A gene cassette, a small MGE, consists of multiple genes and recombination sites and can be captured by integrons (Partridge et al., 2009; Xu et al., 2018). Currently, the genes within gene cassettes are primarily associated with drug resistance traits. Integrons serve as assembly platforms that capture gene cassettes and incorporate exogenous genes through site-specific recombination, allowing for the internalization and expression of these foreign genes (Hall and Collis, 1995; Mazel, 2006; Escudero et al., 2015). Although integrases cannot excise it from the chromosome, rendering the integron itself immobile, they can depend on recombinases or transposases for genome-wide migration.

Transposon elements, including insertion sequences (ISs) and transposons (Tns), are MGEs that mediate intracellular gene transfer (Partridge et al., 2018). ISs and Tns are discrete, mobile DNA segments that can relocate within the genome through a series of processes, including excision and reintegration.

Phages can be classified into two types, virulent and temperate phages, based on their distinct life cycles within the host (Casjens, 2003). Virulent phages follow a classical cycle consisting of five stages: adsorption, invasion, proliferation, assembly, and lysis, ultimately leading to the release of progeny phages. In contrast, temperate phages integrate their genetic material into the host genome and can adopt a lysogenic state. Under specific inducing conditions, temperate phages can enter the lytic phase, resulting in the lysis of the host bacterium. Prophages are the nucleic acids of temperate phages that integrate into the host genome at specific sites, allowing them to replicate alongside the host chromosomes (Canchaya et al., 2003). HGT occurs through the imprecise excision of the prophage, which can carry host genes, leading to cell lysis under induced conditions and subsequent infection of a new host. Prophage-mediated HGT is facilitated by transduction, which includes specific transduction, generalized transduction, and lateral transduction (as described in the section on transduction below). Multiple prophages have been identified on the chromosomes and plasmids of K. pneumoniae, exhibiting diversity and widespread dissemination (Gabashvili et al., 2020; Shen et al., 2020; Kang et al., 2023). Notably, a prophage containing the blaKPC gene has been discovered in the chromosomes of KPC-producing K. pneumoniae strains (Chen et al., 2015).

K. pneumoniae, which mediates the transfer of resistance genes, is primarily associated with various classes of antibiotics to which it develops resistance. These include tetracyclines (tet), sulfonamides (sul), β-lactams (bla), aminoglycosides (AAC), mucormycetes (MCR), quinolones (QNR), and multidrug-resistant bacteria (MDR). Information regarding these plasmids, as collected from the literature, is presented in Table 1. This table was utilized to create plasmid maps that illustrate the relevant resistance gene information in K. pneumoniae. Furthermore, plasmids also facilitate the transfer of genes that encode functions enabling recipient cells to better adapt and survive in their environments (Bellanger et al., 2014). To better understand the contribution of plasmids in HGT, it is important to distinguish between transmissible and non-transmissible plasmids. Transmissible plasmids include conjugative plasmids and mobilizable plasmids (Fang and Zhou, 2020). Specifically, conjugative plasmids carry all the essential genes for self-transmission through conjugation, including a Type IV secretion system (T4SS), a T4SS coupling protein (T4CP), relaxasome accessory factors (RAFs), and a relaxase gene (Ramsay and Firth, 2017). Essential genes are shown on the K. pneumoniae plasmid map in the following description.

Integrative and conjugative elements (ICEs) are prevalent mobile units carrying modules responsible for excision, maintenance, conjugative transfer, and integration into the genome of a new host (Johnson and Grossman, 2015). ICEs are plasmid-like MGEs that intercept chromosomal genes from the host, which are transferred to recipient cells via conjugation, and subsequently integrate into the recipient chromosome (Thomas and Nielsen, 2005). A key distinction between ICEs and plasmids is that ICEs typically do not replicate independently; rather, they integrate into the chromosome and replicate in conjunction with it. ICE gene clusters are classified into four classical modules based on their functions: integration and excision (int-xis), mating-pair formation (mpf), mobilization and processing (mob), and regulatory (reg). Additionally, ICEs contain hot spots (HS) and variable regions (VR) that may include exogenous genes. The processes of ICE segregation and integration are associated with the genes for dissociase (xis) and integrase (int), respectively (Hirano et al., 2011). The widespread presence of the yersiniabactin gene cluster and its associated integrative conjugative elements (ICEKp) has been documented among clinical isolates of K. pneumoniae (Lam et al., 2018; Octavia et al., 2019; Jati et al., 2023). ICEKp1, an integrative and conjugative element identified in K. pneumoniae, plays a significant role in the transmission of the high-pathogenicity island (HPI) and contributes to the genomic heterogeneity associated with K. pneumoniae pyogenic liver abscess (PLA) infections (Lin et al., 2008). ICEKp1 can be divided into three distinct regions: one region resembles the high-pathogenicity island of Yersinia species and encodes the siderophore yersiniabactin; the second region resembles a portion of the large virulence plasmid pLVPK and encodes the glycosidophile salmochelin (iroBCDN) and the capsular polysaccharide regulator RmpA; the third region encodes the type IV secretion system (T4SS) (Shen et al., 2019). Due to their critical role in the transfer of carbapenemase genes (Botelho et al., 2020), ICEs pose a significant threat to human health by potentially facilitating the emergence of hypervirulent, drug-resistant K. pneumoniae.

