Introduction

Antimicrobial resistance (AMR) threatens the utility of our current antimicrobial armamentarium. Without intervention, antibiotic-resistant infections are estimated to cause approximately ten million deaths globally per year by 2050, costing US$100 trillion1. Gram-negative bacterial pathogens, including Enterobacteriaceae, pose the most significant threat and many have been designated critical priority pathogens for research and development by the World Health Organization2. Infections caused by these pathogens are increasingly becoming untreatable or requiring last-resort antibiotics with high rates of severe adverse effects3,4. Two of these organisms are of greatest concern—extra-intestinal pathogenic Escherichia coli (ExPEC) and Klebsiella pneumoniae.

Vaccination targeting ExPEC and K. pneumoniae is a vital approach in the global strategy against AMR, alongside the development of new antibiotics and improved antimicrobial stewardship. Vaccines have the potential to reduce antibiotic-resistant and antibiotic-susceptible infections, subsequent antibiotic usage (which will further delay AMR development), and maintain the effectiveness of current antimicrobials. Vaccines have historically been highly successful at reducing the spread of bacteria with high AMR burden, including Haemophilus influenzae and Streptococcus pneumoniae5,6,7,8. However, the development of vaccines against ExPEC or K. pneumoniae is dependent upon the robust understanding of host-pathogen interactions, AMR mechanisms, and previous vaccine development approaches. Herein, we provide an overview of these two organisms, the current status of vaccine development, and discuss future considerations to inform vaccine development.

Search strategy and selection criteria

References for this review were identified by searches of PubMed and references from relevant articles using the following search terms: “Klebsiella pneumoniae”, “ExPEC”, and “Escherichia coli”. Articles published in English between 1980 and 2024 were included. Articles resulting from these searches and relevant references cited in those articles were included. Additionally, published reports from the World Health Organization and information from the webpages of vaccine developers and clinicaltrials.gov were also included.

Epidemiology and clinical manifestations

ExPEC1,2 causes a variety of types of infections (Fig. 1) and contributes to a large burden of community-acquired (outside healthcare settings) and hospital-acquired (48 h or more after admission) infections9. It was recently estimated that E. coli contributes to 950,000 deaths annually worldwide10. ExPEC is the leading cause of death due to peritoneal and intra-abdominal infections11, and the major cause of UTIs worldwide (Table 1)12. This opportunistic extracellular pathogen exploits host receptors and deploys bacterial virulence factors to colonize the urinary tract13. Once adhered to the urinary epithelium, organisms preferentially colonize the bladder, causing cystitis. If left untreated, ascending infection can occur via the ureters into the kidneys, causing pyelonephritis, which can subsequently lead to invasive diseases such as bacteremia and sepsis. Recurrent UTIs are also common, with a study of 179 adult women who had experienced cystitis caused by E. coli noting that 44% of patients experienced at least one recurrent infection after their index infection14. This is related to the pathogen’s ability to form intracellular biofilm communities that burst out of the urothelial cells11.

Fig. 1: An overview of the clinical manifestations (infections) that are most commonly caused by ExPEC or K. pneumoniae.
figure 1

The figure on the left details those common manifestations of ExPEC infection, while the figure on the right highlights manifestations of K. pneumoniae. Both ExPEC and K. pneumoniae are common causative agents in bloodstream infections (bacteraemia), urinary tract infections (UTIs), pneumonia, and many other types of infections. Pyogenic liver abscesses, however, are most often associated with hypervirulent K. pneumoniae, whereas K. pneumoniae rarely causes meningitis. Life-threatening sepsis is a concern for both pathogens. Note, this figure is not exhaustive but details the most common manifestation of these pathogens. Created in BioRender.

Full size image
Table. 1 Epidemiologic, clinical, and microbiological characteristics of extra-intestinal pathogenic Escherichia coli (ExPEC) and Klebsiella pneumoniae
Full size table

Treatment of ExPEC is becoming more challenging due to the emergence of resistance to third-generation cephalosporin antibiotics. ExPEC carrying beta-lactamase (bla)-encoding genes including blaCTX-M-1 variants (with high prevalence of blaCTX-M-15), blaNDM-1, and blaOXA-48 have been reported worldwide. Multilocus-sequence typing (STs) enables bacteria to be classified into different lineages, or STs, based on the nucleotide sequence of a small number of relatively well-conserved housekeeping genes15. These highly resistant ExPEC mostly belong to the ST131 lineage, followed by ST38, ST405, and ST648 (Table 1)10,11,12,13,14. These STs are a major cause of UTIs and bloodstream infections and are associated with increased morbidity and mortality15,16. In one study, neonatal bacteremia was caused predominantly by ST131 and ST95, while ST1193 and ST127 were lower in prevalence17. In a systematic literature review of ExPEC lineages globally, ST131 has been the most commonly reported ST since 20009. Other STs have been important, including ST73, most commonly associated with bloodstream infections and ST95, known for the ability to colonize and persist within the intestine9.

