selective pressures (Ernst et al., 2020; Farzana et al., 2023).

Despite significant advances in understanding K. pneumoniae pathogenicity, critical knowledge gaps remain regarding precise mechanisms governing MGE-mediated virulence acquisition and dissemination. Regulatory networks controlling virulence gene expression, molecular bases of MGE mobility, and factors determining successful horizontal transfer events remain incompletely characterised (Haudiquet et al., 2021).

This review provides comprehensive overview of MGEs associated with K. pneumoniae virulence. We focus on mechanisms underlying pathogenic trait transfer between plasmids and chromosomes, and amongst different bacterial strains. By highlighting established and predicted virulence factors and analysing their genetic contexts, this review elucidates MGE roles in shaping K. pneumoniae pathogenic potential.

Data systematically collected from eligible studies included: virulence gene location (chromosome or plasmid); MGEs pertinent to virulence, including size and classification (e.g., group for insertion sequences; replicon type for plasmids); bacterial host harbouring virulence genes; sequence type (ST) and K-antigen (capsular type) of the host bacteria; as well as year, country and source of isolation of bacterial host.

Since the discovery of hvKp, how to genetically and phenotypically classify hvKp has become an urgent question. Early investigations identified the mucoid phenotype and siderophore systems as key contributors to hvKp characteristics (Figure 1A) (Nassif and Sansonetti, 1986; Nassif et al., 1989). This review categorise the virulence genes in K. pneumoniae according to their mechanisms of action (Table 1).

The clinical significance of virulence plasmids was first recognised when Nassif et al. identified a 180 kb plasmid encoding aerobactin and the mucoid phenotype, correlating with virulence phenotypes in K1 and K2 isolates (Nassif and Sansonetti, 1986). Subsequent epidemiological studies demonstrated global dissemination (Struve et al., 2015; Lei et al., 2024). Struve et al. showed that all 30 K1/K2 hvKp strains from patients with liver abscess or community-acquired pneumonia across seven countries (Africa, Asia, Europe, North America) during 1996–2012 harboured pLVPK-like plasmids, though some contained gene deletions (Struve et al., 2015). Additional virulence plasmids underscore their global spread and evolutionary significance. For example, pVir_030666 in Klebsiella variicola encodes multiple virulence determinants including mucoid phenotype regulators (rmpA, rmpA2), aerobactin (iucABCD-iutA), salmochelin (iroBCDN), and yersiniabactin (irp1-2, ybtAEPQSTUX), exhibiting enhanced virulence in larval infection models (Lu et al., 2018).

Most plasmid-mediated virulence in K. pneumoniae is largely driven by this conserved set of genes that enhance capsule formation, iron acquisition, and metabolic fitness. As discussed in the previous part, rmpA and rmpA2 play central roles in increasing transcription of the cps locus, leading to the hypermucoviscous phenotype (Hu et al., 2023). This thick capsule protects the bacterium from complement-mediated killing and phagocytosis, enabling invasive disease even in healthy hosts.

In parallel, plasmid-encoded siderophore systems such as aerobactin (iucABCD-iutA) and salmochelin (iroBCDN) provide high-affinity mechanisms for iron uptake, allowing the pathogen to overcome host nutritional immunity and sustain rapid growth during infection (Hong et al., 2024).

Additional loci, including peg-344 and plasmid-associated variants of the yersiniabactin cluster, further enhance fitness and facilitate tissue invasion. Together, these virulence determinants act synergistically to promote virulence plasmid dissemination, persistence, and severe clinical manifestations.

ICEs are MGEs that integrate into bacterial chromosomes and transfer themselves through conjugation. ICEs consist of cargo modules and functional conjugation systems. Cargo genes, unrelated to their maintenance, provide selective advantages such as antimicrobial resistance genes (ARGs), heavy metal tolerance, or enhanced metabolic capabilities (Johnson and Grossman, 2015). Several ICEs carry virulence-associated cargo genes, contributing to pathogenicity. For instance, P. aeruginosa pathogenicity island 1 (PAPI-1) is a functional ICE that mediate horizontal transfer of virulence traits and may enhance ecological fitness, allowing colonisation and adaptation to specific hosts or environments (Carter et al., 2010).

