Multidrug-resistant hypervirulent Klebsiella pneumoniae: an evolving superbug

ABSTRACT

Multidrug-resistant hypervirulent Klebsiella pneumoniae (MDR-hvKP) combines high pathogenicity with multidrug resistance to become a new superbug. MDR-hvKP reports continue to emerge, shattering the perception that hypervirulent K. pneumoniae (hvKP) strains are antibiotic sensitive. Patients infected with MDR-hvKP strains have been reported in Asia, particularly China. Although hvKP can acquire drug resistance genes, MDR-hvKP seems to be more easily transformed from classical K. pneumoniae (cKP), which has a strong gene uptake ability. To better understand the biology of MDR-hvKP, this review discusses the virulence factors, resistance mechanisms, formation pathways, and identification of MDR-hvKP. Given their destructive and transmissible potential, continued surveillance of these organisms and enhanced control measures should be prioritized.

Plain Language Summary

Antibiotics play an important role in treating diseases and saving lives. However, due to their overuse, they are becoming less effective. When antibiotics are used to treat an infection, some bacteria are killed but others are able to survive exposure to the antibiotic and continue to grow. Bacteria are always changing to survive so, over time, more and more bacteria become resistant to antibiotics. This makes infections more difficult to treat. To help patients recover faster, scientists need to study these bacteria and find better ways to control them. As these bacteria can spread and cause serious infections, it’s important to track them and take stronger safety measures.

1. Introduction

Klebsiella pneumoniae is on the rise in terms of incidence and antibiotic resistance and has been included in the World Health Organization (WHO) 2024 list of priority pathogens. The emergence of drug-resistant strains in certain populations or geographic regions (such as China) has the potential to spread globally, posing a serious threat to public health [1–3]. There is currently no FDA-approved vaccine for K. pneumoniae. Classical K. pneumoniae (cKP) has the powerful ability to acquire multiple components which confer antibiotic resistance. In the 1980s, a unique case of liver abscess and endophthalmitis caused by K. pneumoniae was first reported in Taiwan, and the pathogen was designated hypervirulent K. pneumoniae (hvKP) [4]. Since then, hvKP has been recognized as a new hypervirulent variant of K. pneumoniae associated with high mortality due to its high pathogenicity and virulence [5]. HvKP is serologically specific and usually occurs in several limited capsular types, dominated by K1 and K2 [6–8]. The conventional wisdom is that hvKP is sensitive to antibiotics. However, in recent years, owing to the misuse of antibiotics, multidrug-resistant hypervirulent Klebsiella pneumoniae (MDR-hvKP) has emerged. Carbapenem-resistant hvKP (CR-hvKP) was first reported in China, where the number of ST11-type CR-hvKP cases has increased dramatically since 2017 [9,10].

The clinical manifestations of MDR-hvKP infections are complex and diverse, as they include diseases commonly caused by hvKP infection and cKP-associated hospital-acquired infections. Additionally, MDR-hvKP often occurs in hospital- and community-acquired severe infections, as illustrated in Figure 1, and usually presents as multisite infections, including meningitis and epidural abscess, endophthalmitis, pneumonia, bacteremia, liver abscess, necrotizing skin and soft tissue infection, and bone infection [10,11], which can be life-threatening.

Figure 1.

Common sites of primary and metastatic infections caused by multidrug-resistant hypervirulent Klebsiella pneumoniae (created with BioRender.com).

MDR-hvKP is a combination of the hypervirulent characteristics of hvKP and the drug-resistant characteristics of cKP strains. The convergence of high virulence and resistance is a result of multiple mechanisms. These mechanisms include the horizontal transfer of drug-resistant and hypervirulent genes between strains, the insertion of nonsense gene fragments, and mutations. Owing to the limited therapeutic options, patients infected with MDR-hvKP often have long hospitalizations and poor treatment outcomes, placing tremendous pressure on clinicians. In this review, we present new research developments related to the virulence and drug resistance mechanisms of MDR-hvKP strains, especially some new drugs such as ceftazidime/avibactam (CAZ-AVI), ceftolozane-tazobactam and meropenem-vaborbactam. In addition, we have summarized the current knowledge of MDR-hvKP formation pathways in three areas. A variety of methods for identifying MDR-hvKP are described, which will be significant for early and accurate identification.