K. pneumoniae is an opportunistic pathogen that often colonizes the gastrointestinal tract (GI), the nasopharynx, and the skin of humans. It is also commonly found in natural (including soil and water) and hospital environments, such as sinks, drains, and medical devices. Asymptomatic colonization and the propensity of K. pneumoniae to spread within the hospital environment have been shown to be major reservoirs of hospital-acred infections18. Historically, K. pneumoniae has primarily affected individuals with underlying comorbidities, such as diabetes, chronic liver/kidney disease, cancer, and solid organ transplant recipients, causing serious infections such as pneumonia and bacteremia and less invasive infections such as UTIs (Table 1)19,20,21. Increased infections are now also being observed in healthy individuals, including community-acquired infections such as pneumonia, severe skin and soft tissue infections (e.g., cellulitis, necrotizing fasciitis, and myositis), and abscesses in the neck, lungs, liver, and kidneys (Table 1).

There are two major pathotypes of K. pneumoniae named ‘classical’ and ‘hypervirulent’. These classifications are often based on infection types, and classical K. pneumoniae is often associated with extensive drug resistance22. Hypervirulent K. pneumoniae is more often associated with less-common infection types occurring at distinct bodily sites, including pyogenic liver abscesses and necrotizing fasciitis22. Published mortality estimates associated with hypervirulent and classical K. pneumoniae bacteremia are comparable, ranging between 37% and 65%23. For now, these hypervirulent strains are predominantly found in South-East Asia, but their increasing presence in countries such as Ireland is notable24.

In one meta-analysis bacteremia caused by K. pneumoniae in adults had a much higher case-fatality rate (54%) compared with UTIs (14%)25. Factors associated with increased risk of death included admission to the intensive care unit, having an underlying medical condition, age ≥65 years, and requiring mechanical ventilation26. Moreover, K. pneumoniae strains with carbapenem resistance, containing carbapenemase-encoding genes such as blaKPC and blaVIM, had a case-fatality rate of 42%, compared with only 21% in infections caused by carbapenem-susceptible K. pneumoniae25. Alarmingly, the global spread of carbapenem-resistant K. pneumoniae isolates has been increasingly documented, underscoring the ‘resistance epidemic’; these infections have been found in the United States since 1996, United Kingdom (2003), China (2004), Argentina (2006), Poland (2008), Greece (2012), and more27.

K. pneumoniae is a major contributor to death in low- and middle-income countries (LMICs)23. K. pneumoniae was identified as the leading infectious cause of mortality in children under 5 years of age beyond the neonatal period, accounting for 31% of deaths28. K. pneumoniae is also a significant cause of neonatal sepsis in LMICs29. Case-fatality in neonates with bacteremia due to ESBL-producing Enterobacteriaceae, including infections caused by K. pneumoniae, is reported as 36%, compared with 18% among all other neonates with bacteremia30.

Virulence factors and pathogenesis

ExPEC employs a diverse repertoire of virulence factors to facilitate pathogenesis, allowing for adhesion to host cell surfaces outside the GI tract, evasion of the immune system, and nutrient acquisition (Fig. 2A). These include a capsular polysaccharide (K-antigen), lipopolysaccharide (LPS; O-antigen), iron-chelating compounds (siderophores), fimbriae (Type 1, P, S, and F1C pili), flagella (H-antigen), and other adhesins that mediate attachment to host urothelial cells. There are >180 O-antigen serotypes currently described; however, only a subset of 10–12 (O1, 2, 4, 6, 7, 8, 16, 16/72, 18, 25, 50, and 75) are found amongst the majority of ExPEC infections31 The capsule and flagella are less diverse than LPS, but nevertheless over 80 distinct K-antigens and 50 H-antigens have been described. The K1 antigen is a predominant antigen among invasive strains (particularly meningitis)32, but vaccines targeting K1 would unlikely be successful since this molecule resembles sialylated glycoproteins found in neuronal tissue and is thus unlikely to be immunogenic to the human immune system – if it was, eliciting autoimmune responses would be a concern.