The widespread distribution of ICEs across bacterial populations has been revealed through large-scale genomic analysis, which identified over 300 putative ICEs across more than 1,000 bacterial genomes. ICEs are found in both pathogenic and non-pathogenic bacteria, with a broader distribution than conjugative plasmids (Guglielmini et al., 2011). ICEs typically integrate at tRNA loci, with one study reporting 73% of strains harboured ICE insertions at one or more of four asparagine tRNA genes (Marcoleta et al., 2016).

They were first characterised in Enterococcus faecalis during the late 1980s, revealing tetracycline resistance transfer without plasmids (Franke and Clewell, 1981). The ICE in K. pneumoniae (ICEKp), extensively characterised by Lin et al., is defined by biosynthetic genes for siderophore yersiniabactin (Lin et al., 2008). ICEKp1 spans approximately 76 kilobases and was identified in hvKp strain NTUH-K2044. ICEKp1 demonstrated substantially higher prevalence in hvKp (38/42) compared to cKp strains (5/32) (Lin et al., 2008). ICEKp1 was also detected in in several members of the K. pneumoniae species complex (Breurec et al., 2016). Recent studies show ICEs are particularly prevalent in K. pneumoniae but occasionally acquired by other Enterobacteriaceae (Putze et al., 2009; Paauw et al., 2010). In a large-scale genomic survey of 2,498 K. pneumoniae isolates, approximately 40% of ICEKp carried ybt, and around 14% carried clb (Lin et al., 2008).

ICEKp elements are characterised by a conserved backbone including P4-like integrase (int), the 29-kb ybt locus, and a ~ 14-kb mobilisation module encoding xis, virB-type IV secretion system (T4SS), oriT, and mobBC (Lin et al., 2008; Marcoleta et al., 2016). Fourteen distinct structural variants, designated ICEKp1 through ICEKp14, have been identified, each associated with specific lineages of yersiniabactin and colibactin loci (Table 3) (Lam et al., 2018). Distinct cargo gene clusters at the right end allow classification into these structural variants.

Most ICEKp variants are associated with a single ybt lineage, suggesting co-evolution, while ICEKp10, harbouring the clb (colibactin) locus, is linked to multiple ybt lineages, indicating repeated acquisition events. The presence of clb has been strongly associated with enhanced pathogenicity in both hvKp and cKp backgrounds, particularly in severe invasive infections. In contrast to ICEKp1, which primarily harbours ybt, iro, and rmpA, ICEKp10 lacks these additional virulence genes but contributes to hypervirulence through colibactin-mediated genotoxicity. Phylogenetic analysis of clb sequences revealed three distinct lineages (clb1, clb2A, clb2B), each associated with a specific ybt lineage, suggesting independent acquisitions into ICEKp. Certain ICEKp variants lacking typical mobilisation machinery indicate evolutionary divergence or specialisation within specific lineages (Table 3) (Lam et al., 2018).

ICEs spread virulence or resistance traits by excising from the chromosome, forming a circular intermediate, and transferring single-stranded DNA to recipient cells through a T4SS (Lee et al., 2010). Similar to other ICEs, ICEKp integrates into the chromosome at a conserved tRNA-associated attachment site (attB) through an ICE-encoded integrase, usually a tyrosine recombinase (Guérillot et al., 2014). After transfer, the ICE is re-integrated into the recipient chromosome, allowing stable maintenance of the virulence locus. The capability of ICEKp to mobilise virulence-associated loci, including ybt and clb, along with metabolic modules facilitating bacterial nutrient uptake, indicates its critical role in shaping pathogenic potential across various clonal backgrounds.

