The Characteristic of Virulence, Biofilm and Antibiotic Resistance of Klebsiella pneumoniae

Klebsiella pneumoniae is an important gram-negative opportunistic pathogen that causes a variety of infectious diseases, including urinary tract infections, bacteremia, pneumonia, and liver abscesses. With the emergence of multidrug-resistant (MDR) and hypervirulent K. pneumoniae (hvKP) strains, the rapid spread of these clinical strains in geography is particularly worrying. However, the detailed mechanisms of virulence and antibiotic resistance in K. pneumoniae are still not very clear. Therefore, studying and elucidating the pathogenic mechanisms and drug resistance mechanism of K. pneumoniae infection are important parts of current medical research. In this paper, we systematically summarized the virulence, biofilm, and antibiotic tolerance mechanisms of K. pneumoniae, and explored the application of whole genome sequencing and global proteomics, which will provide new clues for clinical treatment of K. pneumoniae.


Introduction
Klebsiella pneumoniae is a class of gram-negative bacterium that is ubiquitously found on the surface of mucosa in animals, or in the environment (such as water, soil, etc.). In humans, K. pneumoniae is concentrated in the gastrointestinal tract, and a few in the nasopharynx, through which the bacteria can enter the blood circulation or other tissues, and then cause infection. In the era of pre-antibiotics, K. pneumoniae was a vital pathogen of community-acquired pneumonia (CAP), especially in diabetics and alcoholics. In the era of antibiotics that followed, it became a major cause of medical-related infections in hospitals [1], and a risk factor of severe community-acquired infections [2]. In Singapore, mortality rates of K. pneumoniae bacteremia ranged from 20-26% [3]. In China, K. pneumoniae accounted for 11.9% of isolated pathogens from ventilator-associated pneumonia (VAP) and intensive care unit (ICU)-acquired pneumonia [4]. In addition, carbapenem-resistant Enterobacteriaceae (CRE) caused by K. pneumoniae have been reported to account for 73.9% of 664 clinical samples in a multi-center clinical study that covered 25 "AAA" hospitals in 14 provinces of China [5]. There is no doubt that such a high prevalence and mortality rate of K. pneumoniae infection caused a great burden on the country's health system.
K. pneumoniae has many accessory genomes of plasmids and chromosome gene loci. According to the accessory genome, K. pneumoniae strains are divided into three types: opportunistic, hypervirulent, and multidrug-resistant (MDR) [6]. Recently, most infections of K. pneumoniae were caused by the classic K. pneumoniae (cKp) strains which remained living in the hospitals, and then caused infection in weak patients. The cKp strain appears to be different from the hvKp. Genetic factors of high virulence phenotype of hvKp are present on a large virulent plasmid, and there may be integrative conjugal In addition, some special virulence factors of K. pneumoniae strains have been considered in recent years. Pks gene clusters can be detected in bloodstream infections (BSIs) caused by pks positive (pks+) strains. Colibactin is a toxic substance encoded by pks gene clusters, which participates in the process of host DNA damage and strains virulence enhancement [21]. By monitoring the prevalence of K. pneumoniae-infected bacterial meningitis, Xu et al. found that the high prevalence and mortality of hvKP producing kpc-2 gene meningitis should be paid attention [22]. Some "high risk" sequence types of K. pneumoniae, such as STs11, 15, and 383 clones, showed a combination of drug resistance and virulence factor. In these strains, many elements are carried on a large virulent plasmid (such as ST147 with blaNDM-1), which usually encodes iron vector genes (such as aerobactin, salmochelin, and enterochelin), heavy metal resistance genes (encoding resistance for tellurite, silver, copper, and lead), and capsule up-regulation genes, rmpA and rmpA2. The emergence of this combination of resistance and virulence is worrying [23]. It is necessary to conduct in-depth study and epidemic monitoring of some highly virulent strains. Although the clinical features of infection by hvKp strains have been fully elucidated, there are only a few reliable biological indicators to distinguish these strains from other K. pneumoniae strains. Aerobactin plays an important role in iron carrying, growth, and