2. Primary virulence factors of MDR-hvKP

2.1. Capsular polysaccharides

Capsular polysaccharides (CPSs) are important virulence factors in hypervirulent strains [12]. Traditionally, the typing and classification of K. pneumoniae has been based on capsular serotyping or K typing, and at least 160 different K types have been identified [13]. When carbapenem-resistant K. pneumoniae (CRKP) strains begin to acquire virulence plasmids, some capsule types previously present only in cKP strains, such as K47 and K64, appear in MDR-hvKP strains [14]. Serotypes K1 and K2 are the most common CR-hvKP serotypes. Current research shows that the K1 serotype is mainly the ST23 serotype, whereas the K2 serotypes are mainly the ST65 and ST86 serotypes [15,16]. Yang et al. reported that both K47 and K64 serotypes are mainly ST11 [17]. ST11 outbreaks have occurred in China, of which K64 is the most common capsular type, accounting for up to 25% of CR-hvKP infections in certain regions [18].

Overexpression of rmpA increases CPS production [19]. Overproduction of CPS helps bacteria escape host immunity through antiphagocytic activity, bacitracin neutralization, and inhibition of complement and inflammatory factor responses and promotes the persistence and reproduction of bacteria in the host [20–23]. Capsules are essential for the intracellular survival of K. pneumoniae [24]. The capsule promotes hvKP survival in the bloodstream by protecting the bacteria from being captured by liver-resident macrophage Kupffer cells [22]. However, hvKP phagocytosis is also a part of the infection process, as neutrophils carrying engulfed pathogens may circulate in the blood of mice and transport hvKP to distant sites, causing abscesses to spread [25]. In addition, the thick capsule acts as a sturdy shell around the bacteria, improves bacterial viability, and affects DNA absorption and thus limits horizontal gene transfer, which may partly explain why drug-resistant strains are derived primarily from cKPs.

2.2. Hypermucoviscosity phenotypes

RmpA was identified as a regulator of the hypermucoviscosity (HMV) phenotype in 1989 [26], and early studies associated this hypercapsule production with the rmpA (or rmpA2) gene [27]. However, subsequent studies indicated that HMV is a complex phenotype associated with excessive CPS production. Zhang et al. reported that HMV was observed in only 6 of 72 (8.3%) CRKP isolates containing rmpA and rmpA2 collected from municipalities across China between 2015 and 2017 [28]. rmpC is another CPS expression regulator which is encoded by the same operon as rmpA, and mutations in rmpC can reduce the expression of the CPS gene and CPS production; however, these strains have wild-type levels of HMV, which supports the hypothesis that capsule hyperproduction is not a requirement for the HMV phenotype [29]. Emerging research suggests that rmpD, located between rmpA and rmpC, is required for HMV [30]. The rmpD gene encodes a protein which increases the length of podoplanar polysaccharide chains [31]. In addition, when plasmids carrying drug resistance genes are obtained, strains can reduce their energy load and adapt rapidly by mutating rfaH and wcaJ to reduce CPS production [32]. wcaJ mutations were found in four ST23-K1 CR-hvKPs, resulting in reduced capsule synthesis, reduced bacterial virulence, and lower fitness costs, suggesting that wcaJ is a potential factor promoting hypervirulence and drug resistance aggregation [33]. This reflects the complex relationship between CPS and HMV, which is not a one-to-one correspondence.

2.3. Iron acquisition systems

Iron uptake lays the foundation for hvKP success hvKP strains have the ability to produce four different siderophores: enterobactin, salmochelin, yersiniabactin, and aerobactin. The genes encoding aerobactin, salmochelin, and yersinibactin are located on the plasmid; thus, these genes can shuttle between strains via horizontal transfer of the plasmid [34]. Russo et al. demonstrated that aerobactin is the dominant siderophore and a critical virulence factor for the hvKP genetic background [35] and has been shown to potentiate invasive hvKP infection in vitro and in mouse models [36]. Some scholars have defined hvKP strains as aerobactin positive [37–39]. Recent studies have shown that the main gene encoding aerobactin, iucB, mediates the inhibition of rmpA expression and influences CPS synthesis [40]. Redundant siderophores are present in hypervirulent strains; their exact mechanism is unknown but may be the result of bacteria adjusting their growth status in different environments according to the cost of adaptation. Ferric uptake regulator (Fur) is often responsible for regulating genes which affect iron uptake and metabolism, influence bacterial virulence, and control bacterial colonization [41]. Type 3 fimbriae are associated with bacterial biofilm formation, and it has been suggested that Fur can inhibit type 3 fimbriae via mrkHI and impede biofilm formation [42]. Chu et al. showed that, in the CG23-I subline, Fur affects the expression of type 3 fimbriae and hypermucoid capsules in the presence of iron through the IroP protein encoded by a virulence plasmid, thereby affecting biofilm formation and bacterial colonization [43]. This regulatory mechanism elucidates how hvKP regulates its capsule and fimbriae, thereby enhancing its adaptation.