Fig. 2: Notable virulence factors and vaccine targets for ExPEC and K. pneumoniae.
figure 2

In both ExPEC (A) and K. pneumoniae (B), capsular polysaccharide K-types, lipopolysaccharide (LPS) O-types, porins, and fimbriae have been investigated in vaccine design. In ExPEC, the adhesin flagellum and various siderophores have been studied, while outer membrane vesicles (OMVs) have been utilized for Klebsiella vaccine design. Both organisms are host to a number of β-lactamases.

Full size image

K. pneumoniae also employs several virulence factors that contribute to its pathogenesis, including the capsular polysaccharide (K-type), LPS (O-type), fimbrial proteins, porins, and siderophores (Fig. 2B)33. The capsular polysaccharide, synthesized by-products of the cps gene locus, is perhaps the best-understood virulence factor of K. pneumoniae, and 141 K-types are used in K. pneumoniae serotyping34. Hypervirulent K. pneumoniae is associated with two predominant capsular polysaccharide types, denoted as K1 and K2. However, associating multidrug resistance with a particular K-type is challenging, as these isolates encompass a range of K-types. Beyond K-types, 13 O-types have been identified in K. pneumoniae34. Serotypes O1, O2, and O3 are associated with most disease-causing isolates35. The O1 serotype has been associated with AMR36.

Immune response to ExPEC

ExPEC is genetically distinct from commensal and intestinal E. coli pathotypes, although the transition from gut flora resident to the pathogen is poorly understood37. Additionally, the existence of ExPEC on the intestinal, vaginal, or oropharyngeal mucosa is insufficient to stimulate a protective bactericidal antibody response, and the host is susceptible to recurrent infections even when previously colonized37. This is due to the evolution of mechanisms to evade the host immune system; understanding these mechanisms is key to the development of vaccines against ExPEC.

A key component of the innate immune response against ExPEC is the recruitment of phagocytes, including neutrophils, macrophages and monocytes, to the site of infection (Fig. 3)38. In cystitis, most bacteria remain in the bladder lumen, although some organisms can be internalized in superficial bladder epithelial cells39. This niche allows for ExPEC sheltering and evasion of phagocytosis by incoming neutrophils39. When compared with commensal E. coli strains, ExPEC suppresses the bladder epithelial cytokine response through disruption of the nuclear factor (NF)-kB pathway via IkB stabilization39. This disruption results in the suppression of NF-kB activity, which blocks the secretion of pro-inflammatory cytokines such as IL-6 and IL-8 by uroepithelial cells40. ExPEC survival within neutrophils has been documented following phagocytosis by these components of the innate immune response38. Additional strategies to subvert neutrophils include toxin-induced neutrophil apoptosis and impaired neutrophil chemotaxis38. Despite these evasion strategies, the host immune system can clear ExPEC infections utilizing antimicrobial peptides and reactive oxygen species.

Fig. 3: Both the host innate and adaptive arms of the immune system are involved in the response to infection with ExPEC.
figure 3

The figure displays the numerous components involved in the immune response to ExPEC, using the bladder as an example site. The innate immune response, which is more immediate, is centered around the recruitment of various phagocytes, including neutrophils, macrophages, and monocytes, to the site of infection. The subsequent adaptive response involves antibodies that prevent bacterial adhesion and invasion. Created in BioRender.

Full size image

Adaptive immunity is also important in the host response to ExPEC. Antibodies against ExPEC adhesins, including the autotransporter SinH, have been found to prevent bacterial adhesion and invasion, while also facilitating uptake by macrophages41. Additionally, mucosal IgG may play an essential role in preventing urinary tract colonization and recurrent infections41. In mouse models of ExPEC UTIs, a TH2-biased immune response has been noted as a bladder repair response, however, this skew may hinder more effective bacterial clearance42.

Immune response to Klebsiella pneumoniae

Innate immunity is essential to the response to K. pneumoniae. Complement plays a crucial role in bacterial clearance, particularly the alternative pathway-mediated membrane attack complex (Fig. 4). In pulmonary infection, surfactant proteins enhance bacterial killing via phagocyte recruitment—surfactant protein A serves as an opsonin and stimulator of alveolar macrophages1. Alveolar macrophages have phagocytic capabilities and further amplify the immune response via pro-inflammatory cytokine expression43. The K. pneumoniae capsular polysaccharide is a major protective component for the bacteria in this regard, conferring resistance to bactericidal peptides, complement, and phagocytes43.