ISs are the simplest form of mobile elements, typically consisting of a transposase gene flanked by short, inverted repeats. Though small, they have significant functional impacts by disrupting genes, modifying promoter activity, or facilitating genomic rearrangements (Wei et al., 2025). In K. pneumoniae, IS element involvement in virulence has been demonstrated in recent years, particularly through disruption of capsule biosynthesis gene wcaJ (Wang et al., 2022). Hypervirulent isolates of ST23-K1 strain harbour IS elements, such as ISKpn26 or ISKpn74, which insert within wcaJ and induce frameshift mutations. This insertion significantly diminishes CPS production (Wang et al., 2022). Furthermore, IS5/ISKox3 elements have been observed associated with key capsule synthesis genes (wcaJ, wza, wzc) in hvKp. This association leads to non-mucoid phenotypes in vitro. Importantly, excision of these IS elements restores capsule production in vivo, thereby recovering hypervirulence capacity (Wei et al., 2025). The specific IS elements associated with K. pneumoniae virulence are detailed in Table 4.

IS-mediated inactivation of capsule biosynthesis genes conferred lower fitness cost and enhanced conjugation frequency of a bla_KPC-2 resistance plasmid (Wang et al., 2022). These findings highlight a dynamic IS-mediated “capsule ON–OFF–ON” mechanism that not only alters virulence but also promotes horizontal transfer of multidrug resistance in hvKp (Wei et al., 2025). For instance, ISKpn74 has been identified as a significant factor altering virulence through two distinct yet contrasting mechanisms. Typically, plasmid-associated, ISKpn74 is integrated upstream of rmpA and rmpA2, enhancing expression of virulence traits such as hypermucoviscosity and increased siderophore production (Huang et al., 2024). Conversely, another study identified ISKpn74 within the chromosomal framework of an ST20 isolate, where its insertion occurs between K and O antigen loci, suggesting potential disruption of capsule biosynthesis pathways (Eger et al., 2021).

Beyond influencing strain virulence, ISs significantly contribute to structural evolution of virulence elements by facilitating transfer between plasmids and chromosomes. Research shows IS elements like ISKpn28 and IS26 are instrumental in forming large fusion plasmids, arising from recombination events between classical virulence plasmids (such as pK2044-like) and resistance plasmids (like pNDM-Mar-like) (Wang S. et al., 2024). These hybrid plasmids contain key virulence genes, including rmpA2 and the iucABCD-iutA operon. Furthermore, IS-mediated integration of these genetic fragments into the chromosome has been documented (Tian et al., 2022; Wang S. et al., 2024).

Transposons mobilise large genetic regions, including virulence cassettes, facilitating horizontal transfer across strains. The rapid evolution of CR-hvKp stems from transposon-mediated co-selection and co-transfer of virulence and resistance determinants (Gray et al., 2024). Tn3 family members frequently mobilise virulence genes in K. pneumoniae (Tian et al., 2023). A conserved ~16.2 kb composite transposon on IncFIB/FII plasmids carries the iuc3 aerobactin operon, enabling dissemination across animal and human reservoirs (Kaspersen et al., 2023). Tn7074-like transposons integrate complete virulence cassettes (rmpA2, iucABCD-iutA, peg-344) into conjugative plasmids, significantly enhancing recipient strain virulence (Li et al., 2025).

Integrons capture, assemble, and express gene cassettes encoding virulence factors (Chen et al., 2018). Integrons lack self-mobility and rely on ISs, transposons, or plasmids for dissemination (Farajzadeh Sheikh et al., 2024). Integrons, particularly class 1 integrons, are capable of shaping K. pneumoniae virulence potential (Chen et al., 2018). Studies found 74% of clinical isolates harboured class 1 integrons, with higher wcaG capsule gene prevalence (Derakhshan et al., 2016). Another study revealed 19.1% of integrons co-localised with virulence loci, including iucABCD-iutA, rmpA2, and peg-344 (Li et al., 2025).