2.4. Virulence plasmids

HvKP strains usually carry two similar plasmids, pK2044 and pLVPK, which have been isolated from K. pneumoniae NTUH-K2044 (K1, ST23) and K. pneumoniae CG43 (K2, ST86), respectively; both are IncFIIK/IncFIBK plasmids [44]. IncFIIK and IncFIBK are the most common plasmids in K. pneumoniae and are important vectors for carbapenem resistance genes, most likely from the horizontal transfer of genes from unidentified species [45,46].

The pLVPK plasmid carries a variety of virulence-associated genes, including the capsule-related rmpA/rmpA2 genes and aerobactin and salmochelin gene clusters (Figure 2). K. pneumoniae virulence is significantly increased by carrying pK2044, pLVPK, or pLVPK-like plasmids [10,47]. pLVPK confers virulence by promoting the spread of K. pneumoniae infection but cannot increase the resistance of strains to serum killing or phagocytosis [48]. The detection of virulence plasmids or virulence plasmid-specific gene loci is helpful for the identification of hvKP strains.

Figure 2.

Virulence features and resistance mechanisms of multidrug-resistant hypervirulent Klebsiella pneumoniae (mdr-hvKP). The upper half of the picture shows the resistance genes and the main drug resistance mechanisms of mdr-hvKP, and the lower half of the picture depicts the hypervirulence factors. Resistance genes are located on chromosomes or plasmids, whereas highly virulent phenotype-associated genes are often encoded on virulence plasmids. The overexpression of efflux pumps to extrude antibiotics, prevention of inwards flow by modulating pore proteins, production of antimicrobial drug- inactivating enzymes, and biofilm formation while antimicrobial drug resistance genes are seeded with the help of plasmids, etc., are among the major mechanisms of resistance in mdr-hvKP. The transcriptional regulators rmpA/rmpA2 are associated with the hypercapsule and HMV phenotypes. Salmochelin and aerobactin are characteristic virulence factors, and their coding genes iro and iuc are both located on virulence plasmids (created with BioRender.com).

2.5. Additional virulence factors

Biofilm-forming capacity also contributes to hypervirulence. Biofilms are a collection of microorganisms which promote bacterial attachment to inert or living surfaces by secreting various active substances including proteins, polysaccharides, and extracellular DNA [49]. Bacteria within biofilms become more resistant to antibiotics and evade host immune responses, making the eradication of bacterial infections more difficult [50]. Wen et al. reported that compared with cKPs, hvKPs produced tighter biofilms with improved viability and antibiotic resistance [51]. Various studies have attributed the hvKP biofilm phenotype to the capsule [52], whereas others have suggested that the absence of a capsule promotes biofilm formation [53]. Biofilm formation has been reported to be related to the O antigen because polysaccharide capsules remain on the outer surface of bacteria by interacting with the O antigen [54]. Fimbriae are other virulence factors of hvKP. Adhesion is an important step in biofilm formation, and type 1 and especially type 3, as the main adhesion structures, can promote the production of biofilms and are necessary for the binding of hvKP to medical devices [20,55]. Based on this finding, some studies have concluded that biofilm formation by hvKP is relatively weak because of the thick capsule covering the fimbriae [56]. Polyphosphate kinase 1 (PPK1) is an enzyme which regulates a variety of physiological processes in bacteria, and a recent study revealed that PPK1 mutations and deletions are associated with impaired biofilm and capsule formation and reduced virulence in hvKP [57].