virulence of K. pneumoniae, so it is considered to be the key virulence factor of hvKP. The high expression level of aerobactin in strain is a significant feature of the hvKP strain. Li et al. revealed that string test combined with aerobactin and rmpA index is helpful to improve the detection rate of hvKP [16]. We systematically summarized the virulence, biofilm and antibotic resistance related genes, and their functions of K. pneumoniae are systematically summarize in Table 1. In addition, some special virulence factors of K. pneumoniae strains have been considered in recent years. Pks gene clusters can be detected in bloodstream infections (BSIs) caused by pks positive (pks+) strains. Colibactin is a toxic substance encoded by pks gene clusters, which participates in the process of host DNA damage and strains virulence enhancement [21]. By monitoring the prevalence of K. pneumoniae-infected bacterial meningitis, Xu et al. found that the high prevalence and mortality of hvKP producing kpc-2 gene meningitis should be paid attention [22]. Some "high risk" sequence types of K. pneumoniae, such as STs11, 15, and 383 clones, showed a combination of drug resistance and virulence factor. In these strains, many elements are carried on a large virulent plasmid (such as ST147 with blaNDM-1), which usually encodes iron vector genes (such as aerobactin, salmochelin, and enterochelin), heavy metal resistance genes (encoding resistance for tellurite, silver, copper, and lead), and capsule up-regulation genes, rmpA and rmpA2. The emergence of this combination of resistance and virulence is worrying [23]. It is necessary to conduct in-depth study and epidemic monitoring of some highly virulent strains. Although the clinical features of infection by hvKp strains have been fully elucidated, there are only a few reliable biological indicators to distinguish these strains from other K. pneumoniae strains. Aerobactin plays an important role in iron carrying, growth, and virulence of K. pneumoniae, so it is considered to be the key virulence factor of hvKP. The high expression level of aerobactin in strain is a significant feature of the hvKP strain. Li et al. revealed that string test combined with aerobactin and rmpA index is helpful to improve the detection rate of hvKP [16]. We systematically summarized the virulence, biofilm and antibotic resistance related genes, and their functions of K. pneumoniae are systematically summarize in Table 1. Efflux pump systems; significant resistance to tobramycin and spectinomycin [29] KpnO Directly involved in aminoglycoside resistance; resistance of tobramycin, streptomycin and spectinomycin [30] rrs or rpsL rpsL mutations associated with high fitness costs and reduced virulence [31] Quinolone resistance DNA gyrase (gyrA-gyrB subunit) Topoisomerase IV (parC-parE subunit) Resistance of nalidixic acid and ofloxacin [32] OmpK36 Cell permeability [33] acrAB Cell permeability [34] kdeA Cell permeability [35] OqxAB Efflux pump; plasmid-mediated quinolone resistance [36] qnr Encoding a family of proteins that protect DNA gyrase and topoisomerase IV from quinolone inhibitory activity [37] aa(6 )-Ib-cr Quinolone modification [37]  mcr-1 Encoding a family of phosphoethanolamine transferases that can bind to phosphoethanolamine [52] Tigecycline resistance

AcrAB-TolC, OqxAB
Overexpression of efflux pumps lead to tigecycline resistance [53] RarA, RamA, RamR and AcrR Regulators of efflux pumps [53] Lon and rpsJ Encoding ribosome protein S10 [54] ompK35K Decreased transcript level of porin ompK35K can also lead to enhanced resistance [55] tetA Encoding tetracycline-resistant efflux pumps [56] 3. Biofilm K. pneumoniae can form biofilm, and it accumulates in such cells embedded in the self-generated matrix of the extracellular polymeric substance ( Figure 1). Extracellular polymeric substances are complex structures containing polysaccharides, proteins, and DNA. Most clinically apparent K. pneumoniae biofilm is formed on the catheter and the inner surface of internal devices [57]. K. pneumoniae biofilm can lead to colonization in the respiratory, gastrointestinal, and urinary tracts, as well as the development of invasive infections (especially in immune-deficient patients). The development of K. pneumoniae biofilm on the surface of hard objects undergoes the adhesion of cells, the formation of small colonies, maturation, and the propagation as ultimately free-living cells.