Fig. 4: The K. pneumoniae host immune response involves several layers, however, complement-mediated clearance and antibody responses are notable.
figure 4

The figure on the left details the generalized innate immune response to K. pneumoniae, focussing on the clearance of the bacteria from the site of infection. The alternative pathway of the complement system is key in leading to phagocytosis and lysis via the membrane attack complex. In the lungs (right-hand figure), tissue-resident memory cells are key in infection clearance. Antibody responses lead to opsonophagocytosis. Created in BioRender.

Full size image

Tissue-localized and systemic T-cell and antibody (Ab) responses are important in adaptive immunity against K. pneumoniae. For example, K. pneumoniae outer membrane proteins stimulate T-helper cells to secrete IL-17 and IFN-gamma, which is thought to prime for granulocyte colony-stimulating factor release and leukocyte recruitment to the site of infection, ultimately contributing to bacterial clearance44. While the presence of T cells is essential for the clearance of K. pneumoniae infection in mouse models, the subtype appears to be less important; either gamma-delta T cells or classical T cells were able to independently facilitate clearance, while the absence of both did not45. In the context of mucosal immunity in the lungs, tissue-resident memory T cells may also play a vital role46. Ab responses appear to have a key role in host defense against Klebsiella. In humans, defects in Ab immunity have been associated with Klebsiella infection47,48. Studies of intravenous immunoglobulin suggest opsonophagocytic Ab responses specifically are critical for protection49. In animal models, the importance of Ab comes from studies of proposed Klebsiella vaccines, including adoptive transfer studies of serum in mice50,51. The capsular polysaccharide induces Ab responses52, and capsule-based vaccines show protection against infection via bactericidal Abs in macaques53.

Extended-spectrum and AmpC beta-lactamase enzymes

Treatment of ExPEC and K. pneumoniae infections is complicated by the high burden of AMR. Resistance occurs primarily via beta-lactamase enzymes, which are classified into distinct classes (A, B, C, and D) based on their amino acid sequence by Ambler molecular classification54. Ambler classes A, C, and D are serine beta-lactamases, whereas class B comprises metallo-beta-lactamases54.

Ambler Class A enzymes comprise broad-spectrum serine beta-lactamases, which have hydrolytic activity against penicillins and cephalosporins, mainly due to two commonly encountered beta-lactamases designated blaTEM-1 and blaSHV-1 (Supplementary Table 1). Class A beta-lactamases also comprise ESBLs and carbapenemases that hydrolyze both narrow and broad-spectrum penicillins cephalosporins, and/or carbapenems. The most clinically relevant and globally disseminated group of ESBLs include those from the blaCTX-M family, in particular the most globally dominant blaCTX-M-15 found predominantly in E. coli and K. pneumoniae, in addition to other increasingly prevalent ESBL ales such as blaCTX-M-14, and blaCTX-M-2755. The blaSHV family includes beta-lactamases with ESBL activity, such as blaSHV-5, blaSHV-7, blaSHV-12, which are classically associated with K. pneumoniae but can also be found less commonly in other Enterobacteriaceae. Other ESBL alleles include those from the blaTEM family (blaTEM-3, blaTEM-10, blaTEM-52) often found in E. coli and K. pneumoniae, and other less frequent alleles from the PER- (blaPER-1 and blaPER-2) and blaOXA (blaOXA-10, blaOXA-14) families most firstly associated with Pseudomonas aeruginosa and less commonly with E. coli and K. pneumoniae56,57. Although most ESBLs are susceptible to beta-lactamase inhibitors (BLIs) (e.g., tazobactam, avibactam, relabactam, and varbobactam)58,59,60, some blaTEM and blaSHV alleles are resistant to clavulanic acid and sulbactam61,62. Ambler Class C includes AmpC beta-lactamases that confer resistance to most cephalosporins including those with extended-spectrum activity such as ceftazidime, cefotaxime and ceftriaxone as well as cephamycins such as cefoxitin, but cefepime still remains an option63.