The type VI secretion system (T6SS) has been extensively documented in Vibrio cholerae, Bacteroidetes, Proteobacteria, and especially Enterobacteriaceae [58,59], and studies have shown that T6SS is positively correlated with K. pneumoniae virulence [60,61]. It can inhibit the growth of other species and interfere with bacterial colonization by disrupting the cell envelope, mediating target cell lysis, affecting DNA integrity, interfering with cell division, and depleting energy resources [62]. A recently published study revealed that T6SS promotes hvKP infection by affecting the local immune microenvironment of lung tissues [63], providing novel insights into how K. pneumoniae evades host innate immune responses. T6SS can enhance the competitive advantage of hvKP strains over gut microbiota constituents [64] and aid in the colonization of the gastrointestinal tract [65]. Capsule, type 1 fimbriae, and biofilm formation are all beneficial for the intestinal colonization of hvKP, but CPS-mediated colonization increases the fitness cost [66]. T6SS enhances bacterial adhesion and invasion by mediating the expression of type 1 fimbriae, which are essential for hvKP survival and subsequent colonization in vivo [64]. Studies have shown that T6SS-positive strains have greater biofilm-forming ability than T6SS-negative strains [67]. Since the genes involved in iron absorption mechanisms are near T6SS-related genes, it is possible that T6SS is involved in iron import by hvKP [64].

A recent study revealed that microcin E492, a product of the genomic island E492 (GIE492), and colibactin, a product of the integrative conjugative element Kp10 (ICEKp10), act synergistically in the intestinal colonization of CG23 hvKP [68]. In addition, virulence factors, such as moaR, mrk fimbriae, kvgAS signaling systems, and kva15 regulators, have been identified [69] and experimentally validated in mouse disease models. Although different strains have been used to identify virulence factors in different mouse models, these virulence factors constitute only a fraction of those used by hvKP to infect healthy hosts. Adjustments and adaptive changes in virulence factor expression in MDR-hvKP strains under different survival conditions require further exploration. A comprehensive understanding of MDR-hvKP virulence factors is needed to improve diagnosis and identify new antimicrobial targets.

3. Resistance mechanism of MDR-hvKP

3.1. Resistance mechanism of CR-hvKP

Carbapenemase production is essential for carbapenem resistance in CR-hvKP. According to an epidemiological analysis of CR-hvKP, the distribution of carbapenemases is characterized by certain geographical features [47]. Globally, most CR-hvKP isolates are Klebsiella pneumoniae carbapenemases (KPC)-positive, followed by OXA-positive and multiple positive strains [70]. Asia, represented by Singapore and especially China, is a major endemic area for CR-hvKP, and a report showed that more than 80% of the cases were bla_KPC-2 [28]. Although KPC is the most prevalent carbapenemase in China; NDM, OXA, and VIM have also been observed in CR-hvKP [71–74]. CR-hvKP strains which produce two or more carbapenemases have been reported from several countries and regions. For example, a CR-hvKP strain carrying three carbapenem resistance genes, blaOXA-48, and blaNDM-1/5, was reported in Italy in 2022 [75]. In Shanghai, China, a KPC-2 and NDM-5-coproducing hypermucoviscous strain named RJ-8061 was reported [76]. CR-hvKP harboring multiple carbapenemase resistance genes poses a new challenge for infection control.

The mechanisms underlying resistance to CR-hvKP include reduced or absent expression of outer membrane porins, especially ompK35 and ompK36, and activation of the efflux pump system [71,77]. Studies have shown that CRKP strains isolated from the urinary system are more capable of forming biofilms (91.07%) and producing CPSs (78.57%) [78]. However, conflicting results have been reported in the literature. Cusumano et al. [79] reported that the CRKP strains were 91% less likely to form strong biofilms. In other words, virulence may be a trade-off for bacterial survival. Therefore, the relationship between antibiotic resistance and biofilm formation remains unclear, and the underlying mechanisms require further exploration.

3.2. Resistance to polymyxin and tigecycline

Polymyxins are considered the last line of defense against metalloid carbapenemase-producing strains. Unfortunately, an hvKP resistance mechanism against polymyxins has been previously described [80]. In particular, polymyxin resistance emerges independently in non-outbreak settings, suggesting that the emergence of resistance may be a result of polymyxin B or colistin use, rather than through the clonal spread of polymyxin-resistant isolates [81]. Polymyxins have been reported to bind to lipopolysaccharides, penetrate the outer cell membrane, and ultimately cause cell death [82]. K. pneumoniae can modify its lipopolysaccharide, such as by modifying lipid A with 4-amino-4-deoxy-L-arabinose [83]. This process is controlled by a complex network involving two-component regulatory systems (CrrAB, PmrAB, and PhoPQ). In addition, in K. pneumoniae, mgrB is a negative regulator of PhoPQ; thus, any loss-of-function genetic alterations in mgrB could lead to polymyxin resistance [84,85].

