Abstract
Klebsiella spp. are early colonisers and commensals of human skin, oral, nasal, throat and gut microbiotas, but are also opportunistic pathogens. Much of our information on these bacteria is derived from genomic studies focused on characterising antimicrobial resistance and virulence genes. Often neglected is the contribution of the bioactive molecules of Klebsiella spp. to mammalian host–microbiota interactions. This article reviews our current knowledge on metabolites and peptides produced by Klebsiella spp. and their interactions with host systems, and highlights areas warranting further studies. New information is also provided on the prevalence of the recently described leupeptin biosynthetic gene cluster among members of the Klebsiella oxytoca species complex.
Introduction
The genus Klebsiella encompasses 17 species of bacteria with validly published names, one species with a non-valid name (‘Klebsiella quasivariicola’) and several uncharacterised taxa (Table 1). While a reliance on phenotypic testing led to misidentification of Klebsiella spp. in the past (Kanki et al. 2002, Ponce-Alonso et al. 2016), characterisation of these facultative bacteria based on marker genes (e.g. rpoB, gyrB, rrs or gyrA), conserved signature indels and the increasing availability of whole-genome sequence data have greatly improved identification and genotyping of Klebsiella spp. (Moradigaravand et al. 2017, Wyres et al. 2020, Cosic et al. 2021, Ma et al. 2021, Shibu et al. 2021).
Taxonomic affiliations of Klebsiella spp. as defined by Ma et al. (2021) based on analyses of genomic data.
| Cluster | Species complex/species* | Phylogroup(s) |
|---|---|---|
| I | K. pneumoniae | |
| K. pneumoniae | Kp1 | |
| K. quasipneumoniae | Kp2, Kp4 | |
| K. variicola | Kp3, Kp5 | |
| ‘K. quasivariicola’ | Kp6 | |
| K. africana | Kp7 | |
| I | K. aerogenes | |
| K. aerogenes | Ka1, Ka2 | |
| Klebsiella sp. (Ka3) | Ka3 | |
| Klebsiella sp. (Ka4) | Ka4 | |
| II | K. terrigena | |
| K. terrigena | Kterr1, Kterr2, Kterr3 | |
| Klebsiella sp002270295 | Kterr4 | |
| II | K. planticola | |
| K. planticola | Kplan1 | |
| K. ornithinolytica | Kplan2 | |
| K. electrica | Kplan3 | |
| Klebsiella sp003752615 | Kplan4 | |
| III | K. oxytoca | |
| K. michiganensis | Ko1, Ko5 | |
| K. oxytoca | Ko2 | |
| K. spallanzanii | Ko3 | |
| K. pasteurii (Ko4) | Ko4 | |
| K. grimontii (Ko6) | Ko6 | |
| Klebsiella sp. (Ko7) | Ko7 | |
| K. huaxiensis (Ko8) | Ko8 | |
| Klebsiella sp. (Ko10) | Ko10 | |
| Klebsiella sp. (Ko11) | Ko11 | |
| Klebsiella sp013705725 | Ko12 | |
| III | K. indica | |
| K. indica | None defined |
Due to difficulties encountered in the storage of K. granulomatis, there is no type culture currently available for this species, so the species has not been included in genome-based studies.
Klebsiella spp. are found in many different environments and sample types (Fig. 1). They are early colonisers and commensals of human skin, oral, nasal, throat and gut microbiotas, but contribute to a wide range of nosocomial infections (e.g. pneumonia, wound, urinary tract or bloodstream infections, sepsis) (Wyres et al. 2020, Appel et al. 2021, Dong et al. 2022, Yang et al. 2022). K. pneumoniae infections are associated with multidrug resistance and/or hypervirulent clones, with hypervirulent strains increasingly associated with community-acquired infections (e.g. pneumonia or urinary tract infections), pyogenic liver abscesses and rarer complications such as endophthalmitis, necrotising fasciitis and meningitis (Dong et al. 2022). K. pneumoniae is by far the most studied species of the genus Klebsiella because of its contribution to the global burden of antimicrobial resistance (AMR) (Antimicrobial Resistance Collaborators 2022). However, reports on AMR in all Klebsiella spp. are increasing, and most members of the genus Klebsiella have been reported as emerging opportunistic pathogens of humans and other animals (Joainig et al. 2010, Appel et al. 2021, Yang et al. 2022, Feng et al. 2024).
Klebsiella spp. detected in publicly available shotgun metagenomic datasets. (A) Summary of different Klebsiella spp. in metagenomic datasets derived from different settings. (B) ‘Animal, bird, insect, plant’ data from (A) broken down by source of metagenome. The colour legend for species applies to both (A) and (B). K. ornithinolytica, K. pasteurii and K. indica were not detected in the dataset presented. To generate the figure, species-level searches were made (14 November 2024) via sandpiper 0.3.0 (Eisenhofer et al. 2024, Woodcroft et al. 2024), with data downloaded and filtered to remove generic ‘organism’ descriptors (e.g. metagenome, metagenomes, synthetic, unidentified, uncultured virus, viral or uncultured bacterium). (A) Data for environmental samples were aggregated and represent a range of sources (i.e. air, aquatic, aquifer, cold seep, compost, food, freshwater, hospital, hot springs, indoor, landfill, marine, mine drainage, sediment, silage, sludge, soil, terrestrial, wastewater or wetland). (B) Data were aggregated into broad descriptive groups (e.g. Chloris|Dromaius|Gallus|Sylvia > bird) to simplify visualisation of the data.
Citation: Microbiota and Host 3, 1; 10.1530/MAH-24-0011
Determination of carriage rates for Klebsiella spp. in global populations has been hampered by misclassification of species, low abundance of these bacteria in adult faeces and (often) a focus only on antimicrobial-resistant strains in faecal contents. Large-scale analyses of metagenomic datasets have shown a decline in the abundance of Klebsiella species between infancy and adulthood (Tarracchini et al. 2022) (Fig. 2). Reports in the literature regarding gastrointestinal carriage with K. pneumoniae vary, but adults in community settings tend to have lower carriage rates compared with those in hospital settings (23 vs 70%) (Bray & Zafar 2024). Gastrointestinal carriage in community settings is dependent on geography, with 19–88% of individuals in Asia and 5–25% of individuals in Western countries colonised with K. pneumoniae (Bray & Zafar 2024). Between 2 and 9% of healthy adults are colonised with K. oxytoca (Joainig et al. 2010). A recent study comparing cultivation- and PCR-based approaches for detection of K. oxytoca-related bacteria suggests their prevalence is much higher in infants (49 and 73%, respectively) (Greimel et al. 2022), decreasing with maturation of the microbiome in the first year of life (Zechner & Kienesberger 2024). In the first 21 days of life, infants delivered by caesarean section harbour significantly higher levels of K. pneumoniae species complex (KpSC) and K. oxytoca species complex (KoSC) bacteria (both ∼10% total microbiota) than their vaginally delivered counterparts (∼5 and ∼2% total microbiota, respectively) (Shao et al. 2019). Representatives of both the KpSC and the KoSC are also detected at higher frequency in infants delivered by caesarean section compared with those delivered vaginally in the first 21 days of life: KpSC, ∼15 vs ∼12%; KoSC, ∼20 vs ∼10% (Shao et al. 2019).
Mean abundance and prevalence of Klebsiella species across the life course based on analyses of metagenomic data. Data are summarised based on results presented by Tarracchini et al. (2022), who analysed infant and adult metagenomic datasets. Life stage: 0–1 month, n = 732 samples; 1–6 months, n = 1,209 samples; 6–12 months, n = 788 samples; 12–24 months, n = 922 samples; 24–48 months, n = 411 samples; adult (18–70 years old), n = 2,555 samples.
Citation: Microbiota and Host 3, 1; 10.1530/MAH-24-0011
The KoSC first received attention due to toxigenic strains of K. oxytoca causing non-Clostridioides difficile antibiotic-associated haemorrhagic colitis (AAHC), a rare condition. More recently, the KoSC has been associated with respiratory infection (Li et al. 2020), necrotising enterocolitis (NEC; reviewed below) and Crohn’s disease (Elmassry et al. 2024). Toxigenic strains are now known to produce metabolites that induce mutations and drive somatic genetic variation in colonic epithelial stem cells in animal models, implicating the KoSC in the development of gastrointestinal cancers (Turocy & Crawford 2025).