CR-hvKP strains exhibit both hypervirulence and high-level AMR, spreading globally and presenting significant clinical threats (Karampatakis et al., 2023; Chen et al., 2020; Ahmed et al., 2021; Chen et al., 2021). An ST11 K. pneumoniae strain gained virulence after acquiring pLVPK-like plasmid pVir-CR-hvKP4, despite a 41,231-bp deletion including rmpA and iro loci (Gu et al., 2018). Salmochelin (iro) appears non-essential for systemic infection (Russo et al., 2015), and rmpA/rmpA2 redundancy maintains pathogenicity (Cheng et al., 2010; Russo et al., 2015). Aerobactin production was predicted to be essential for hypervirulence (Russo et al., 2014).

There is evidence of harbouring hybrid plasmid pVir (297,984 bp), combining sequences from virulence plasmids pK2044/pLVPK and resistance plasmid pPMK-NDM in ST11 strain from Taiwan (Fang et al., 2025). Despite harbouring iroBCDN, iucABCD-iutA, rmpA, and rmpA2, it showed limited virulence in mouse models, suggesting unidentified virulence factors exist in the truncated regions from pK2044/pLVPK.

Tn3-family transposons mediate fusion between virulence and resistance plasmids, creating self-transmissible hybrids encoding both determinants (Ramirez et al., 2014; Nicolas et al., 2015; Tian et al., 2023). Integrons within multidrug-resistant IncFII plasmids, flanked by IS26, facilitate simultaneous horizontal transmission of resistance and virulence genes (Li et al., 2025).

K. pneumoniae exemplifies a high-risk pathogen combining AMR and hypervirulence through MGEs acquisition, threatening both immunocompromised and healthy populations globally (García-Cobos et al., 2025). Despite extensive research on resistance mechanisms, critical knowledge gaps persist in the current understanding of this organism: studies remain confined to limited clonal groups and serotypes (Wyres et al., 2019); the hypervirulence-resistance relationship is underexplored (García-Cobos et al., 2025); virulence gene functions require deeper characterisation (Derakhshan et al., 2016); and MGEs-mediated mobility mechanisms remain incompletely understood (Haudiquet et al., 2021).

While over 200 published reviews have examined K. pneumoniae virulence since 1977, few address how virulence traits are acquired, mobilised, and maintained molecularly (Russo and Marr, 2019; Yang et al., 2021). Critically, plasmid-chromosome exchanges and smaller mobile elements (ISKpn74, Tn3 transposons) facilitating gene dissemination have received insufficient scientific attention (Haudiquet et al., 2021).

Understanding MGEs-mediated virulence dissemination is essential for advancing surveillance and therapeutics. Characterisation of MGE-mediated virulence gene mobility, including the roles of integrases, transposases, insertion sequences, and recombination hotspots, can enhance molecular diagnostic platforms by enabling the detection of both established virulence markers (such as rmpA, iucABCD, and clb) and their associated genetic mobility signatures (Farajzadeh Sheikh et al., 2024). Such integrated surveillance approaches would improve the identification of high-risk hypervirulent strains with enhanced horizontal transfer potential, thereby informing more effective infection control measures by distinguishing between clonal dissemination and independent MGE-mediated acquisition events across diverse lineages (Farajzadeh Sheikh et al., 2024). Furthermore, elucidating the molecular mechanisms governing ICE excision, plasmid conjugation, and IS-mediated genomic rearrangements may reveal novel therapeutic targets, including MGE-encoded integrases, relaxases, and T4SS components, whose inhibition could disrupt the horizontal spread of hypervirulence and AMR determinants (Wang Q. et al., 2024). Given the convergence of hypervirulence and carbapenem resistance traits mediated by hybrid plasmids and composite transposons, targeted disruption of MGE mobility machinery represents a promising strategy to limit the emergence and dissemination of CR-hvKp strains, which pose formidable challenges to global public health.