A study from Taiwan revealed that tigecycline resistance was caused mainly by the overexpression of AcrAB and OqxAB with the upregulation of RamA or RarA, respectively [86]. A study of K. pneumoniae isolates from 26 provinces in China demonstrated that frameshift mutations in the TetR/AcrR family of transcriptional regulators resulted in tigecycline resistance [87]. Moreover, these strains can carry virulence plasmids without much adaptation cost, suggesting an urgent need to increase clinical awareness and epidemiological surveillance to prevent dissemination.

3.3. Resistance of MDR-hvKP to “new” drugs

With the rapid development of MDR-hvKP, new antibacterial drugs have been developed. Ceftazidime/avibactam (CAZ-AVI) is a combination of a β-lactam and a β-lactamase inhibitor and is mostly used to treat severe CRKP infections. The efficacy of CAZ-AVI, a novel antibacterial agent, has been established. Recently, CAZ-AVI, imipenem/cilastatin/relebactam, and meropenem/vaborbactam were recommended by the Infectious Diseases Society of America (IDSA) as the preferred drugs for the treatment of KPC-producing Enterobacteriaceae infections arising from the urinary tract [88]. They are characterized by their activity against ESBL, KPC and OXA-48 MDR strains but lack of activity against metallo-β-lactamases (MBLs). Currently, resistance to CAZ-AVI is still uncommon, and ST258 strains have been reported, which mostly occur in patients with severe tumors or solid organ transplantation [89]. A major reason for the resistance to CAZ-AVI is the presence of MBLs, whose activity cannot be restored by avibactam. Others include increased bla_KPC gene expression, mutations in genes encoding carbapenemases, altered cell permeability due to porin loss, and the expression of efflux pumps [90,91].

In addition, ceftolozane-tazobactam, which is a new combination of anti-pseudomonal cephalosporin and β-lactamase inhibitors [92], is also used for the treatment of MDR-hvKP. In vitro studies have shown that ceftolozane-tazobactam is effective against MDR Pseudomonas and ESBL-producing Enterobacterales but not against KPC-, OXA-48-, or MBL-producing strains [92,93]. Although K. pneumoniae is less sensitive to it than Escherichia coli, it still provides a new option for limited treatment options.

Meropenem-vaborbactam is an important serine β-lactamase inhibitor which can effectively reduce the minimal inhibitory concentration of carbapenem antibiotics [94], and the incidence of adverse events is very low [95]. It is active against ESBL and KPC MDR strains but not against OXA-48 or MBLs [96]. Therefore, meropenem/vaborbactam is a good candidate for the treatment of ESBL- and KPC-producing K. pneumoniae, especially for CAZ-AVI-resistant strains.

Rapid advances in artificial intelligence (AI) have provided new hope for antibiotic researchers and clinicians. For example, the new antibiotic halicin, originally developed as a treatment for diabetes, was predicted by AI models to be bactericidal against a wide range of bacteria, including Mycobacterium tuberculosis and carbapenem-resistant Enterobacteriaceae, and successfully cured drug-resistant bacterial infections in a mouse infection model [97]. This opens up new ideas for antibiotic research and increases the speed of the discovery of novel drug candidates. In the future, combining CRISPR and automated platforms is expected to enable AI in the entire design, synthesis, and testing process. AI also plays a unique role in antigen identification, which can aid vaccine development [98]. Although the use of artificial intelligence has opened a new window for anti-infective research, testing the efficacy and safety of these molecules in the laboratory remains a necessary step. The integration of genomic and clinical data may help improve the accuracy of predictions, while high-quality raw data and sound interpretation of the results of the analyses are important.

4. Pathways of MDR-hvKP formation

4.1. Evolution of MDR-hvKP

MDR-hvKP strains emerged in the 2010s and thereafter became increasingly prevalent [99]. Although reported in Chile [100], Germany [101], Switzerland [102], Italy [103], and other countries, approximately 50% of carbapenem-resistant strains in China carry virulence genes, which is largely related to the prevalence of the ST11 strain [10,73,104,105]. Despite its ubiquity in the environment, the reservoir seeding human infections with K. pneumoniae is often the patient’s gut [106]. In China, the hvKP carriage rate among healthy individuals ranges from 4% to 5.19% [107]. Osama et al. reported that intestinal commensal hvKPs isolated from healthy individuals carry a variety of virulence and drug resistance genes, indicating a potential risk of severe invasive disease [108]. This finding highlights the importance of actively monitoring gut-colonizing hvKPs in the population, especially for colonization by novel high-risk MDR-hvKP strains.