Carbapenemases

Increasing prevalence of infections producing ESBLs and AmpC enzymes complicates therapy and limits treatment options, which often leads to the use of broader-spectrum antimiials such as carbapenems64. Unfortunately and predictably, this has led to the emergence of more resistant organisms, such as carbapenemase-producing Enterobacteriaceae. Carbapenemase production, which confers resistance to carbapenems, is a significant concern. This is due to the hydrolytic activity of Ambler Class A serine carbapenemases, Class B metallo-beta-lactamases (MBLs), and some Class D oxacillinases65. Serine carbapenemases of Ambler Class A include blaKPC, blaSME, blaIMI, blaNMC-A, and some variants of blaGES. blaKPC is the most widely disseminated carbapenemase worldwide, found primarily in K. pneumoniae, but also in E.coli and other Enterobacteriaceae conferring resistance to carbapenems, penicillin, and cephalosporins66. Apart from the blaGES carbapenemase reported with increasing frequency in several species (including K. pneumoniae, E.coli, and P. aeruginosa), the other serine carbapenemases are almost exclusively found in a particular species. The blaSME is primarily found in Serratia marcescens while blaIMI and blaNMC-A remain almost exclusively identified in Enterobacter spp67,68. However, a few case reports to date have identified blaIMI in E. coli and Klebsiella69,70. Similar to Class A carbapenemases, MBLs hydrolyze most beta-lactams, including carbapenems, but do not analyze monobactams (e.g., aztreonam). The blaNDM, blaIMP, and blaVIM are the most widespread MBLs reported in Enterobacteriaceae, including K. pneumoniae, E. coli, and Enterobacter cloacae, as well as Pseudomonas aeruginosa, and Acinetobacter baumannii71,72. Class D oxacillinases are oxacillin-hydrolyzing enzymes with carbapenamase spectrum activity, including blaOXA-48 and related enzymes (e.g., blaOXA-162, blaOXA-163, blaOXA-181 and blaOXA-232) that have spread amongst Enterobacteriaceae73. The vast diversity of antibiotic resistance enzymes makes the development of new antibiotics a major challenge, as resistance can rapidly emerge via the evolution of new types of beta-lactamases.

Antimicrobial therapies against highly resistant organisms

Current antimicrobials against most ESBL-producing organisms include beta-lactam-beta-lactamase inhibitor (BLBLI) combinations such as piperacillin-tazobactam, ceftazidime-avibactam, imipenem-cilastatin/relebactam, and meropenem-vaborbactam74. However, other antimicrobial agents can also be used for the treatment of such infections and include cephamycins (e.g., cefepime), temocillin, aminoglycosides, tigecycline, fosfomycin, fluoroquinolones, and trimethoprim-sulfamethoxazole3,74,75. In the case of CRE, “second-line” drugs, such as polymyxins, tigecycline, aminoglycosides, and fosfomycin may be considered74.

Between 2017 and 2023, only meropenem-aborbactam, imipenem-cilastatin/relebactam, plazomicin, eravacycline, and a siderophore cephalosporin (cefiderocol) received regulatory approval from the US Food and Drug Administration (FDA) and/or the European Medicines Agency (EMA) for treatment of CRE-associated infections (Table 2)76,77. Currently, 32 new antimicrobials based upon traditional mechanisms of action are in phase 1–3 clinical trials targeting WHO priority pathogens (excluding Mycobacterium tuberculosis)77. Traditional mechanisms refer to the inhibition of bacterial cell wall synthesis, bacterial nucleic acid/protein synthesis, metabolic pathways or membrane function78. Of those 32 agents, 14 are expected to have some activity against CRE organisms79. The majority of antibiotics in clinical trials, which are mainly BLBLI combinations, have limited or no activity against metallo-beta-lactamases such as the blaNDM (Table 2). This highlights that the antibiotic pipeline remains insufficient to address the challenge of AMR. Further, pharmaceutical industry innovation has largely moved away from the development of novel antimicrobials, due to a lack of profit incentives resulting from short dosing regimens and the rapid bacterial evolution of drug resistance80. Although antibiotic stewardship programs do lead to more appropriate antibiotic use and are a necessary part of any AMR strategy81, these interventions are also fragile as these are dependent on healthcare providers’ perception of antibiotic use, adequate infection control practices, and available infrastructure (e.g., microbiology results provided to clinical care teams in a timely fashion)82. Thus, a systems-level intervention such as the use of vaccines against Enterobacteriaceae may be the key long-term strategy against AMR.