Outside the context of AMR, gut-associated KpSC and KoSC bacteria have been studied most intensively with respect to preterm infants (McCartney & Hoyles 2023). Between 1 and 10% of preterm infants harbour Klebsiella spp. in their faecal microbiota, but prevalence of these bacteria can be higher depending on geographical location. A mechanistic understanding of how Klebsiella spp. influence disease onset and/or progression in preterm infants is lacking, because isolates recovered from healthy and sick preterm infants display similar AMR and virulence profiles (Chen et al. 2020b ). However, a Klebsiella/Enterococcus-dominated faecal microbiota is known to be associated with an increased risk of developing NEC (McCartney & Hoyles 2023).
Beyond hypervirulence – linked to iucA, iroB, peg-344, rmpA and rmpA2 encoded on the K. pneumoniae hvKp-specific virulence plasmid – and AMR, adhesins (i.e. type 1 and 3 fimbriae), toxins, lipopolysaccharide, capsule, proteases and iron-harvesting siderophores produced by Klebsiella spp. contribute to their virulence. The type VI secretion system of K. pneumoniae has been shown to mediate intra- and inter-species bacterial competition, and to manipulate host mitochondria to elicit an anti-eukaryotic response (Storey et al. 2020, Sá-Pessoa et al. 2023). In addition, it is becoming increasingly clear that metabolites produced by Klebsiella spp. contribute to the virulence of these bacteria and mammalian microbiota–host interactions, and these will be the focus of the remainder of this review.
Histamine
Enteric bacteria, including K. planticola, K. ornithinolytica and K. aerogenes, are among the dominant histamine-producing bacteria in fish. Some strains of these bacteria, and of K. terrigena, encode a pyridoxal phosphate-dependent histidine decarboxylase (hdc; H+ + l-histidine → histamine + CO2) that can decarboxylate free histidine in the skeletal muscles of fish (Kanki et al. 2002, Rachmawati et al. 2022). In two-thirds of bacteria encoding histidine decarboxylase (including K. aerogenes), a histidine–histamine antiporter facilitates uptake of l-histidine and export of histamine out of the bacterial cell (Engevik et al. 2024) (Fig. 3). Consumption of histamine-rich fish can lead to histamine fish poisoning (scombroid poisoning), a foodborne illness that can cause mild illness or severe cardiac or respiratory distress in those with pre-existing conditions.
Proposed mechanisms through which metabolites produced by Klebsiella spp. interact with the mammalian histamine GPCRs HRH4 and HRH1, and the responses elicited. Some strains of K. planticola, K. ornithinolytica, K. terrigena and K. aerogenes encode a histidine decarboxylase (HDC) that converts histidine to histamine in the skeletal muscles of fish (Kanki et al. 2002, Rachmawati et al. 2022); a histidine (His)/histamine antiporter facilitates uptake of l-histidine (light-blue dots) and export of histamine (red dots) out of the bacterial cell. When histamine-rich fish is ingested, the histamine interacts with HRH4 on mast cells in the gut to produce a proinflammatory response, which can lead to respiratory and cardiac distress in some individuals. High-histamine-producing, K. aerogenes-rich microbiotas in mice have been shown to activate HRH4, leading to mast-cell infiltration and abdominal pain (i.e. visceral hypersensitivity) in these animals (De Palma et al. 2022). Activation of the leupeptin operon by N-acetylneuraminic acid (Neu5Ac; red diamonds) in K. michiganensis, via the nan operon (nanRATEK), leads to the production of leupeptin (a protease inhibitor; dark-blue dots), pyrazinones and pyrazines Li et al. (2020). Hamchand et al. (2024) proposed that leupeptin produced by K. michiganensis increases survival of the bacterium during infections because the leupeptin inhibits subunit beta2i of the immunoproteasome, thereby decreasing immune activity. Hamchand et al. (2024) also showed that pyrazinone 3 (yellow dots) is a selective agonist of HRH4, potentially eliciting histamine release from mast cells and activating proinflammatory and allergic responses similar to those seen via activation of HRH4 by histamine. Presence of HRH4 and HRH1 (plus HRH2 and HRH3) on a range of proinflammatory cells suggests that there is still much work to be done with respect to host–microbiota interactions in the context of histamine production. Information used to generate the figure was taken from Thurmond et al. (2008), Li et al. (2020), Banafea et al. (2022), Vanuytsel et al. (2023), Engevik et al. (2024) and Hamchand et al. (2024). Created in BioRender by L Hoyles (2025) https://BioRender.com/0c9qgu7.
Citation: Microbiota and Host 3, 1; 10.1530/MAH-24-0011
Bacteria from all phyla associated with the human gut microbiota have been found to encode histidine decarboxylase, with members of the phylum Pseudomonadota (particularly Gammaproteobacteria) making the greatest contribution to the production of the biogenic amine histamine from l-histidine (Engevik et al. 2024). G-protein-coupled receptors (GPCRs) are cell-surface receptors that can be activated or inhibited by microbiome-associated metabolites, and are thought to be among the main mediators of mammalian host–microbiota interactions. Bacterially produced histamine can activate the mammalian histidine GPCRs HRH1 to HRH4: HRH1 is expressed by dendritic cells, eliciting the production of pro-inflammatory cytokines; HRH2 is expressed by gut neutrophils and macrophages, eliciting an anti-inflammatory response and suppressing pro-inflammatory cytokines; and activation of HRH4 in mast cells elicits chemotaxis in these cells (Engevik et al. 2024). In a study of individuals with irritable bowel syndrome (IBS), K. aerogenes was identified as a major histamine-producer in individuals with high but not low urinary histamine (De Palma et al. 2022). The bacterium was highly abundant in the faecal microbiota of three independent cohorts of IBS individuals compared with healthy controls. Faecal microbiota transplantation from the IBS individuals to germ-free mice led to establishment of a microbiota that produced high amounts of histamine, with this metabolite associated with activation of HRH4, mast-cell infiltration and abdominal pain in the mice (Fig. 3). Inhibition of HRH4 attenuated the effects of histamine, suggesting modulation of bacterial histamine production could be a means of treating a subset of IBS individuals affected by chronic abdominal pain.
Engevik et al. (2024) noted that no study has linked K. aerogenes to intestinal inflammation, but hypothesised that other pro-inflammatory stimuli (e.g. lipopolysaccharide) produced by this bacterium could drive immune cell infiltration, with histamine potentially activating gut-associated pro-inflammatory cascades via HRH1 and HRH4. As such, in vivo studies in this area are warranted.
Ethanol
Metabolic dysfunction-associated steatotic liver disease (MASLD; previously known as non-alcoholic fatty liver disease) is the most prevalent chronic liver disease globally and increases the risk of developing co-morbidities associated with cardiometabolic disorders (Hoyles 2024). Liver cells contribute to the development of steatosis by converting ethanol into acetate and triglycerides. In a cohort of Chinese patients, ethanol-producing K. pneumoniae strains were found in moderate to high abundance in over 60% of individuals with MASLD/steatohepatitis (Yuan et al. 2019). Ethanol production by K. pneumoniae isolates recovered from patients’ faecal samples was assessed aerobically and anaerobically using yeast extract–peptone–dextrose medium including either fructose or glucose as the sole carbon source; high-ethanol-producing strains generated less ethanol anaerobically than they did aerobically (∼34 vs ∼62 mmol/L). Transfer of the high-alcohol-producing strains of K. pneumoniae to mice caused the rodents to develop hepatic steatosis, demonstrating the bacterium’s metabolic capability contributed to disease. While this study is interesting, its unusual findings warrant the need for further studies to determine the exact contribution of high-alcohol-producing Klebsiella spp. to the development of human hepatic steatosis, as they may only contribute to liver disease in a small subset of patients affected by auto-brewery syndrome, a very rare medical condition (Yuan et al. 2019).
Another mouse study showed that high-ethanol-producing K. pneumoniae induces mitochondrial dysfunction (i.e. reduced ATP and increased accumulation of reactive oxygen species and DNA damage) in hepatocytes, supporting a role for bacterially generated ethanol in development of endogenous alcohol steatotic liver disease (Chen et al. 2020a ). In vivo (mice), production of ethanol by K. pneumoniae is mediated by adh (alcohol dehydrogenase); high-ethanol-producing strains catabolise fructose or glucose to alcohol and 2,3-butanediol via the 2,3-butanediol fermentation pathway (Li et al. 2021).