MDR-hvKP can be produced in numerous ways, including the acquisition of resistance genes by highly virulent strains, acquisition of highly virulent genes by resistant strains, and transformation of nonresistant non-highly virulent strains [109]. For a long time, hvKP strains have rarely developed resistance to commonly used antimicrobials, which may be related to some limitations in gene acquisition, such as the CRISPR system [110]. Significantly thickened CPSs limit horizontal gene transfer, hvKP rarely undergoes chromosomal recombination, and reduced plasmid diversity leads to reduced pangenomic diversity [111]. These factors lead to much lower clonal diversity in hvKP than in multidrug-resistant strains, and the efficiency of MDR clones in obtaining virulence plasmids far exceeds that of hvKP clones in obtaining resistance plasmids [32,112]. However, antibiotic-resistant hvKP isolates have been reported worldwide. The MDR-hvKP formation patterns are shown in Figure 3. Elucidating the underlying mechanisms which promote the spread of antimicrobial resistance and virulence elements is important for the prevention and control of MDR-hvKP, which are currently poorly understood.

Figure 3.

Patterns of multidrug-resistant hypervirulent Klebsiella pneumoniae (mdr-hvKP) formation. The process of mdr-hvKP formation can be categorized into three modes: mge-mediated, intermediate vector-mediated horizontal gene transfer and the insertion of nonsense fragments of genes and stimulus-induced mutations (created with BioRender.com).

4.2. Mobile genetic element (MGE)-mediated MDR-hvKP formation

MGEs, such as plasmids and insertion sequence elements (ISs), play important roles in the acquisition of pathogenicity and spread of drug resistance genes. The major conjugative plasmids carrying resistance genes for horizontal gene transfer (HGT) include hybrid incompatibility group F (IncF), IncX, and IncN, whereas Incfik and IncHI1B are the major conjugative plasmids carrying virulence genes. Plasmid migration leads to the development of MDR-hvKP [113,114]. A hybrid plasmid formed by integrating a hypervirulent plasmid fragment into a transferable plasmid can be transferred into CRKP strains via conjugation, enabling them to simultaneously express carbapenem resistance- and hypervirulence-associated phenotypes [70,115]. Plasmids lacking a conjugation mechanism can be mobilized by other mobile elements or can form hybrid plasmids via intermolecular translocation or homologous recombination [116,117]. Wang et al [118]. reported that non-conjugative plasmids could be mobilized by IncN3 conjugative plasmids by fusion with IncN3 into a hybrid plasmid or by co-transfer with IncN3. Acquisition of virulence plasmids by MDR or extremely drug-resistant cKP strains has been previously reported [116,119]. Hypervirulent strains can acquire integrative conjugative elements or conjugative resistance plasmids containing determinants of antimicrobial resistance and integrate them into chromosomes or virulence plasmids, benefitting from genome plasticity and openness [72,120,121]. The ureC:IS5075 insertion is present in 14.1% of the 9,755 K. pneumoniae genomes, because these transposons often carry antibiotic resistance genes, which may explain the frequent occurrence of urease loss in fluoroquinolone-resistant and carbapenemase-carrying isolates [72]. Researchers have shown that cKP becomes a highly virulent drug-resistant strain by obtaining a fusion plasmid carrying virulence and carbapenem resistance genes [122], and the existence of this hybrid plasmid is undoubtedly a great risk.

Phages also play an important role in HGT. Phages act as drug-resistance gene repositories. However, only 1.2% of K. pneumoniae strains contain virulence factor genes from phages [123]. Because virulence genes are species-specific, resistance genes are highly conserved across species and are located near transposases and/or integrases.