Table 2 Antibiotics and beta-L inhibitors recently approved and currently in the developing pipeline targeting ESBL-producing Enterobacteriaceae and carbapenem-resistant Enterobacteriaceae (modified from the 2021 and 2023 WHO clinical pipeline reports with updates)77,79
Full size table

Vaccines against extra-intestinal pathogenic Escherichia coli

Attempts to develop a vaccine against ExPEC targeting the LPS O-antigen started in the 1990s but were unsuccessful49,83,84,85. The first attempt was a monovalent O18 vaccine-derived by acid hydrolysis detoxification to reduce LPS reactogenicity. The antigen was conjugated to exotoxin A from Pseudomonas aeruginosa (EPA) to improve immunogenicity84, and followed on from the highly successful conjugate vaccines against H. influenzae, S. pneumoniae and Neisseria meningitidis86. This O18 conjugate vaccine-induced antigen-specific IgG responses in rabbits, and promoted opsonophagocytic killing (OPK) of an O18 E. coli84,87. In a phase 1 human trial of 15 individuals, the vaccine was well-tolerated and induced IgG antibodies against O1884. Passive transfer of sera from vaccinated volunteers to mice protected against sepsis49. These studies led to the development of a polyvalent conjugate vaccine comprising 12 O-antigens responsible for the majority of invasive infections in the United States (O1, O2, O4, O6, O7, O8, O12, O15, O16, O18, O25, and O75)85,88. This 12-valent vaccine conjugated to EPA induced 6- to 74-fold increases in IgG antibody titers in pre-clinical studies in rabbits. Passive transfer of immunized-rabbit sera to mice protected against challenge with 9/12 vaccine O-antigens85. In a phase 1 human trial of the polyvalent vaccine, there was a 4-fold increase in antibody responses to only 6/12 targets83. This disappointing result, in addition to challenges in LPS purification, has prevented further progress of this formulation.

Since the early 2000s, new strategies have been investigated for ExPEC vaccine development. Technologies exploiting the natural glycosylation systems of prokaryotic bacteria have enabled faster and more efficient methods for glycoconjugation89,90,91,92,93. This process, termed bioconjugation, uses conjugating enzymes known as oligosaccharyl transferases to transfer polysaccharide from lipid-linked precursors to a carrier protein in the periplasm, resulting in conjugate vaccine that can be further purified91. Using this technology, a tetravalent O-antigen conjugate vaccine, ExPEC4V, has been developed for the prevention of ExPEC disease (Table 3). ExPEC4V comprises four of the most common ExPEC O-antigens associated with UTIs in the United States (O1A, O2, O6A, and O25B) and is conjugated to a detoxified variant of exotoxin A from P. aeruginosa91.

Table 3 ExPEC4V human trials
Full size table

The first clinical trial conducted in women with recurrent UTI demonstrated that ExPEC4V was well-tolerated and elicited robust IgG antibody and OPK responses90. There was a significant reduction in UTI incidence in the ExPEC4V group compared with the placebo group, though there was no significant efficacy against vaccine-specific serotypes90. An additional phase 1 trial of ExPEC4V used higher doses, resulting in robust antibody response90,94. A subsequent phase 2 trial reported serotype-specific antibody responses up to 1 year post-immunization95. Some reactogenicity was noted in the highest dose of 8:8:8:16 μg per serotype95. A second phase 2 trial evaluating a two regimen of 4:4:4:8 μg per serotype was completed in 2019, with results indicating robust and serotype-specific IgG and OPK responses at days 15 and 195 (post-dose 1 and post-dose 2, respectively)95,96.

ExPEC10V, a 10-valent vaccine targeting the ExPEC4V serotypes plus 6 additional serotypes (O4, O8, O15, O16, O18, and O75), has shown promising results in phase 1 and 2 trials in healthy adults aged 60–85 years (NCT03819049). Robust IgG production was demonstrated against all vaccine serotypes, and opsonic killing was successful in all but O897. The increase in valency is expected to significantly enhance the coverage of invasive ExPEC disease strains, from 46·7% to 73·3%, based on data from the United States31,98. The addition of O16 (ST131) is particularly important because this is associated with significant AMR31. To date, clinical trials for these multivalent ExPEC conjugate vaccines have focused on UTIs. However, future studies are crucial to evaluate the potential of this vaccine to protect against other infectious syndromes.

A phase 3 clinical trial of a vaccine including 9 O-serotypes (specific serotypes unknown) is currently in progress at a number of sites across the globe99. This clinical trial is focused on preventing invasive ExPEC disease in adults over the age of 60 years99. The trial is assessing prevention of infection in blood, urine, and other sterile sites and will follow participants for up to 3 years. The trial is expected to conclude in 2028.