Colibactin
Colibactin is a polyketide–non-ribosomal peptide hybrid genotoxin that induces DNA double-strand damage and cell cycle arrest of eukaryotic cells in the G2/M phase (Faïs et al. 2018, Strakova et al. 2021). It was first reported in Escherichia coli strain IKE3034 (Nougayrède et al. 2006), and its biosynthesis depends on the 19-gene (clbA–clbS; Fig. 4A), 54 kbp pks gene cluster (Auvray et al. 2021). Colibactin has been demonstrated to have a tumourigenic effect, both in epithelial cell culture and in a colorectal cancer (CRC) mouse model (Strakova et al. 2021). However, other studies have shown that pks-positive bacteria can elicit analgesic, anti-inflammatory and antimicrobial effects. For example, the probiotic E. coli Nissle 1917 encodes a functional pks genomic island (Nougayrède et al. 2021) that is essential for its ability to colonise the gut and elicit its anti-inflammatory and antimicrobial effects in mouse models (Faïs et al. 2018, Strakova et al. 2021).
Organisation of characterised BGCs in Klebsiella spp. (A) pks gene cluster in K. pneumoniae 1084. Annotated according to Auvray et al. (2021), with the 8 bp repeat region (Putze et al. 2009) identified using Geneious Prime 2024.0.5. (B) til BGC of K. grimontii MH43-1. The strain was originally thought to be K. oxytoca (Tse et al. 2017), but was later found to represent K. grimontii upon in-depth analyses of the til BGC (Shibu et al. 2021). cAMP regulatory protein (CRP) binds to the aroX–npsA intergenic region, activating the transcription of both genes (Rodríguez-Valverde et al. 2021). marR is also a transcriptional regulator (Tse et al. 2017). The Leu-responsive regulatory protein (LRP) enhances production of TM/TV by binding to the intergenic region of the genes aroX and npsA and activating their transcription when Leu is present in the growth medium (Cruz et al. 2024), but it is not essential for production of TM/TV by members of the KoSC. (C) leup (leupeptin) BGC of K. michiganensis KCTC 1686 (annotated according to Li et al. (2020)), and the predicted pyr operon of K. michiganensis KCTC 1686 (annotated using primer information derived from Hamchand et al. (2024)). The PGAP-annotated GFF file available from GenBank (assembly GCF_000240325) was used as the source of coding sequences. This cluster type was not detectable with antiSMASH v7.1 (Blin et al. 2023), but will be included in the next release based on data reported here and in Fig. 5 and Supplementary Fig. 1.
Citation: Microbiota and Host 3, 1; 10.1530/MAH-24-0011
Since the gene cluster’s initial description, pks-positive E. coli strains have been recovered from numerous clinical samples (e.g. CRC, newborn meningitis, septicaemia or urinary tract infections) and from the commensal gut/faecal microbiota of asymptomatic carriers/healthy subjects (Faïs et al. 2018). More recently, the pks gene cluster has been detected in other Enterobacterales, such as K. pneumoniae, K. aerogenes, Erwinia oleae, Serratia marcescens and Citrobacter koseri (Auvray et al. 2021). Among Klebsiella spp. colibactin has only been found to be produced by some strains of K. pneumoniae (3.5% prevalence, 5/141 isolates tested) and K. aerogenes (27.3% prevalence, 3/11 isolates tested) (Putze et al. 2009).
In HeLa cells, E. coli-produced colibactin induced DNA damage akin to that seen with cisplatin (a DNA cross-linking drug) (Bossuet-Greif et al. 2018). It is important to note that a close interaction (i.e. cell-to-cell contact) is required between bacteria encoding a functional pks gene cluster and other cells (bacterial or eukaryotic) for effects of colibactin to be elicited. The effects of colibactin cannot be detected when screening bacterial culture supernatants against cells or by metabolomics approaches. Deletion mutants for clbA, clbH, clbP or clbQ were not genotoxic to HeLa cells, but both activities could be restored by complementation of bacterial strains (Bossuet-Greif et al. 2018). Infection of HeLa cells with colibactin-producing E. coli induced an ATR-dependent replication stress response in the mammalian cells like that seen with the interstrand-cross-linking agent mitomycin C.
Colibactin-induced DNA interstrand cross-links covalently link the two strands of DNA and block DNA unwinding, stalling replication forks and DNA replication in eukaryotes. A recent study has demonstrated this stalling leads to replisome disassembly and activation of the Fanconi anaemia interstrand cross-link repair pathway, which promotes resistance to colibactin-induced DNA damage (Altshuller et al. 2024). The repair pathway mediates the introduction of nucleolytic incisions flanking the DNA interstrand cross-links, releasing the cross-links and creating a DNA double-strand break (thought to be repaired by homologous recombination) and an interstrand cross-link remnant that is bypassed by the translesion synthesis polymerase(s). Bypass of the colibactin adducts can be error-free, but can frequently introduce T>A point mutations at the site of alkylation, and these are thought to drive the development/progression of CRC and inflammatory bowel disease (IBD). The gut microbiota of ∼60% of patients with CRC and ∼40% of patients with IBD harbours colibactin-producing bacteria (Altshuller et al. 2024).
Colibactin contributes to interference competition in microbial communities at the small scale, directly harming members of the microbiota it comes into direct contact with. For example, it has been shown to exert antimicrobial activity against Staphylococcus aureus in wound infections by causing irreversible DNA damage. In vitro, colibactin has also been shown to induce prophages in gram-positive and gram-negative members of the gut microbiota, suggesting that the antibacterial activities of this metabolite may contribute to shaping microbiota community composition (Zechner & Kienesberger 2024).
Klebsiella, colibactin and the pks gene cluster
Most of the literature regarding colibactin and/or the pks gene cluster and Klebsiella comprises genetic-based retrospective investigations or surveillance studies (Putze et al. 2009, Conlan et al. 2014, Chen et al. 2017, Khaertynov et al. 2018, Lan et al. 2019, Morgan et al. 2019, Shi et al. 2020, Wami et al. 2021, El-Ashry et al. 2022, Jati et al. 2023, Luo et al. 2023, Wang et al. 2023). These either involve i) bioinformatics analyses of bacterial genomes, sometimes focusing on enterobacteria or Klebsiella, for the pks gene cluster, or ii) PCR for selected pks genes encoded by clinical Klebsiella isolates. Many of these studies do not provide information as to whether strains can produce colibactin, nor whether they did in vivo (in the context of clinical isolates). Furthermore, most of the PCR-based analyses only examine a limited number of the 19 clb genes (Fig. 4A) of the pks gene cluster: most commonly, clbA, clbB, clbN and clbQ are selected as they are from both ends of the pks gene cluster, but this does not mean the full complement of the gene cluster is encoded by specific Klebsiella strains. The clbS gene is responsible for production of the colibactin self-resistance protein (Bossuet-Greif et al. 2018), essential for protecting the bacterium from any colibactin it produces. The gene product of clbP deacetylates pre-colibactins to colibactin and again is essential for colibactin production. Lu et al. (2017) demonstrated that colibactin production contributes to the hypervirulence of K. pneumoniae 1084 and the development of meningitis in mice using a clbP deletion mutant in orogastric, intranasal and intravenous models of infection. Deletion of clbA, which encodes a 4′-phosphopantetheineyl transferase, significantly attenuates the genotoxicity of K. pneumoniae 1084 (Lai et al. 2014, Lu et al. 2017), so inclusion of clbA in pks PCR-based investigations is relevant to potential genotoxicity.
In addition to assaying for only a few pks-relevant genes, many of the above-mentioned studies characterise virulence genes (particularly yersiniabactin, hypervirulence-associated genes and siderophores), commonly associated with Klebsiella spp. Therefore, the contribution of pks to clinical infections cannot be determined, as it is likely that virulence factors such as aerobactin or traits associated with hypervirulence make a greater contribution to disease and patient outcomes than a complete and (potentially) functional pks gene cluster. Common findings from genetic-based investigations of the pks gene cluster are that pks-positive Klebsiella isolates (in most cases K. pneumoniae) display more susceptibility to antimicrobial agents than pks-negative Klebsiella.
In some of the retrospective studies, the demographic and clinical data associated with patients from whom the Klebsiella strains were isolated have been compared between pks-positive and pks-negative groups. Some researchers suggest these associations demonstrate certain demographic or clinical factors increase susceptibility to pks-positive Klebsiella. However, in the absence of mechanistic data, identifying that a higher proportion of clinical Klebsiella isolates from diabetic patients are pks-positive than pks-negative does not mean ‘individuals with diabetes mellitus are more susceptible/vulnerable to pks-positive Klebsiella’ (Lan et al. 2019, El-Ashry et al. 2022).