The most important surface structures of K. pneumoniae involved in the formation process are type 3 pili and capsular polysaccharides (CPs) [58]. Pili mediate steady adhesion, while CPs ultimately affect biofilm structure and intercellular communication. Given the dynamics of biofilm formation and the variability of environmental stimuli, embedded bacterium must have the ability to rapidly and extensively change gene expression. Transcriptional regulation is regulated by quorum sensing, a system coordinating the signals and responses that control gene expression in a microbial population. The putative quorum-sensing-related regulators and autoinducers in K. pneumoniae have been identified, but the relevant available data is still incomplete [58]. K. pneumoniae with biofilm is protected from the host immune response, in part. This matrix inhibits the proximity of antibodies and antimicrobial peptides, and reduces the effects of complement and phagocytosis [59]. Mutations in some specific genes of K. pneumoniae can also affect the function of biofilm. Mostafavi et al. found that fabZ and lpxC mutations lead to lpxC inhibitor-dependent growth of K. pneumoniae, which leads to the loss of biofilm homeostasis [24]. In recent years, the research on the outer membrane protein of K. pneumoniae has attracted the attention of researchers. Hsieh et al. reported that YfgL (BamB) lipoprotein is involved in the biofilm formation of K. pneumoniae and the transcriptional expression of type 1 pili, which is critical for the anti-phagocytosis of K. pneumoniae in vivo [25]. Besides this, K. pneumoniae's outer membrane protein A (KpOmpA) is reported to be involved in cell adhesion and cell-cell recognition, as well as immune response of host. Meanwhile, the L3 site on the KpOmpA surface may be related to pathogenicity of K. pneumoniae [26]. Oxidative stress may oxidize and destroy the biofilm of K. pneumoniae, resulting in the loss of major membrane proteins and activity of K. pneumoniae [60]. The above related studies have revealed that K. pneumoniae's biofilm is one of the crucial conditions for maintaining its activity.

Aminoglycoside Resistance Gene
Aminoglycosides, which can inhibit bacteria protein synthesis, were used in antibacterial chemotherapy extensively from 1940 to 1980, until they were replaced by following carbapenems, cephalosporins, and fluoroquinolone drugs [61]. During this time, K. pneumoniae acquired antibiotic resistance mechanisms, including drug-modifying enzymes with diverse activities (adenylation, acetylation or phosphorylation). Within ten years, all types of plasmid-mediated resistance genes of aac, aph, and ant gene families in K. pneumoniae were found. The reduced use of aminoglycoside slowed the evolution of the novel resistant genes until the 16SrRNA methylase occurred, which was belonged to the armA gene family, encoding an enzyme that blocks the binding of aminoglycoside antibiotics to the 16SrRNA [27]. These resistant genes are encoded by plasmids in K. pneumoniae, whereas drug-modifying enzymes have narrow-spectrum activity, and 16 rRNA methylase confers resistance to almost all aminoglycosides, including plazomicin and recently developed aminoglycosides [62]. However, an epidemic surveillance, conducted by Cirit et al. in Turkey and Syria in 2019, found that the aminoglycoside resistance of clinically isolated K. pneumoniae was mainly mediated by aac (3)-II, rather than 16SrRNA methylase [63]. The chromosomal location of armA gene family in K. pneumoniae has merely been reported once before [64]. Other known plasmids that mediate the activity of 16S rRNA methylases, including the NpmA and Rmt gene family, have also been identified in K. pneumoniae [61], but there is still no evidence of chromosomal localization.
The mechanism of K. pneumoniae's chromosomal tolerance to aminoglycosides includes modification of cell permeability (because of changes in the KpnEF efflux pump systems and AcrAB-TolC, assuming loss of porins KpnO) and aminoglycoside-modifying enzymes (AMEs) genes. Deletion of AcrAB-TolC increased the sensitivity of tobramycin and gentamicin [28], whereas the kpnEF mutant showed significant resistance to tobramycin and spectinomycin [29], but only slightly tolerated gentamicin and streptomycin. It indicated that the osmotic device has a different affinity for different aminoglycosides. Studies have reported that in vitro KpnO pore proteins are directly involved in aminoglycoside resistance, and they can lead to the tolerance of tobramycin, streptomycin, and spectinomycin [30]. No mutations conferring resistance by target modification, such as rpsL or rrs, have been identified in the K. pneumoniae clinical strains. The causes may be high fitness costs and reduced virulence associated with rpsL mutations, so they are relatively rare [31]. Among gentamicinor amikacin-resistant K. pneumoniae strains, acc (6 )-Ib was considered to be the most widespread, followed by acc (3 )-II, aph (3 )-IV, and ant (3 )-I [65]. In addition, aminoglycoside resistance was also found in strains producing β-lactamases, which was realized by multiple drug resistance coding genes in plasmids, and K. pneumoniae strains coexisting with rmtC and blaNDM-1 gene were isolated from Turkish clinical samples [66]. Hence, we should not only pay attention to the emergence of aminoglycoside resistance genes, but also pay attention to multidrug-resistant K. pneumoniae.