While much of the pre-clinical and clinical development of ExPEC vaccines has focused on O-antigen conjugate vaccines, other vaccine types have also been assessed. Such strategies include whole-cell, protein subunits, and nucleic acid vaccines. Live attenuated whole-cell vaccines have had mixed results in pre-clinical models of infection and have thus not made it to clinical assessment. 1 Protein subunit vaccines coing K-antigens (i.e., K1) have been described preclinically to be minimally immunogenic in rats; little protection combined with potential challenges of autoimmunity as described earlier in this manuscript likely contribute to the ultimate failure of these vaccines100. Subunit vaccines have been investigated using alternative outer membrane proteins or virulence factors that are widely conserved across ExPEC strains (e.g., adhesins), however, they have largely been found to be ineffective clinically101. Future strategies may also ide mRNA vaccines of antigen proteins of interest, though feasibility of eukaryotic cellular expn of prokaryotic protein antigens remains a challenge for mRNA bacterial vaccine production102.

Vaccines against Klebsiella pneumoniae

Efforts to develop a K. pneumoniae polysaccharide vaccine started in 1985, with the development of a monovalent vaccine targeting the K1 antigen103 followed by subsequent vaccines with higher valency such as a 6-valent and 24-valent polysaccharide vaccines104,105, all which were shown to be safe and immunogenic in humans104,105,106,107,108. The 24-valent polysaccharide vaccine, which was developed in 1988, included the most prevalent K-types (2, 3, 5, 9, 10, 15–18, 21, 25, 28, 30, 35, 43, 52, 53, 55, and 60–64) identified from Klebsiella associated with bacteremia, with a predicted coverage of ~70%107. Evidence from a Swiss study assessing the 24-valent vaccine in seven healthy volunteers, revealed that this vaccine was safe and immunogenic, eliciting an ≥2-fold IgG antibody levels to the vaccine K-antigens, in addition to cross-reactive IgG against 11 other immunologically related K-types (7, 11–14, 26, 31, 37, 46, 68, 69)107. Subsequent studies evaluated this 24-valent polysaccharide vaccine in combination with an 8-valent P. aeruginosa O-antigen vaccine covalently linked to EPA106,108. The vaccines were well-tolerated in a study of 41 individuals, and elicited robust antibody responses to 33 vaccine antigens (24 K-antigens, 8 O-antigens and toxin A) at 2 months post-immunization106. While these results were promising, no further development of this vaccine occurred; it appears that a combination of lack of commercial interest and pressure to include additional K. pneumoniae serotypes contributed to a halt in vaccine development.

After a long hiatus in K. pneumoniae vaccine development, there has been recently renewed interest in the evidence of the high burden of disease and increasing prevalence of AMR10,109. Multiple K. pneumoniae vaccines are now in pre-clinical evaluation, and some have entered clinical development. Vaccine developers have primarily targeted the two major virulence factors of this pathogen: the capsular polysaccharide K-antigen110,111 and the LPS O-antigen112,113. Other strategies have also been investigated, such as the assessment of outer membrane vesicles (OMVs) from K. pneumoniae46,51,114, inactivated whole-cell K. pneumoniae115,116, targeting outer membrane proteins63,115,116,117,118,119,120, and major virulence factors121,122.

KlebV4 is a tetravalent bioconjugate subunit vaccine adjuvanted with ASO3 and targets the most predominant K. pneumoniae O-types123,124. The vaccine entered Phase 1/2 clinical trials in 2021 in adults in initial dose-ranging studies125. A sublingual immunotherapy, MV140 KlebV4 has been licensed for use in Spain to reduce UTIs caused by various pathogens, including K. pneumoniae, utilizing heat-killed bacteria. The dosing regimen is unusual compared with most vaccines, requiring 2 sprays daily under the tongue for a period of 3 months126. A prospective study has documented a decrease of UTI episodes to zero or 1 episode at 3- and 6-month follow-up investigations in 71·7% and 64·7% of individuals respectively127, but longer-term studies and overall vaccine characterization are lacking. The longest follow-up period of 12 months noted that 59/75 women (78%) who completed the entire dosing regimen reported no recurrent UTI episodes128. No data are available regarding antigen-specific immunity, and protection against other infectious syndromes has not been evaluated.