Determination of whether pks-positive isolates are active in vivo is challenging because colibactin cannot be detected using metabolomics, the active molecule is only transferred from cell to cell, and genes of the pks gene cluster have additional functions distinct from colibactin biosynthesis. The pks gene cluster of strain K. pneumoniae 51-5, isolated from a healthy pre-term infant, was shown to contribute to colitis-associated tumourigenesis, with ΔclbP abolishing this effect in cell culture (Pope et al. 2019). Inoculation of a spontaneous colitis-associated tumourigenesis mouse model with wild-type or ΔclbP K. pneumoniae 51-5 demonstrated similar faecal counts of wild-type and ΔclbP, and significantly more tumours present than in specific-pathogen-free inoculated controls. However, there was no significant difference between the wild-type and mutant groups, highlighting the difficulty of disentangling the contribution of colibactin to disease in the presence of other potential virulence factors encoded by K. pneumoniae. No genome sequence is available for K. pneumoniae 51-5, so it is not possible to comment on the additional virulence genes encoded by this particular strain.
More recently, the in vitro effects of K. pneumoniae isolates recovered from stool and/or biopsy samples of CRC patients (n = 15/43; 3/15 K. pneumoniae pks-positive – all K1) and healthy individuals (n = 25/72; 7/25 K. pneumoniae pks-positive – 5 K2, 1 K5, 1 unknown capsular type) on tumourigenesis have been examined (Kaur et al. 2023). In general, a higher proportion of the pks-positive K. pneumoniae isolates were PCR-positive for siderophores and iron-uptake systems than their pks-negative counterparts. Almost all the K. pneumoniae isolates were PCR-positive for entB1 (enterobactin). Notably, all pks-positive K. pneumoniae isolates from healthy individuals lacked the Klebsiella iron-uptake determinants kfu and kfuBC, and iroN (salmochelin) was not encoded by any of the pks-positive K. pneumoniae from cancer patients. All pks-positive K. pneumoniae were also PCR-positive for ybtS (yersiniabactin). The pks-positive strains from the cancer patients were PCR-positive for the mucoid phenotype determinants rmpA-1 and rmpA-2, and cytotoxin necrotising factor (cnf-1). One of seven of the pks-positive isolates from the healthy individuals was rmpA-1+ and 2/7 were rmpA2 +. Proliferation of human primary colon and CRC (SW1116, SW480, HT29 and HCT116) cell lines was enhanced by exposure to K. pneumoniae antigens compared to PBS controls (MTT assay). pks-positive K. pneumoniae (from cancer-affected and healthy individuals) elicited significantly higher proliferation in cell lines than the pks-negative K. pneumoniae in the early-cancer cell line SW116. Notably, statistical significance was only presented within subject groups and not across all K. pneumoniae groups. However, K. pneumoniae isolated from healthy subjects elicited higher proliferation than K. pneumoniae recovered from cancer patients in HT29 cell lines. pks-positive K. pneumoniae from cancer patients elicited significantly higher proliferation than their pks-negative counterparts for the stage II CRC cell line SW480, but the opposite was true for K. pneumoniae from healthy subjects (pks-negative significantly higher than pks-positive) for SW480. These findings highlight the complexity of disentangling cause from effect in studies involving Klebsiella spp. encoding a functional pks gene cluster.
Tilimycin and tilivalline
Tilimycin (TM; also called kleboxymycin) and tilivalline (TV) are cytotoxic pyrrolobenzodiazepine non-ribosomal peptides that are produced by strains of some species of the KoSC (Zechner & Kienesberger 2024). Both TM and TV can be detected in a range of biological samples at nanomolar and picomolar concentrations, respectively (Table 2). Detection of TV in human serum at higher levels than seen in human faeces has led to the suggestion that TV is efficiently translocated from the gut to the bloodstream (Glabonjat et al. 2021), but its mode of transport and systemic effects have not been determined to date.
Quantification of TM and TV in different biological samples (Glabonjat et al. 2021). All values are presented in terms of wet weight of sample.
| Matrix | TM (nmol/g) | TV (pmol/g) |
|---|---|---|
| Human faeces* | 1.1 ± 0.1 | 8 ± 1 |
| Human serum* | Not detected | 13 ± 1 |
| Mouse faeces* | 51 ± 8 | 22 ± 4 |
| CASO medium † | 24 ± 1 | 94 ± 2 |
Samples collected in active phase of AAHC.
Inoculated with K. oxytoca AHC-6.
TM is a DNA-damaging agent that is encoded by the til biosynthetic gene cluster (BGC), comprising two operons of biosynthesis genes (Fig. 4B). In the presence of indole, TM spontaneously reacts with this peptide to form TV, which targets tubulin and disrupts the spindle apparatus of eukaryotic cells (Unterhauser et al. 2019). TV also induces caspase-3-dependent apoptosis in mice and Hep2 cells and impairs the epithelial barrier (Schneditz et al. 2014). TM is approximately nine times more cytotoxic than TV (Tse et al. 2017). The genes aroX, adsX, icmX and dhbX and both modules of the NRPS operon are required for TM production, along with an immunity protein (encoded by uvrX) and a 3-dehydroquinate synthase (aroB) located apart from the BGC (Schneditz et al. 2014, Dornisch et al. 2017, Tse et al. 2017, Kienesberger et al. 2022).
K. oxytoca encoding a functional til BGC fulfils Koch’s postulates as the causative agent of AAHC (Högenauer et al. 2006), which is characterised by diffuse mucosal oedema, haemorrhagic erosions and bloody diarrhoea. AAHC is caused by the overgrowth of cytotoxin-producing strains secondary to the use of penicillins in older children and adults (Schneditz et al. 2014). Approximately half of all human-associated isolates belonging to the KoSC have been shown to have a functional til BGC via cell-line assays (e.g. MTT, LDH cytotoxicity) or detection of TM/TV by high-performance liquid chromatography or mass spectrometry (Zechner & Kienesberger 2024). TM not only contributes directly to AAHC but also introduces mutations into the DNA of intestinal and bacterial cells. TM exposure in mice, after a single round of K. oxytoca overgrowth, has been shown to lead to the accumulation of mutations in colonic stem cells (Pöltl et al. 2023). In addition, de novo mutations introduced into the genomes of E. coli and K. pneumoniae upon TM exposure led to resistance of these ESKAPE pathogens to rifampicin and nalidixic acid (Kienesberger et al. 2022).
No antibacterial effect has been demonstrated for TV to date. However, TM can inhibit the growth of members of the gut microbiota including Lactobacillus, Bacteroides, Fusobacterium, Proteus and Bifidobacterium spp. (Unterhauser et al. 2019, Ledala et al. 2022). TM production is thought to confer a competitive advantage upon strains encoding a functional til BGC when in the presence of an appropriate carbon source (e.g. glucose). TM (1–170 μM) added to in vitro systems inoculated with a mixed faecal microbiota exerted broad-spectrum antimicrobial activity against a range of gram-positive and gram-negative bacteria (Kienesberger et al. 2022). TM also influenced the composition of the murine gut microbiota, reducing its species richness and evenness (Kienesberger et al. 2022).
Growth medium greatly influences the production of TM/TV, with tryptone lactose broth considered optimal for determining whether KoSC strains produce the cytotoxins (Tse et al. 2017, Rodríguez-Valverde et al. 2021). Maximal production of TM/TV is seen in the late-exponential to stationary phases of growth (Joainig et al. 2010, Rodríguez-Valverde et al. 2021). Expression of aroX and npsA is upregulated in the presence of glycerol or lactose and downregulated in the presence of glucose, high osmolarity or absence of divalent cations (MgCl2 or CaCl2). Because glucose availability is high in the small intestine and glycerol is produced by members of the colonic microbiota (Rodríguez-Valverde et al. 2021), it has been suggested that TM/TV may be synthesised in the glycerol-rich colonic environment under regulation of the cAMP receptor protein (CRP; Fig. 4B). The biosynthesis of TV can be inhibited by salicylic and acetylsalicylic acid in liquid cultures of K. oxytoca. It is thought that this effect is mediated by blocking the peptidyl carrier protein ThdA, and could be exploited as an anti-virulence agent in a combination therapy with antimicrobial(s) to prevent or alleviate the symptoms of AAHC (von Tesmar et al. 2018).