Integrative and mobilizable elements (IMEs) exhibit structural similarities to integrative conjugative elements (ICEs), paralleling the relationship between conjugative plasmids and mobilizable plasmids. Both IMEs and ICEs are classified as mobile genetic elements (MGEs) that possess their own excision and integration modules. IMEs can utilize the mating pair formation (Mpf) machinery of conjugative elements, whether plasmids or ICEs, to facilitate their transfer between bacterial cells (Guédon et al., 2017). Importantly, the IME should not be regarded as a disabled component of the ICE; rather, it functions as a hitchhiking passenger.

Natural bacterial transformation refers to the process by which a recipient bacterium directly assimilates a specific DNA fragment containing genes from a donor bacterium present in the environment. The blaIMP-68 gene, identified in K. pneumoniae strain TA6363 isolated in Tokyo, was inserted into the pHSG398 vector plasmid and subsequently used to transform E. coli DH5α cells. The minimum inhibitory concentration (MIC) of carbapenems was observed to increase in the recipient strain (Kubota et al., 2019). This finding suggests that when K. pneumoniae undergoes fragmentation, plasmids harboring resistance genes can escape and be transformed under specific conditions, facilitating the transfer of resistance genes.

This process can be delineated into four distinct sequential stages: competence induction, DNA uptake, translocation of DNA across the inner membrane (IM) and outer membrane (OM), and the integration or recombination of the incoming DNA into the genome or plasmid (Lorenz and Wackernagel, 1994; Averhoff et al., 2021). Research has been conducted to construct a comprehensive transformation mechanism by analyzing specific transformation processes in various bacterial species (Johnston et al., 2014). In a study by Averhoff et al. (2021), preliminary transformation models were developed for the systems of Acinetobacter and Thermus thermophilus. The transformation process can be summarized as follows: a bacterial pseudopilus penetrates a designated channel in the outer membrane and carries a DNA receptor at its tip, which enables recognition and binding of DNA from the external environment. Subsequently, the pseudopilus translocates the DNA through the secretin channel into the periplasm, where it is transported to the inner membrane by binding to a periplasmic protein. At this juncture, the DNA associates with an inner membrane protein that functions as a channel protein, facilitating the transfer of DNA into the cytoplasm. It is noteworthy that there are variations in the functional proteins involved in transformation systems across different species. In the extensively studied Acinetobacter, the functional proteins implicated in the transformation process include the fimbrillin proteins ComP, ComB, ComE, ComF, PilV, PilX, and FimT. The pili tip proteins are represented by ComC, while the secretin channel proteins consist of the multimeric secretin subunit known as PilQ. Furthermore, Acinetobacter employs the periplasmic binding protein ComEA and the endosomal protein ComA.

Transduction is a biological process whereby a bacteriophage transfers genetic material from a host bacterium to another bacterium during the infection cycle. K. pneumoniae exhibits a high degree of lysogenicity and harbors prophages, which are likely associated with phage transduction (Wang et al., 2019; Kang et al., 2023). This genetic transfer involves the integration of the acquired genetic material into the genome of the recipient bacterium. The incorporation of the phage genome into the bacterial chromosome is referred to as lysogeny. Prior research has elucidated the mechanisms by which bacteriophages facilitate HGT (Borodovich et al., 2022). This overview aims to summarize key aspects of phage propagation, which are essential for understanding the transduction process. During replication, the phage genome undergoes replication and translation to produce structural and functional proteins, leading to the formation of concatemers. These concatemers are subsequently introduced into an empty viral capsid, where the terminase complex excises a copy for packaging. The filled capsid then assembles with the tail to form a progeny phage, while the terminase complex binds to the next empty capsid to continue the packaging process (Rao and Feiss, 2015). Transduction can be categorized into three distinct types based on the mechanisms involved: (i) specialized transduction, which occurs in temperate phages when the prophage integrated into the bacterial chromosome undergoes imprecise excision, resulting in the replication and packaging of adjacent host chromosomal genes into the phage. This process is restricted to genes located near the prophage and may lead to the loss of phage genes; (ii) generalized transduction, primarily observed in pac-type phages, involves the terminase complex recognizing a homologous region at the pac site within the host genome, excising a gene fragment of phage genome size, and packaging it into the phage; and (iii) lateral transduction, where the phage genome is replicated within the host chromosome without excision. In this case, the terminase complex acts on the pac site located in the middle of the phage genome, packaging the replicated DNA into daughter phages that contain both host and phage genes, resulting in multiple phages that incorporate segments of the host genome.