Quinolone Resistance Gene
Quinolone antibiotics target bacteria topoisomerases in order to inhibit DNA replication. These antibiotics have been applied clinically since the 1960s, but they are more widely used after the use of the first class of fluoroquinolones in the 1980s, which leads to the evolution of bacterial fluoroquinolone resistance [67]. The mechanism of K. pneumoniae resistance to fluoroquinolone mainly includes mutation of target gene, MDR efflux pump production, and modification of enzymes and target protection proteins [68].
K. pneumoniae treated with the first-generation quinolone drug nalidixic acid and the firstgeneration fluoroquinolone drug ofloxacin was also accompanied by chromosomal resistance of nalidixic acid and ofloxacin. The primary resistance mechanism is the mutation of DNA gyrase (gyrA-gyrB subunit) binding to the quinolone on the chromosome and topoisomerase IV (parC-parE subunit). Mutations in parC and gyrA in K. pneumoniae were found earlier, and more commonly than those in gyrB and parE [32,69].
Studies have reported that changes in cell permeability in K. pneumoniae are also involved in quinolone resistance, mainly including OmpK36 deficiency [33], overexpression of the multidrug efflux pump gene acrAB [34], and non-alteration of kdeA [35]. Induction of these multidrug efflux pump genes into sensitive E. coli in vitro will increase their drug resistance. Another important efflux pump from OqxAB in K. pneumoniae, which was originally thought to derive from the chromosome, has now been shown to be involved in plasmid-mediated quinolone resistance (PMQR), and is widespread in other bacteria [36]. The efflux pump regulator was also revealed to be participated in the quinolone resistance of K. pneumoniae [70].
Another class of quinolone resistance genes includes the PMQR determinant, which is present not only in K. pneumoniae, but also in other Enterobacteriaceae species. PMQR determinant is considered to be one of the key factors mediating fluoroquinolone resistance in K. pneumoniae [71,72]. These genes contain the qnr gene, which encodes a protein family that protects DNA gyrase and topoisomerase IV from the inhibition of quinolone. The first qnr gene was found on the K. pneumoniae plasmid isolated in USA in 1994. QnrS1, qnrD, qnrB, and oqxAB pump genes could be detected in K. pneumoniae multidrug-resistant strains [73]. Although these proteins are chromosomally encoded in other gram-negative bacteria (including Shewanella algae, Citrobacter spp, Stenotrophomonas maltophilia, and Serratia marcescens) [37], there is still no evidence that the qnr gene is on the chromosome of K. pneumoniae. This suggests that the qnr gene may be one of the genes encoding K. pneumoniae quinolone-resistant plasmids.
Another PMQR gene, aa(6 )-Ib-cr, is considered as the only gene in K. pneumoniae, which is responsible for quinolone modification [37]. It can inactivate narrow-spectrum quinolones, such as ciprofloxacin and norfloxacin, which carry the substrate of this enzyme, the unsubstituted piperazine group. Originally it was found to be encoded by the plasmid of K. pneumoniae, but recently was also found on chromosomes as well. Aminoglycoside resistance is usually accompanied by β-lactam and quinolone resistance; aa(6 )-Ib-cr confers aminoglycosides (kanamycin, tobramycin, and amikacin) and quinolones (norfloxacin and ciprofloxacin) resistance [74].

β-lactam Resistance Gene
β-lactam antibiotics have been used clinically since the 1940s, and are a major class of human medicine-specified antimicrobials. Since then, hundreds of various β-lactamases in intestinal pathogens have evolved, and the number is quite alarming (>2000) and diverse [75].
The penicillin resistance of K. pneumoniae was first found in 1960s, which also led to the discovery of the first class of β-lactamase genes (blaSHV-1 and blaTEM-1). Twenty years later, the first broadspectrum β-lactamase gene blaSHV-2 was identified in K. pneumoniae obtained from ICU patients in Germany. This gene exhibits a broad spectrum of β-lactam tolerance activities (cephalosporins and monochosamines included) and is defined as the first discovered ESBL gene. Soon, another plasmid-mediated ESBL variant, blaTEM-3, was found in K. pneumoniae in France [38]. Since the emergence of ESBLs in K. pneumoniae during the period 1990-2000s, it has become the main ESBL-borne pathogen in the outbreak of nosocomial infection. Of the clinical strains of K. pneumoniae in all Iraq and Spain hospitals, 40% of the strains are ESBL [76]. During this period, the K. pneumoniae strain, which contains major blaTEM and blaSHV β-lactamases, has a high incidence, and is prevalent in many countries [77]. In 2000, due to the availability of plasmids and transposons encoding blaCTX-M type ESBLs, the types of ESBLs in K. pneumoniae caused by iatrogenic outbreaks was altered [39]. Some studies have found that β-lactam resistance can be activated by ramA, and the excessive production of ramA in K. pneumoniae plays an important role in improving acquired β-lactamase-mediated β-lactam resistance. Proteomic analyses show that this enhancement is mainly achieved by activating efflux pump production [40].