4.3. Intermediate vector-mediated MDR-hvKP formation

Almost all gram-negative bacteria can produce and releasing outer membrane vesicles (OMVs) containing large amounts of lipids, proteins, and genetic material, which can be transferred horizontally to recipient bacteria or host cells and play a role in intercellular communication and modulation of the host immune response [124]. OMVs exert a protective effect on luminal genes or plasmids and are effective horizontal gene transfer systems [125]. Recent studies have shown that hvKP-OMVs can carry key virulence genes, resulting in increased viscosity and virulence expression levels in resistant strains which have acquired virulence determinants [126]; this is another mechanism for the horizontal transfer of virulence factors between strains. Another intermediate vector is E. coli because an excess capsule of hvKP is not conducive to the transfer of virulence plasmids to CRKP, and the use of E. coli strains as intermediate carriers can eliminate this restriction. E. coli carrying the conjugated IncF plasmid indirectly delivers the virulence plasmid from hyperviscous hvKP to CRKP, resulting in a hyperviscous phenotype [116].

4.4. Nonsense fragment inserts and gene mutations

Although rarely reported, damage to or mutations in chromosomal genes can result in the development of drug-resistant strains. Mutations are the most common molecular mechanism underlying colistin resistance and are caused by either nonsense mutations or insertion sequences [127]. mgrB is associated with colistin resistance, and its disruption leads to the emergence of hvKP, which is simultaneously colistin and carbapenem resistant [119]. In one study analyzing two strains of K. pneumoniae isolated from a patient with a severe infection, whole-genome sequencing revealed that high-virulence-related genes such as rmpA/A2, iroBCDN, peg, and iucABCD/iutA were successfully inserted into specific regions of the bacterial chromosome [128]. This finding suggests that hvKP can transfer genes horizontally and spread vertically to evolve continuously.

5. Identification of hypervirulence in MDR-hvKP

To date, multiple virulence factors and phenotypes have been well characterized, including clinical manifestations, sequence and capsule typing, HMV phenotype, and the presence of virulence-associated genes, all of which can be used to differentiate hypervirulent strains from cKPs [129–131]. Overall, confirmation of hvKP strains involves two methods: phenotypic and genetic testing. The details are presented in Table 1. The methods used to determine the hypervirulence of MDR-hvKP are essentially identical to those used to determine that of hvKP.

Table 1.

Experiments to identify hypervirulence of hypervirulent Klebsiella pneumoniae (mdr-hvKP).

	Test name	Scope of application	References
Phenotype Detection	Animal experimentation	Gold standard; Research recommendation	[132,133]
	Serum resistance assay	Research recommendation	[134]
	Neutrophil phagocytosis and bactericidal activity	Research recommendation	[135]
	String test	Need for joint accreditation	[136]
	capsule typing	Need for joint accreditation	[137]
	Siderophores testing	Research recommendation	[138]
Genetic testing	Gene expression detection	Need for joint accreditation	[104,138,139]
	NGS	Both samples and strains can be tested, with simultaneous subspecies identification and virulence gene detection.	[140,141]
	MALDI-TOF MS	High specificity, rapid differentiation of K1 hypervirulent K. pneumoniae	[142]

NGS: Next-generation sequencing.

Matrix-assisted laser desorption ionization-time-of-flight mass spectrometry: MALDI-TOF MS.

Animal trials are the gold standard for confirming the presence of hvKP strains. Animal models which can be used include Galleria mellonella, mice, and zebrafish. The Galleria mellonella infection model and the mouse lethal dose 50 test are most commonly used to verify strain virulence [132,133,143]. However, this method is difficult to use in a clinical microbiology laboratory. In vitro validation of virulence includes neutrophil phagocytosis, bactericidal activity tests, and serum resistance assays, which are relatively simple to perform and have short test cycles [134,135]. In previous studies, a positive string test was often used as a criterion for identifying hvKP. Studies have shown that the agreement between string tests and clinical feature identification ranges from 51% [144] to 98% [44]. The string test can identify HMV, but a positive result does not indicate high virulence, which causes some confusion. Capsule typing helps identify highly virulent strains and consists mainly of immunological methods, that is, serological typing based on the K antigen and wzi typing. Siderophores are important indicators of cKPs and hvKPs [131]. The iron carrier assay includes semi-stereotypical experiments to observe the morphology of colonies using chromium-containing azure S dye, and quantitative analysis to calculate the iron carrier content by preparing standards with different concentrations of iron carriers [145].