Conjugation represents a significant mechanism of HGT that necessitates direct contact between bacterial cells. As previously noted, in K. pneumoniae, MGEs such as conjugative plasmids and conjugation elements (ICEs) can be exchanged between bacterial cells through a cell-to-cell conjugation system (Navon-Venezia et al., 2017). This process is facilitated by various modules encoded by the plasmid, which include genes responsible for the type IV secretion system (T4SS), the T4SS coupling protein (T4CP), relaxosome accessory factors (RAF), and relaxase genes. Further details regarding the F-plasmid conjugation process and the roles of the associated molecules have been comprehensively documented by Virolle et al. (2020).

The initial phase of conjugation involves the expression of the tra gene, which is situated within the plasmid transfer region. This gene is responsible for the synthesis of all proteins associated with the conjugative pilus and the T4SS required for the establishment of the mating pair. Additionally, the tra gene produces components of relaxase that facilitate the processing of the plasmid before its transfer. In the subsequent phase, conjugative pilus is generated, which senses the extracellular environment and identifies recipient cells, thereby facilitating their aggregation with donor cells to form the mating pair (Curtiss, 1969; Folkhard et al., 1979; Willetts and Skurray, 1980; Clarke et al., 2008). Notably, it has been observed that the pilus may also serve as a conduit for the direct transfer of single-stranded DNA (Brinton, 1965). Following the establishment of the mating pair, the relaxosome—a complex comprising relaxases and auxiliary proteins—initiates the processing of the plasmid, which includes relaxosome cleavage and the isolation of T-DNA (Lanka and Wilkins, 1995; Dostál and Schildbach, 2010). The relaxosome binds to and cleaves a specific site known as the nik site located in the origin region of the plasmid, resulting in the relaxation of the double-stranded DNA. The processing of the plasmid entails the cleavage by the relaxosome and the subsequent isolation of the T-DNA (Cohen et al., 1968; Willetts and Skurray, 1980). The T-DNA subsequently forms a covalent bond with the relaxase, which, in conjunction with the Type IV coupling protein (T4CP) anchored to the cell membrane, facilitates its recruitment to the conjugative pore (Curtiss, 1969; de la Cruz et al., 2010). Thereafter, the T-DNA and relaxase are transferred to the recipient cell through the conjugation pore, with the leading sequence of the plasmid being the first to be transferred to the recipient (Frost et al., 1994). This concludes the description of plasmid transfer.

In the donor cell, the single-stranded cyclic plasmid is restored to its original structure through the process of rolling circle replication (RCR) (Waters and Guiney, 1993). In contrast, upon transfer to the recipient cell, the T-DNA initially forms a cyclized single-stranded DNA via the ligation of both ends of the oriT, a process facilitated by relaxases (Draper et al., 2005; Chandler et al., 2013). Subsequently, this cyclized single-stranded DNA is converted into double-stranded DNA through the coordinated action of the host’s DNA and RNA polymerases (Baharoglu and Mazel, 2014). In response to the incoming T-DNA, the host activates various defense mechanisms, perceiving it as exogenous DNA. These defense mechanisms encompass restriction modifications, nucleic acid exonucleases, recombinant systems, and adaptive immunity (Casjens, 2003). Importantly, the host’s single-stranded binding protein (SSB) plays a crucial role in protecting the T-DNA from nuclease degradation by enveloping it (Virolle et al., 2020). Additionally, the plasmid may encode PsiB proteins that inhibit SOS induction, thereby providing further protection for the plasmid (Bailone et al., 1988).

Extracellular vesicles (EVs), which are generated through cellular metabolic activities, are spherical nanoparticles that encapsulate nucleic acids, proteins, lipids, and various other biomolecules (Kalluri and LeBleu, 2020). These vesicles play a crucial role in evading the immune response, transmitting virulence factors, facilitating intercellular communication, mediating HGT, facilitating nutrient and electron transfer, and promoting biofilm formation. Bacteria secrete a diverse array of extracellular vesicles, including outer membrane vesicles (OMVs), outer-inner membrane vesicles (OIMVs), and explosive outer membrane vesicles (EOMVs) (Toyofuku et al., 2019).