Another type of ESBL gene is transferred to K. pneumoniae through horizontal gene transfer (HGT), blaOXA type ESBLs included [41] and scarce genes blaGES and blaSFO [42], or blaPER, blaTLA, and blaVEB [43] and blaKLUC-5 [44]. In addition, β-lactamase genes that are tolerant to inhibitors have emerged, and they encode enzymes that are partially inhibited by tazobactam and clavulanic acid [75]. In the 30 years since their appearance, the K. pneumonia that produced ESBL have increased worldwide. Based on the reports of World Health Organization (WHO), K. pneumoniae with ESBL has reached the rate of 50% prevalence in numerous parts of the world, and a resistance rate of 30% in the community, indicating its ubiquity of resistance.

Polymyxin Resistance Gene
Polymyxin disrupts membrane integrity by displacement of extracellular ions (Ca 2+ /Mg 2+ ) with binding to negatively-charged lipopolysaccharide (LPS), and causing cell lysis [78]. The history of K. pneumoniae polymyxin resistance is shorter compared with other antibiotics, due to the limiting use of medicine in the 1980s and 2000s. Clinically, the first colistin-resistant Klebsiella aerogenes (now categorized as Streptococcus pneumoniae) was isolated when firstly used [79]. In the early 2000s, as more and more carbapenemase XDR K. pneumonia (CPKP) strains appeared, polymyxin became the last class of drugs used [80]. The first nosocomial infection of polymyxin non-sensitive MDR K. pneumoniae was found in Greece in 2004 [80]. Since then, there are more and more clinical reports on the emergence of polymyxin-resistant strains.
The main mechanism of K. pneumoniae polymyxin resistance is target modification by chromosomal mechanism, also known as "LPS modification system". Strains with this complex system can change the structure of LPS and cause negative ion reduction by affecting on polymyxin binding [10]. These changes in LPS are due to mutations in core genes responsible for the maturation of lipid A (lpxM and its regulator ramA) [45,46] and lipid A neutralization, amino arabinose (pbgP, pmrE), an additional combination of phosphoethanolamine (pmrC), or palmitate (pagP) [47,48]. Multiple LPS modified gene regulators, such as phoPQ, pmrA, and pmrD, are also involved in polymyxin resistance. Mutations in one of the two other regulatory genes (resulting in overexpression of prmB or inactivation of mgrB) are also sufficient to cause polymyxin resistance [49,50]. In 2015, Wright et al. reported that another underlying pathway for LPS modification is TupA-like/glycosyltransferase and CrrAB regulatory systems [81]. Another mechanism of K. pneumoniae colistin resistance is capsular polysaccharide (which may camouflage charged molecules on the outer membrane) [29], as well as high expression of efflux pumps AcrAB-TolC and KpnEF (due to positive regulation by RarA) [46]. Pal et al. found that the deletion of glycosyltransferase WcaJ in K. pneumoniae made the mutant strain show non-mucus phenotype, which increased polymyxin resistance [51]. Plasmid-mediated polymyxin resistance has merely been reported recently, and the mcr-1 gene has been identified in China. This gene encodes a family of phosphoethanolamine transferases that share the same activity as PmrC and can bind to phosphoethanolamine [52]. Previous studies have shown that polymyxin resistance occurs in patients, rather than being caused by infection, so limiting the use of polymyxin can reduce mcr-1-mediated polymyxin resistance [82]. Phenotypic screening of acquired polymyxin resistance mediated by the mcr-1 gene includes Ethylene Diamine Tetraacetic Acid (EDTA) and dipicolinic acid (DPA) test [83] and MALDIxin test [84].