Indole is readily available in the gastrointestinal tract. In addition, KoSC (but not KpSC) bacteria can convert tryptophan to indole via tryptophanase (tnaA), contributing to indole availability in the gut lumen. Indole downregulates expression of npsAB in the NRPS component of the til BGC (Fig. 4B), suppressing TM synthesis and enhancing conversion of TM to TV (Ledala et al. 2022). The host pregnane X receptor PXR (official gene symbol NR1I2, nuclear receptor subfamily 1 group I member 2) is a transcription factor characterised by a ligand-binding domain and a DNA-binding domain. It regulates the expression of the cytochrome P450 gene CYP3A4 and the ATP-dependent drug efflux pump ABCB1 (also known as MDR1), playing an important role in detoxification of xenobiotics. TV (10 μM) is a strong agonist of PXR, with this interaction upregulating expression of CYP3A4 and ABCB1 and blunting the effects of TV on tubulin acetylation. TM is not a PXR ligand (Ledala et al. 2022).
TV stabilises microtubules by increasing the rate of polymerisation of tubulin in the GTP-bound state, which arrests cells in the G2/M phase (Unterhauser et al. 2019). Stabilising microtubules impedes mitosis, thereby inducing cell apoptosis. TV’s mechanism of action is similar to, but different from, that of paclitaxel, a common chemotherapeutic agent that increases polymerisation of GDP-bound tubulin. This difference in mechanism of action would allow TV to be used as an alternative treatment in paclitaxel-resistant tumour cells (von Tesmar et al. 2018, Unterhauser et al. 2019, Kienesberger et al. 2022, Ledala et al. 2022). To the best of our knowledge, no animal or human trials using TV as an alternative chemotherapeutic agent have been conducted to date.
Preterm infants have an immature immune system and, as such, are susceptible to a range of infections (i.e. NEC or early- and late-onset sepsis), with KpSC and – to a lesser extent – KoSC contributing to these (McCartney & Hoyles 2023). Immediately before obvious signs of late-onset sepsis and NEC, blooms of Enterobacteriaceae including Klebsiella spp. are observed in the faecal microbiota of neonates. These blooms impair epithelial barrier integrity and influence intestinal homoeostasis more widely. Translocation of Enterobacteriaceae from the infant gut into the bloodstream can also cause sepsis (McCartney & Hoyles 2023). K. grimontii and K. pasteurii strains encoding the til BGC have been recovered from faecal samples of preterm infants, along with metagenome-assembled genomes of K. michiganensis and K. oxytoca (Chen et al. 2020b , Paveglio et al. 2020, Shibu et al. 2021). Isolates of K. grimontii and K. pasteurii recovered from preterm infants with NEC have been shown to produce both TM and TV by mass spectrometry analysis and apoptosis assays (Paveglio et al. 2020). NEC shares several features with AAHC including blood in the stool and intestinal necrosis. It is, therefore, possible that cytotoxin-producing strains of KoSC in the gastrointestinal microbiota of a subset of preterm infants may contribute to NEC (Paveglio et al. 2020, Shibu et al. 2021). This is an intriguing possibility that warrants further enquiry (beyond that detailed below for (Subramanian et al. 2024)), given that a study of intestinal Klebsiella isolates (K. pneumoniae, n = 8; K. grimontii, n = 3 (all encoding the til BGC); K. michiganensis, n = 2; K. quasipneumoniae, n = 1) recovered from both sick and healthy preterm infants found no difference between the virulence or colonisation potential of these bacteria. They were all able to reside, persist and replicate in macrophages and they all produced siderophores (iron scavengers) in vitro (Chen et al. 2020b ).
Preterm infants are given empiric antibiotics in the first days to weeks of life. Based on 16S rRNA gene-based microbiota profiling, the presentation of NEC symptoms is associated with high abundance of K. oxytoca immediately before disease onset (Paveglio et al. 2020). However, no robust data are available with respect to the gut carriage of K. oxytoca sensu lato (cytotoxic or otherwise) by preterm infants (McCartney & Hoyles 2023). A highly sensitive (15 cfu/mL of sample) real-time PCR assay has been developed for detection of the npsAB genes of the til BGC (Leitner et al. 2022), but this is not currently used routinely for detection of potential cytotoxin-producing bacteria in the faeces of preterm infants. A reanalysis of the shotgun metagenomic data generated by (Shao et al. 2019) from the faecal microbiota of healthy infants at days 7 and 21 of life showed that 76/504 (15%) and 74/325 (23%) samples, respectively, harboured toxigenic Klebsiella. 46/76 samples encoded the til BGC at day 7 (K. grimontii > K. michiganensis > K. oxytoca > K. pasteurii), with prevalence of each BGC-positive species increasing from 9 to 13% between days 7 and 21 of life. Across all BGC-positive samples, the relative abundance of BGC-encoding bacteria was high (range 0.92–94.1% total microbiota, median 12.6% total microbiota) (Pöltl et al. 2023).
Formula feeding of premature infants increases the likelihood that these neonates will develop NEC. A recent elegant and comprehensive study in mice has shown the complex interplay involving formula feeding, the microbiota, viral infection and immunophenotypes (Subramanian et al. 2024). Formula feeding predisposes conventionally raised mouse pups to NEC in the small intestine upon infection with murine norovirus, supported by activation of TLR3, TLR8 and TLR9 but not TLR4 (known to be activated by bacterial lipopolysaccharide) (Subramanian et al. 2024). Colonisation of the murine small intestine with Klebsiella spp. is enhanced by formula feeding, with significantly increased levels of TV detected in the lumen of the small intestine of formula-fed mouse pups compared with their breast-fed counterparts. Colonisation specifically with TV-producing K. oxytoca makes mouse pups vulnerable to NEC-like tissue damage in the small intestine. Formula feeding downregulates expression of Nlrc5 in the murine small intestine, with this downregulation maintained over time compared with breast-fed animals. Human NEC tissues also exhibit a decrease in NLRC5 expression in intestinal epithelial cells. The gene product of Nlrc5 is a key regulator of processes mediating tissue resistance to natural-killer-cell cytotoxicity. A downstream target of NLRC5 is Tap1, involved in protecting against natural-killer-cell-associated injury of intestinal epithelial cells. High expression of Nlrc5 and Tap1 in intestinal epithelial cells is seen in breast-fed mouse pups, whereas this immunophenotype is suppressed in formula-fed animals and is kept in this state by TV. The formula-feeding-associated immunophenotype (NLRC5LOW, TAP1LOW) results in mice that are predisposed to gut tissue damage upon inflammation caused by enteric viruses. Luminal TV activates peroxisome-proliferator activator receptor (PPAR-γ) signalling in intestinal epithelial cells, inhibiting lamina propria leucocyte-derived IFN-γ-mediated induction of the signal transducer and activator of transcription 1 (STAT1) signalling cascade in intestinal epithelial cells required for NLRC5 expression, leading to the NLRC5LOW, TAP1LOW immunophenotype. Blocking the TV and PPAR-γ interaction may be a therapeutic target to prevent the development of NEC. Microbiota colonisation concomitant with breast feeding promotes intestinal epithelial cells to mature from the low to the high immunophenotype after birth, protecting infants from virus-evoked NEC (Subramanian et al. 2024).
Given our increasing knowledge on the deleterious effects of TV in the mammalian gut, the high occurrence of toxigenic KoSC in the faecal microbiota of healthy infants warrants consideration in future disease association (and exposome) studies (Greimel et al. 2022).