OMVs are nanoscale proteoliposomes characterized by a single membrane and are secreted by bacteria. These vesicles predominantly consist of components derived from the bacterial outer membrane and the periplasmic space (Amano et al., 2010; Ellis and Kuehn, 2010; Toyofuku et al., 2023). OMVs facilitate the differential intracellular release of various toxins and virulence factors, including adhesins, invasins, outer membrane proteins, lipopolysaccharides (LPS), flagellin, and proteases. Additionally, they may encapsulate cytoplasmic proteins and DNA (Kulp and Kuehn, 2010). Recently, a novel class of double-membrane bilayer EVs, known as outer-inner membrane vesicles (OIMVs), has been identified in Gram-negative bacteria (Pérez-Cruz et al., 2015). OIMVs not only contain the traditional components found in OMVs but also incorporate cytoplasmic elements, including DNA and plasmids. Consequently, OIMVs have been proposed as a significant type of EV involved in DNA (Toyofuku et al., 2019). The mechanisms underlying the production of EVs and the packaging of their contents remain inadequately understood; however, several hypotheses have been proposed regarding the presence of genetic material within these vesicles (Pérez-Cruz et al., 2013). The first hypothesis suggests that extracellular free DNA is internalized into the vesicles via a mechanism similar to bacterial transformation (Renelli et al., 2004). The second hypothesis posits that DNA is transported across the inner membrane and cell wall into the extraplasmic space, where it is subsequently encapsulated within the OMV (Kadurugamuwa and Beveridge, 1995). The third hypothesis, which has gained widespread acceptance, proposes that the formation of OIMVs involves the rolling of DNA from the cytoplasm of the bacterial cell (Pérez-Cruz et al., 2015). Thus, EVs, encompassing both OMVs and OIMVs, have the potential to facilitate horizontal gene transfer.

EVs have emerged as a novel mechanism of HGT in various bacterial species, functioning as carriers for the transfer of genetic material (Dorward et al., 1989; Yaron et al., 2000; Rumbo et al., 2011). These vesicles can adhere to the outer membrane of recipient cells, thereby facilitating the transmission of genetic elements between different bacterial species, which may result in the acquisition of new pathogenic traits or drug resistance. The uptake pathways for extracellular vesicles into host cells may involve several mechanisms, including macropinocytosis, clathrin-mediated endocytosis, caveolin-mediated endocytosis, or non-caveolin, non-clathrin-mediated endocytosis (O’Donoghue and Krachler, 2016). Research on K. pneumoniae has demonstrated that outer membrane vesicles (OMVs) play a critical role in mediating the transfer of drug-resistant or virulence plasmids both within and between bacterial species. Dell’Annunziata et al. (2021) was the first to report the transfer of plasmids containing resistance genes via OMVs derived from K. pneumoniae. Furthermore, research conducted by Wang et al. (2022) indicated that OMVs facilitate the transfer of the virulence plasmid phvK2115, which harbors various virulence genes, including rmpAp, rmpA2p, iucA, iroB, and peg344, within and among HVKP species. Additionally, it has been demonstrated that OMVs can disseminate two CRK3022 resistance plasmids, which contain genes such as blaKPC-2 and blaCTX-M-1, among K. pneumoniae and Escherichia coli strains within the Enterobacteriaceae family (Tang et al., 2023). Notably, all of the aforementioned plasmids possess conjugation transfer regions, yet it remains unclear whether these regions play a corresponding role in the fusion of the outer vesicle with the recipient cell. OMVs isolated from K. pneumoniae NUHL30457 (K2, ST86) have also been shown to transfer both resistance and virulence genes, resulting in a phenotype characterized by increased drug resistance and virulence (Li P. et al., 2022). Moreover, the plasmid IncFIBpKPHS1, which carries blaNDM-1 within OMVs, is also transmitted among K. pneumoniae strains (Tang et al., 2023). It is important to note that the aforementioned studies must exclude the possibility of free plasmids entering K. pneumoniae through transformation during the experimental procedures. Furthermore, the specific type of extracellular vesicles involved has not been delineated; it remains uncertain whether they are single-membrane OMVs or double-membrane outer-inner membrane vesicles (OIMVs). Therefore, future investigations should aim to differentiate between these vesicle types to enhance our understanding of the mechanisms by which genetic material enters outer membrane vesicles and to ascertain which type of outer membrane vesicles predominantly mediates gene transfer.

EVs can mediate the transmission of drug-resistance genes and virulence genes between K. pneumoniae, presenting additional challenges for clinical treatment. However, EVs offer several advantages over traditional plasmid-mediated gene transfer (Bonnington and Kuehn, 2014). Firstly, EVs do not require direct cell-to-cell contact for long-distance transport, significantly enhancing transfer efficiency. Secondly, they protect the molecular biological components contained within them from external environmental factors, thereby improving the stability of these components. For instance, EVs safeguard DNA from degradation by extracellular nucleases (Pérez-Cruz et al., 2015).