Tigecycline Resistance Gene
Tigecycline is the first generation of glycylcycline [85], which has been used since 2005 and is used to treat K. pneumoniae infection. Tigecycline is a promising drug with broad-spectrum antibacterial activity, and is effective even for ESBL-derived strains [86]. Shortly after the first usage, a K. pneumoniae MDR strain (reduced tigecycline sensitivity, MIC = 4 µg/mL) was isolated in a hospital [87]. The known tigecycline tolerance mechanisms are chromosomally encoded, including modifications of the 30S and 16S ribosomal unit targets, as well as changes in cell permeability. Studies have shown that overexpression of efflux pumps AcrAB-TolC and OqxAB and changes in the expression levels of their regulators (RamA, RamR, RarA, and AcrR) can also lead to tigecycline resistance [53]. The mutation of RamR gene can increase the expression of RamA [88]. In addition, Lon and rpsJ genes were also associated with tigecycline resistance in K. pneumoniae [54]. Decreased transcript levels of porin ompK35K can also lead to enhanced resistance of K. pneumoniae strains [55]. The first mutation in the ribosome protein that causes a decrease in sensitivity occurs on the rpsJ-encoded protein S10 [89]. The tetracycline-resistant efflux pumps encoded by the tetA gene were found in non-sensitive K. pneumoniae isolates, but their mechanism of resistance to tigecycline remains unclear [56].

Application of Whole Genome Sequencing
Whole genome sequencing enables in-depth identification of bacteria and will provide an effective tool to study and compare in-hospital infections and outbreaks [90] and epidemiological surveillance [91]. To better understand the virulence and prevalence of the K. pneumoniae strain, Brisse et al. developed a free BIGSdb-Kp database [92], which can help scientists to quickly obtain medical and epidemiological information from the K. pneumoniae genome sequence. In addition, several studies have used high-throughput sequencing to obtain and compare the genomes of endemic highvirulent (CG23) and almost avirulent MDR strains (CG258). Bialek-Davenet et al. had shown that CG258 strain is almost virulence-free but exists various drug resistance-associated genes, such as mutated gyrA and parC genes in the quinolone resistance-determining region. Meanwhile, most of the genes of MDR and high virulent strains do not overlap [92]. Struve et al. have shown that the CG23 clinical strain is a highly efficient colonizer of the human intestinal tract [93]. They detected large virulence plasmids encoding two iron vectors (bacteriocin and salmochelin) in all hvKP strains, and found that these strains also have other iron carriers, such as ICE-related Yersin, colibactin, and microbacteria E492. At present, the application of whole genome sequencing in the field of K. pneumoniae also includes exploring the virulence, biofilm formation, and drug resistance mechanism of K. pneumoniae at the genome level. Rimoldi et al. found that the existence of blaKPC gene is related to carbapenem resistance and the relationship between the type 3 pili of K. pneumoniae and iron carrier genes by virulome profile [94]. Meletis et al. revealed that the NDM-1-encoding K. pneumoniae capsular serotype is determined by the nucleotide sequence of the wzc gene. The carbapenem-resistant K. pneumoniae genome includes 16 drug resistance genes; 12 located in plasmids and 4 in the chromosome [95]. Founou et al. sequenced the whole genome of ESBL-producing K. pneumoniae by Illumina MiSeq platform, and found that ESBL-K. pneumoniae contains a variety of β-lactamase genes, including blaOXA-1, blaTEM-1b, blaSHV-1, blaCTX-M-15, and other drug resistance genes. The replicon plasmid types of ESBL-K. pneumoniae were also detected [96]. Generally, whole genome sequencing has been used in various fields of K. pneumoniae study, and we believe that it will become a major tool for K. pneumoniae study in the future.

Application of Global Proteomics
Proteomics has been widely used in major aspects of the bacteria field, such as the identification of bacteria [97] and the development of vaccines [98]. This technique can help us to better analyze the molecular mechanisms of bacterial infection and bacteria-host interaction [99]. Kamaladevi et al. showed that K. pneumoniae mainly affected the metabolic pathway of the host during infection, and further experiments showed that K. pneumoniae destroyed the host's defense system by inhibiting the PI3K/AKT/mTOR pathway [100]. Proteomics has been applied to investigate the mechanism of drug resistance in K. pneumoniae. Sharma et al. used proteomics (LC-MS/MS) and bioinformatics methods to analyze the possible relationship between K. pneumoniae and carbapenem resistance, and found that 52 related overexpressed proteins and their interactive proteins may have contributed to the survival of K. pneumoniae under meropenem stress and the emergence of meropenem resistance through diverse mechanisms or multiple pathways [101]. The application of proteomics can also be used as an effective treatment for K. pneumoniae. Anand et al. used SDS-PAGE and NLC-MS/MS to find that phage therapy of K. pneumoniae through the nasal approach is an effective measure [102]. Therefore, we can believe that more breakthroughs in the field of K. pneumoniae will be brought by the application of proteomics.