There are a number of vaccines targeting K. pneumoniae in pre-clinical development (Table 4). Many of these protect against K. pneumoniae challenge in mice and include various vaccine strategies. Different K. pneumoniae outer membrane proteins (OMPs) have been successful in inducing Th1, Th2, and Th17-type immune responses. OMPs that have been evaluated in these studies include OmpA, OmpX, OmpK17, OmpK36, Omp001, Omp002, and Omp005, suggesting that these proteins may be candidates for multivalent vaccines116,117,120,129. Conjugate vaccines have also been studied with a focus on broad coverage of K. pneumoniae. VXN-319, which is based on synthetically produced O-antigens, aims to target greater than 80% of carbapenem-resistant K. pneumoniae strains130. A nanoparticle glyoconjugate vaccine based on K. pneumoniae O2 LPS has demonstrated immunogenicity and protection in systemic and pneumonia mouse models of K. pneumoniae disease113. Vaccines based on K1 and K2 antigens were protected against challenge by K. pneumoniae of the same capsular type in mice111. Finally, two whole-cell inactivated vaccines in pre-clinical development include one candidate that is an inactivated whole-cell Acinetobacter baumannii displaying P. aeruginosa and K. pneumoniae antigens, and an auxotrophic K. pneumoniae vaccine candidate131,132. While a number of these vaccines in pre-clinical development appear to provide protection against challenge, the immune response to these candidates must be better characterized prior to any clinical trials.

Table 4 K. pneumoniae vaccines in pre-clinical development
Full size table

Future research on vaccines

Beyond the existing research, it is useful to note some future directions for vaccine research surrounding ExPEC and K. pneumoniae, including multi-pathogen vaccines, improving gaps in current immunological understandings of infection, and improving vaccine efficacy against numerous infection types. For one, and as mentioned, these pathogens share much in common, causing similar types of infections, and sharing many similar vaccine targets, such as the fimbrial fimH adhesin. New vaccination strategies may wish to take advantage of this and target multiple Enterobacteriaceae. However, it has also been seen that high sequence homology amongst vaccine targets between ExPEC and K. pneumoniae may still result in proteins which trigger different biological responses in the host133. Thus, major work is needed to understand which target antigens are both shared by these pathogens and whether the physiological functions of these antigens may result in their inclusion in vaccines favouring one bacterium over the other. Additionally, there is a current literature gap surrounding the immunological response necessary to provide protection from Klebsiella pneumoniae infections in particular. Establishing a correlate of protection may strengthen vaccine research against this pathogen. Pre-clinical vaccine design often focuses on key antigen selection to provide broad coverage and stimulate humoral responses, perhaps not considering the other specific immunological arms necessary to combat a pathogen. Gathering a broader understanding of the human response to K. pneumoniae infection is key to improving this. Finally, it is very possible to see a focus on improving the efficacy of both ExPEC and K. pneumoniae vaccines such that they are strongly protective across a range of infection types. The specific direction of this work may benefit from following vaccine value profiles, such as the recent K. pneumoniae vaccine value profile134. Such research models the health impacts of different vaccine platforms with different levels of efficacy, duration, and immunization strategy and may help further research with the most possible benefit.

Concluding remarks

Novel strategies are clearly needed to combat resistant Enterobacteriaceae, with vaccination the most promising option to reduce both the prevalence of resistant bacteria and the usage of antibiotics. There have been numerous attempts to develop vaccines against ExPEC and K. pneumoniae, with challenges including target antigen choice and the organization of clinical trials. Further studies of multivalent ExPEC conjugate vaccines are needed to determine vaccine efficacy against UTIs and feasibility as a long-term preventive strategy. Current ExPEC vaccine trials have primarily focused on preventing UTIs, and although these infections constitute a significant proportion of ExPEC disease, the prevention of neonatal meningitis, sepsis, and severe invasive infections through vaccination needs to be considered. K. pneumoniae vaccines have predominantly remained at the pre-clinical stage of development. The gap in complete understanding surrounding the protective immune response necessary to prevent K. pneumoniae disease needs to be investigated further to optimize vaccine development. The heterogeneity of disease caused by K. pneumoniae means that different aspects of host immunity may be necessary to promote infection clearance in the lungs compared with UTI or sepsis. Ongoing global molecular epidemiology surveillance is important to identify which STs contribute to the most disease globally, and which are associated with AMR. Additionally, given the similarities between K. pneumoniae and ExPEC both microbiologically (both Enterobacteriaceae) and pathogenically (range of infectious syndromes), the development of a universal Gram-negative vaccination strategy may be a feasible avenue of development. Future investigations focusing on the development of a universal Enterobacteriaceae vaccine, targeting ExPEC, K. pneumoniae, and other prominent, resistant Gram-negative pathogens will help to reduce the spread of antimicrobial-resistant pathogens worldwide and reduce the associated morbidity and mortality.