Pyrazinone and pyrazine autoinducers
Novel pyrazinone and pyrazine autoinducers have recently been characterised in K. michiganensis using a combination of ‘one strain many compounds’ and host sensing approaches (Hamchand et al. 2024). Leupeptin is a protease inhibitor first identified as prevalent in clinical lung isolates of K. michiganensis (Li et al. 2020). In K. michiganensis ATCC 8724 (=KCTC 1686), leupeptin is encoded by a 7.2 kbp BGC (leup; Fig. 4C). Its expression is selectively activated by the human mucin-capping sugar N-acetylneuraminic acid, leading to the production of an Arg–Leu-derived pyrazinone analogous to the quorum-sensing signal molecules dimethylpyrazine-2-ol (DPO) and autoinducer-3 (AI-3), and activation of a pyrazine pathway via an aminoketone species that resembles precursors of DPO and AI-3. Leupeptin expression is also weakly induced by porcine mucin. DPO and AI-3 contribute to regulation of biofilm formation and virulence in vibrio cholerae and enterohaemorrhagic E. coli, respectively. However, in K. michiganensis, the leupeptin pyrazinone upregulates iron acquisition (enterobactin and yersiniabactin) and metabolic pathways. That the pyrazinones contribute to regulation of the virulence factor yersiniabactin implies these autoinducers are involved in the pathogenesis of K. michiganensis. Leupeptin pyrazinone is a selective agonist of the mammalian histidine GPCR HRH4. HRH4 is associated with proinflammatory responses, allergic disorders and asthma (Hamchand et al. 2024) (Fig. 3), consistent with the finding that the leupeptin BGC is more prevalent in respiratory K. michiganensis isolates than intestinal isolates of the KoSC (Li et al. 2020). Our own genome-based analyses support the almost exclusive association of the leup BGC with K. michiganensis, with 90.6% of all genomes screened encoding the complete operon (Fig. 5, Supplementary Fig. 1 (see section on Supplementary materials given at the end of the article)). Whether the leup BGC is functional in non-K. michiganensis species of the KoSC remains to be determined. The pyrazine (pyr; Fig. 4C) pathway promotes upregulation of iron uptake (including enterobactin) in K. michiganensis ATCC 8724, and activation of this pathway is mediated via simple (e.g. d-galactose) to complex sugars. Components of the pyrazine biosynthetic machinery appear to be conserved in K. pneumoniae (shown to be functional), K. variicola and K. aerogenes, suggesting this family of autoinducers may contribute to biological processes and pathogenesis of other Klebsiella spp.
Prevalence of the complete leupeptin (leup) BGC across the KoSC. PhyloPhlAn v3.1.68 (Asnicar et al. 2020) was used to generate a phylogenetic tree (low diversity; rooted at the midpoint) from the Bakta-annotated proteomes of 2,716 genome sequences available from AllTheBacteria v1 (Hunt et al. 2024): K. michiganensis, n = 1,298 (1176 BGC-positive, 90.6%); K. oxytoca, n = 865; K. grimontii, n = 443 (4 BGC-positive, 0.9%); K. pasteurii, n = 82 (2 BGC-positive, 2.4%); K. spallanzanii, n = 12; Klebsiella taxon 3, n = 7 (4 BGC-positive, 71.4%); Klebsiella taxon 1, n = 5; K. huaxiensis, n = 4. Only genomes encoding contiguous transporter–LeupA–LeupB–LeupC–LeupD sequences were annotated as having a complete leup BGC (identified by using BLASTP searches against the reference sequence used in Fig. 4C). The inner coloured circle represents Klebsiella species, while the outer circle shows those genomes that encode the complete leup BGC (refer to colour legends). Subsequent phylogenetic analysis of concatenated transporter–LeupA–LeupB–LeupC–LeupD sequences showed species-specific clustering of the leup BGC (Supplementary Fig. 1). Supplementary materials associated with this figure are available from https://doi:10.6084/m9.figshare.27922128.
Citation: Microbiota and Host 3, 1; 10.1530/MAH-24-0011
N-acyl amides
Cohen et al. (2017) mined Human Microbiome Project genomic data to identify bacteria that encoded members of the N-acyl synthase protein family (Pfam PF13444), responsible for producing lipid-signalling molecules and neurotransmitters implicated in a range of biological processes (Cohen et al. 2015). Most of their N-acyl synthase hits were associated with the phylum Pseudomonadota. The gastrointestinal microbiota was enriched for these synthases compared with other body sites, with the GPCRs targeted by the encoded N-acyl amides predominating in the gastrointestinal tract and its associated immune cells. Cohen et al. (2017) identified six different types of N-acyl amides could be produced by gut bacteria: N-acyl glycine, N-acyloxyacyl lysine, N-acyloxyacyl glutamine, N-acyl lysine/ornithine, N-acyl alanine and N-acyl serinol. Among their genomic data, two genomes of K. pneumoniae (gut, oral) and one of K. michiganensis (oral) were predicted to encode for N-acyloxyacyl lysine (potentially activating S1PR4), while a genome of K. variicola (gut) was predicted to encode for N-acyloxyacyl glutamine (shown to inhibit both PTGIR and PTGER4). Attempts were made to confirm production of N-acyloxyacyl glutamine by K. pneumoniae WGLW1-5 under aerobic conditions, but the N-acyl metabolite could not be identified in the culture broth extracts from this strain, though the metabolite was detected after heterologous expression of a synthesised related gene (sourced from K. variicola 342) in E. coli (Cohen et al. 2017). Production of N-acyloxyacyl lysine was confirmed by expression of a synthetic gene from K. ornithinolytica 10-5246 in E. coli (Cohen et al. 2017). A recent search (February 2025) of Pfam shows that 397/18806 sequences containing the acetyltransferase (GNAT) domain (Pfam PF13444) used in the analyses of Cohen et al. (2017) are drawn from the Enterobacteriaceae, with Klebsiella spp. representing 186 of these (unreviewed AlphaFold predictions covering 20 different species of Klebsiella). The conditions under which these genes are expressed in strains harbouring them is unknown. Activation of S1PR4 (sphingosine-1-phosphate receptor 4) has been linked to cancers, dermatitis and immune deficiency disease (Wang et al. 2014), while PTGIR (prostaglandin I2 receptor) is associated with vasodilation and inhibition of platelet aggregation, and PTGER4 (prostaglandin E receptor 4) can activate T-cell factor signalling. The activation of these three and other GPCRs by microbially produced N-acyl amides in disease contexts has not been examined to date, but such studies may provide novel insights into metabolite-associated virulence factors of Klebsiella spp. Co-localisation studies to determine the distribution of GPCRs in the body and proximity to Klebsiella N-acyl synthase gene expression are also warranted. This may provide insights as to how Klebsiella-produced metabolites act as agonists or antagonists in infectious or metabolic diseases.
Future perspectives
Most of our recent knowledge on Klebsiella spp. is from genotypic data derived from clinical isolates. As such, much of the literature focuses on predictions on AMR or virulence genes, often with little to no accompanying phenotypic data. Few data are available on the genotypic and phenotypic traits of commensal Klebsiella spp., though it is clear from the literature covered in this review that bioactive molecules produced by these bacteria contribute to host–microbiota interactions. It is entirely possible that, in some instances, commensal Klebsiella spp. may confer beneficial traits on their mammalian hosts. As such, we need to study commensal members of the genus Klebsiella as well as those associated with disease.
The increasing availability of curated metabolic models for K. pneumoniae (Hawkey et al. 2022) is likely to contribute to a more holistic consideration of how Klebsiella-associated metabolites contribute to host–microbiota interactions in the coming years. Much of our knowledge on secondary metabolites encoded by Klebsiella spp. is related to the KoSC. Continuing improvements to tools available for identifying BGCs in genomic data will lead to greater appreciation of secondary metabolites produced by KpSC bacteria, K. aerogenes and members of the K. terrigena and K. planticola complexes.
Recently, a new class of genotoxic metabolite (the indolimines) was identified, produced by the gut bacterium Morganella morganii (Cao et al. 2022). M. morganii produces several genotoxic indolimines in vivo, which exacerbate CRC in mice. Indolimine(s) in spent culture medium were found to induce DNA double-strand breaks. The genes encoding these newly identified metabolites remain to be identified, but it is thought the indolimines are synthesised from indole-3-aldehyde and condensation with a primary amine via an amino acid decarboxylase (Cao et al. 2022). Indolimine-200, indolimine-214 and indolimine-248 are ligands for the aryl hydrocarbon receptor (AHR), activating it in a dose-dependent manner (Patel et al. 2023). By activating the AHR, indolimines are proposed to mediate both carcinogen metabolism and inflammatory signalling within the tumour microenvironment via a non-genotoxic mechanism. Given the indolimines are indole-related metabolites, as is TV, it is entirely possible these or analogues of indolimines could be encoded by Klebsiella spp. To date, no studies have been undertaken examining the interactions of TV with the AHR.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/MAH-24-0011.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was supported by the European Union’s Horizon 2020 research and innovation programme (grant agreement number 874583).
Author contribution statement
ALM and LH researched and wrote the article. LH did all the bioinformatics work presented.
Acknowledgements
This publication reflects only the authors’ view and the European Commission is not responsible for any use that may be made of the information it contains.