Confirmation of the hvKP-specific gene sequence includes hvKP-related virulence gene detection, hvKP high-throughput sequencing, and mass spectrometry. Based on its practical clinical importance, optimal categorization is a key factor for a highly virulent phenotype. Studies have shown that the genes peg-344, iroB, and iucA; plasmid-carried rmpA; and siderophore production greater than 30 µg/mL can accurately identify hvKP with a diagnostic accuracy > 95% and correlate well with the severity of infection and the risk of death [138]. This approach is regarded as the gold standard for identifying hvKP virulence. Yu et al. developed a rapid multiplex PCR assay capable of analyzing common sequences and capsular types of cKP and hvKP, and virulence genes [104]. The whole genome of the isolated strain or the original sample is directly detected via next-generation sequencing, and the pLVPK-like or pK2044-like virulence plasmid is subsequently identified by bioinformatics; the pod synthesis locus (K locus) can also be identified for rapid confirmation of hvKP [140]. Matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) can identify K1 hvKP with high specificity, and in combination with other assays, it can provide rapid and early diagnosis of hvKP [142].

Additionally, new detection techniques have been developed for hvKP identification. Studies have developed convolutional neural networks which use raw Raman spectrum data to quickly identify antimicrobial resistance genes, high virulence-associated factors, and resistance phenotypes of K. pneumoniae and design highly accurate and reasonable treatment plans for bacterial infections within a few hours [146]. The peg-344 loop-mediated isothermal amplification technique established by Liao et al. also shows good application prospects for the rapid diagnosis of hvKP [139]. Overall, these new techniques complement the current rapid identification methods for MDR-hvKP.

6. Conclusion

The convergence of drug resistance and hypervirulence has made clinical treatment of hvKP infections extremely difficult, and MDR-hvKP has become an emerging “superbug” of great clinical interest. Although MDR-hvKP strains have been reported globally, a large knowledge gap remains. As shown in Figures 1–3 and Table 1 in the text, we compiled the virulence factors and resistance mechanisms of MDR-hvKP, discussed how it has developed in the context of recent research advances, and finally summarized current methods for identifying highly virulent strains. We hope that this summary will help increase the awareness of researchers, clinical microbiologists, and medical workers regarding MDR-hvKP strains and emphasize the rational use of antibiotics.

7. Future perspective

MDR-hvKP strains are highly pathogenic and multidrug resistant, leaving clinicians with limited options. In the current context of increased antibiotic pressure, horizontal transfer of virulence and resistance components between strains is inevitable. Based on clinical needs, such as preventing the development of invasive infections, improving patient prognosis, and promoting the rational use of antimicrobials, appropriate assays or combinations should be selected to rapidly identify MDR-hvKP infections. Clinical microbiologists should clarify its definition and develop appropriate identification criteria as soon as possible, which will also help avoid confusion in the research process. These findings suggest that we should continue to emphasize the surveillance and detection of MDR and hvKP strains, as well as the development of novel antimicrobial drugs in the coming decades. In addition, as the research progresses, we expect to identify more molecules for detection.

Funding Statement

This project was supported by the Science and technology project of Zhangjiagang (ZKS2041).

Article highlights

• Multidrug-resistant hypervirulent Klebsiella pneumoniae (MDR-hvKP) combines high pathogenicity and multidrug resistance to become a new superbug.

• The main virulence factors of MDR-hvKP include capsular polysaccharides, hypermucoviscosity phenotypes, iron acquisition systems, virulence plasmids, biofilm-forming capacity, and type VI secretion system.

• Mechanisms of resistance to MDR-hvKP include resistance to carbapenemases, polymyxin and tigecycline, and “new” drugs.

• Three mechanisms are involved in MDR-hvKP formation: mobile genetic element (MGE)-mediated, intermediate vector-mediated, and nonsense fragment insertion and mutation.

• Overall, confirmation of hvKP strains involves two methods: phenotypic and genetic testing.

• Given their destructive and transmissible potential, continued surveillance of these organisms and enhanced control measures should be prioritized.

• Based on practical clinical needs, such as preventing the occurrence of invasive infections, improving patient prognosis, and promoting the rational use of antibiotics, appropriate assays or combinations should be selected to rapidly identify MDR-hvKP strains. This will also help to avoid confusion during the research process.

• As research progresses, we expect to identify new and more identifiable molecules for detection.

Authorship contribution

Yu Zheng: writing – original draft, investigation, conceptualization; Xiaojue Zhu: writing – original draft; Chao Ding: writing – original draft and review & editing; Weiqiang Chu, Xiaoxiao Pang, Ruxia Zhang and Jiucheng Ma: writing – review & editing; Guoxin Xu: conceptualization.

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.