References
Altshuller M , He X , MacKrell EJ , et al. 2024 The Fanconi anemia pathway repairs colibactin-induced DNA interstrand cross-links. BioRxiv Prepr Serv Biol 2024.01.30.576698. (https://doi.org/10.1101/2024.01.30.576698)
Antimicrobial Resistance Collaborators 2022 Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet 399 629–655. (https://doi.org/10.1016/S0140-6736(21)02724-0)
Appel TM , Quijano-Martínez N , De La Cadena E , et al. 2021 Microbiological and clinical aspects of Raoultella spp. Front Public Health 9 686789. (https://doi.org/10.3389/fpubh.2021.686789)
Asnicar F , Thomas AM , Beghini F , et al. 2020 Precise phylogenetic analysis of microbial isolates and genomes from metagenomes using PhyloPhlAn 3.0. Nat Commun 11 2500. (https://doi.org/10.1038/s41467-020-16366-7)
Auvray F , Perrat A , Arimizu Y , et al. 2021 Insights into the acquisition of the pks island and production of colibactin in the Escherichia coli population. Microb Genomics 7 000579. (https://doi.org/10.1099/mgen.0.000579)
Banafea GH , Bakhashab S , Alshaibi HF , et al. 2022 The role of human mast cells in allergy and asthma. Bioengineered 13 7049–7064. (https://doi.org/10.1080/21655979.2022.2044278)
Blin K , Shaw S , Augustijn HE , et al. 2023 antiSMASH 7.0: new and improved predictions for detection, regulation, chemical structures and visualisation. Nucleic Acids Res 51 W46–W50. (https://doi.org/10.1093/nar/gkad344)
Bossuet-Greif N , Vignard J , Taieb F , et al. 2018 The colibactin genotoxin generates DNA interstrand cross-links in infected cells. mBio 9 e02393-17. (https://doi.org/10.1128/mBio.02393-17)
Bray AS & Zafar MA 2024 Deciphering the gastrointestinal carriage of Klebsiella pneumoniae. Infect Immun 92 e0048223. (https://doi.org/10.1128/iai.00482-23)
Cao Y , Oh J , Xue M , et al. 2022 Commensal microbiota from patients with inflammatory bowel disease produce genotoxic metabolites. Science 378 eabm3233. (https://doi.org/10.1126/science.abm3233)
Chen Y-T , Lai Y-C , Tan M-C , et al. 2017 Prevalence and characteristics of pks genotoxin gene cluster-positive clinical Klebsiella pneumoniae isolates in Taiwan. Sci Rep 7 43120. (https://doi.org/10.1038/srep43120)
Chen X , Zhang Z , Li H , et al. 2020a Endogenous ethanol produced by intestinal bacteria induces mitochondrial dysfunction in non-alcoholic fatty liver disease. J Gastroenterol Hepatol 35 2009–2019. (https://doi.org/10.1111/jgh.15027)
Chen Y , Brook TC , Soe CZ , et al. 2020b Preterm infants harbour diverse Klebsiella populations, including atypical species that encode and produce an array of antimicrobial resistance- and virulence-associated factors. Microb Genomics 6 e000377. (https://doi.org/10.1099/mgen.0.000377)
Cohen LJ , Kang H-S , Chu J , et al. 2015 Functional metagenomic discovery of bacterial effectors in the human microbiome and isolation of commendamide, a GPCR G2A/132 agonist. Proc Natl Acad Sci U S A 112 E4825–E4834. (https://doi.org/10.1073/pnas.1508737112)
Cohen LJ , Esterhazy D , Kim S-H , et al. 2017 Commensal bacteria make GPCR ligands that mimic human signalling molecules. Nature 549 48–53. (https://doi.org/10.1038/nature23874)
Conlan S , Deming C , Tsai Y-C , et al. 2014 Complete genome sequence of a Klebsiella pneumoniae isolate with chromosomally encoded carbapenem resistance and colibactin synthesis loci. Genome Announc 2 e01332-14. (https://doi.org/10.1128/genomeA.01332-14)
Cosic A , Leitner E , Petternel C , et al. 2021 Variation in accessory genes within the Klebsiella oxytoca species complex delineates monophyletic members and simplifies coherent genotyping. Front Microbiol 12 692453. (https://doi.org/10.3389/fmicb.2021.692453)
Cruz MADla , Valdez-Salazar HA , Robles-Leyva N , et al. 2024 The transcriptional regulator Lrp activates the expression of genes involved in tilivalline enterotoxin biosynthesis in Klebsiella oxytoca. BioRxiv Prepr Serv Biol 2024.02.29.582825. (https://doi.org/10.1101/2024.02.29.582825)
De Palma G , Shimbori C , Reed DE , et al. 2022 Histamine production by the gut microbiota induces visceral hyperalgesia through histamine 4 receptor signaling in mice. Sci Transl Med 14 eabj1895. (https://doi.org/10.1126/scitranslmed.abj1895)
Dong N , Yang X , Chan EW-C , et al. 2022 Klebsiella species: taxonomy, hypervirulence and multidrug resistance. EBioMedicine 79 103998. (https://doi.org/10.1016/j.ebiom.2022.103998)
Dornisch E , Pletz J , Glabonjat RA , et al. 2017 Biosynthesis of the enterotoxic pyrrolobenzodiazepine natural product tilivalline. Angew Chem Int Ed Engl 56 14753–14757. (https://doi.org/10.1002/anie.201707737)
Eisenhofer R , Alberdi A & Woodcroft BJ 2024 Large-scale estimation of bacterial and archaeal DNA prevalence in metagenomes reveals biome-specific patterns. (https://doi.org/10.1101/2024.05.16.594470)
El-Ashry AH , Hendawy SR & Mahmoud NM 2022 Prevalence of pks genotoxin among hospital-acquired Klebsiella pneumoniae. AIMS Microbiol 8 73–82. (https://doi.org/10.3934/microbiol.2022007)
Elmassry MM , Sugihara K , Chankhamjon P , et al. 2024 A meta-analysis of the gut microbiome in inflammatory bowel disease patients identifies disease-associated small molecules. bioRxiv 2024.02.07.579278. (https://doi.org/10.1101/2024.02.07.579278)
Engevik KA , Hazzard A , Puckett B , et al. 2024 Phylogenetically diverse bacterial species produce histamine. Syst Appl Microbiol 47 126539. (https://doi.org/10.1016/j.syapm.2024.126539)
Faïs T , Delmas J , Barnich N , et al. 2018 Colibactin: more than a new bacterial toxin. Toxins 10 151. (https://doi.org/10.3390/toxins10040151)
Feng Y , Yang Y , Hu Y , et al. 2024 Population genomics uncovers global distribution, antimicrobial resistance, and virulence genes of the opportunistic pathogen Klebsiella aerogenes. Cell Rep 43 114602. (https://doi.org/10.1016/j.celrep.2024.114602)
Glabonjat RA , Kitsera M , Unterhauser K , et al. 2021 Simultaneous quantification of enterotoxins tilimycin and tilivalline in biological matrices using HPLC high resolution ESMS2 based on isotopically 15N-labeled internal standards. Talanta 222 121677. (https://doi.org/10.1016/j.talanta.2020.121677)
Greimel TM , Stampfer L , Leitner E , et al. 2022 Toxin-producing Klebsiella oxytoca in healthy infants: commensal or pathobiont? J Pediatr Gastroenterol Nutr 74 e1–e7. (https://doi.org/10.1097/MPG.0000000000003299)
Hamchand R , Wang K , Song D , et al. 2024 Mucosal sugars delineate pyrazine vs pyrazinone autoinducer signaling in Klebsiella oxytoca. Nat Commun 15 8902. (https://doi.org/10.1038/s41467-024-53185-6)
Hawkey J , Vezina B , Monk JM , et al. 2022 A curated collection of Klebsiella metabolic models reveals variable substrate usage and gene essentiality. Genome Res 32 1004–1014. (https://doi.org/10.1101/gr.276289.121)
Högenauer C , Langner C , Beubler E , et al. 2006 Klebsiella oxytoca as a causative organism of antibiotic-associated hemorrhagic colitis. N Engl J Med 355 2418–2426. (https://doi.org/10.1056/NEJMoa054765)
Hoyles L 2024 Gut microbiome and hepatic steatosis (steatotic liver disease). In Gut Microbiome, Microbial Metabolites and Cardiometabolic Risk, pp 177–197. Eds M Federici & R Menghini . Cham: Springer International Publishing. (https://doi.org/10.1007/978-3-031-35064-1_7)
Hunt M , Lima L , Shen W , et al. 2024 AllTheBacteria – all bacterial genomes assembled, available and searchable. BioRxiv Prepr Serv Biol 2024.03.08.584059. (https://doi.org/10.1101/2024.03.08.584059)
Jati AP , Sola-Campoy PJ , Bosch T , et al. 2023 Widespread detection of yersiniabactin gene cluster and its encoding integrative conjugative elements (ICEKp) among nonoutbreak OXA-48-producing Klebsiella pneumoniae clinical isolates from Spain and The Netherlands. Microbiol Spectr 11 e0471622. (https://doi.org/10.1128/spectrum.04716-22)
Joainig MM , Gorkiewicz G , Leitner E , et al. 2010 Cytotoxic effects of Klebsiella oxytoca strains isolated from patients with antibiotic-associated hemorrhagic colitis or other diseases caused by infections and from healthy subjects. J Clin Microbiol 48 817–824. (https://doi.org/10.1128/JCM.01741-09)
Kanki M , Yoda T , Tsukamoto T , et al. 2002 Klebsiella pneumoniae produces no histamine: Raoultella planticola and Raoultella ornithinolytica strains are histamine producers. Appl Environ Microbiol 68 3462–3466. (https://doi.org/10.1128/AEM.68.7.3462-3466.2002)
Kaur CP , Iyadorai T , Sears C , et al. 2023 Presence of polyketide synthase (PKS) gene and counterpart virulence determinants in Klebsiella pneumoniae strains enhances colorectal cancer progression in-vitro. Microorganisms 11 443. (https://doi.org/10.3390/microorganisms11020443)
Khaertynov KS , Anokhin VA , Rizvanov AA , et al. 2018 Virulence factors and antibiotic resistance of Klebsiella pneumoniae strains isolated from neonates with sepsis. Front Med 5 225. (https://doi.org/10.3389/fmed.2018.00225)
Kienesberger S , Cosic A , Kitsera M , et al. 2022 Enterotoxin tilimycin from gut-resident Klebsiella promotes mutational evolution and antibiotic resistance in mice. Nat Microbiol 7 1834–1848. (https://doi.org/10.1038/s41564-022-01260-3)
Lai Y-C , Lin A-C , Chiang M-K , et al. 2014 Genotoxic Klebsiella pneumoniae in Taiwan. PLoS One 9 e96292. (https://doi.org/10.1371/journal.pone.0096292)
Lan Y , Zhou M , Jian Z , et al. 2019 Prevalence of pks gene cluster and characteristics of Klebsiella pneumoniae-induced bloodstream infections. J Clin Lab Anal 33 e22838. (https://doi.org/10.1002/jcla.22838)
Ledala N , Malik M , Rezaul K , et al. 2022 Bacterial indole as a multifunctional regulator of Klebsiella oxytoca complex enterotoxicity. mBio 13 e0375221. (https://doi.org/10.1128/mbio.03752-21)
Leitner E , Bozic M , Kienesberger S , et al. 2022 Improved diagnosis of antibiotic-associated haemorrhagic colitis (AAHC) in faecal specimens by a new qualitative real-time PCR assay detecting relevant toxin genes of Klebsiella oxytoca sensu lato. Clin Microbiol Infect 28 690–694. (https://doi.org/10.1016/j.cmi.2021.09.017)
Li J-H , Oh J , Kienesberger S , et al. 2020 Making and breaking leupeptin protease inhibitors in pathogenic Gammaproteobacteria. Angew Chem Int Ed Engl 59 17872–17880. (https://doi.org/10.1002/anie.202005506)
Li N-N , Li W , Feng J-X , et al. 2021 High alcohol-producing Klebsiella pneumoniae causes fatty liver disease through 2,3-butanediol fermentation pathway in vivo. Gut Microbes 13 1979883. (https://doi.org/10.1080/19490976.2021.1979883)
Lu M-C , Chen Y-T , Chiang M-K , et al. 2017 Colibactin contributes to the hypervirulence of pks(+) K1 CC23 Klebsiella pneumoniae in mouse meningitis infections. Front Cell Infect Microbiol 7 103. (https://doi.org/10.3389/fcimb.2017.00103)
Luo C , Chen Y , Hu X , et al. 2023 Genetic and functional analysis of the pks gene in clinical Klebsiella pneumoniae isolates. Microbiol Spectr 11 e0017423. (https://doi.org/10.1128/spectrum.00174-23)
Ma Y , Wu X , Li S , et al. 2021 Proposal for reunification of the genus Raoultella with the genus Klebsiella and reclassification of Raoultella electrica as Klebsiella electrica comb. Nov Res Microbiol 172 103851. (https://doi.org/10.1016/j.resmic.2021.103851)
McCartney AL & Hoyles L 2023 The role of Klebsiella populations in preterm infants. Biochem Soc Trans 51 887–896. (https://doi.org/10.1042/BST20200325)
Moradigaravand D , Martin V , Peacock SJ , et al. 2017 Population structure of multidrug-resistant Klebsiella oxytoca within hospitals across the United Kingdom and Ireland identifies sharing of virulence and resistance genes with K. pneumoniae. Genome Biol Evol 9 574–584. (https://doi.org/10.1093/gbe/evx019)
Morgan RN , Saleh SE , Farrag HA , et al. 2019 Prevalence and pathologic effects of colibactin and cytotoxic necrotizing factor-1 (Cnf 1) in Escherichia coli: experimental and bioinformatics analyses. Gut Pathog 11 22. (https://doi.org/10.1186/s13099-019-0304-y)
Nougayrède JP , Homburg S , Taieb F , et al. 2006 Escherichia coli induces DNA double-strand breaks in eukaryotic cells. Science 313 848–851. (https://doi.org/10.1126/science.1127059)
Nougayrède JP , Chagneau CV , Motta J-P , et al. 2021 A toxic friend: genotoxic and mutagenic activity of the probiotic strain Escherichia coli Nissle 1917. mSphere 6 e0062421. (https://doi.org/10.1128/mSphere.00624-21)
Patel D , Murray IA , Dong F , et al. 2023 Induction of AHR signaling in response to the indolimine class of microbial stress metabolites. Metabolites 13 985. (https://doi.org/10.3390/metabo13090985)
Paveglio S , Ledala N , Rezaul K , et al. 2020 Cytotoxin-producing Klebsiella oxytoca in the preterm gut and its association with necrotizing enterocolitis. Emerg Microbes Infect 9 1321–1329. (https://doi.org/10.1080/22221751.2020.1773743)
Pöltl L , Kitsera M , Raffl S , et al. 2023 Microbiota-derived genotoxin tilimycin generates colonic stem cell mutations. Cell Rep 42 112199. (https://doi.org/10.1016/j.celrep.2023.112199)
Ponce-Alonso M , Rodríguez-Rojas L , Del Campo R , et al. 2016 Comparison of different methods for identification of species of the genus Raoultella: report of 11 cases of Raoultella causing bacteraemia and literature review. Clin Microbiol Infect 22 252–257. (https://doi.org/10.1016/j.cmi.2015.10.035)
Pope JL , Yang Y , Newsome RC , et al. 2019 Microbial colonization coordinates the pathogenesis of a Klebsiella pneumoniae infant isolate. Sci Rep 9 3380. (https://doi.org/10.1038/s41598-019-39887-8)
Putze J , Hennequin C , Nougayrède JP , et al. 2009 Genetic structure and distribution of the colibactin genomic island among members of the family Enterobacteriaceae. Infect Immun 77 4696–4703. (https://doi.org/10.1128/IAI.00522-09)
Rachmawati N , Powell SM , Triwibowo R , et al. 2022 Modelling growth and histamine formation of Klebsiella aerogenes TI24 isolated from Indonesian pindang. Int J Food Microbiol 362 109459. (https://doi.org/10.1016/j.ijfoodmicro.2021.109459)
Rodríguez-Valverde D , León-Montes N , Soria-Bustos J , et al. 2021 cAMP receptor protein positively regulates the expression of genes involved in the biosynthesis of Klebsiella oxytoca tilivalline cytotoxin. Front Microbiol 12 743594. (https://doi.org/10.3389/fmicb.2021.743594)
Sá-Pessoa J , López-Montesino S , Przybyszewska K , et al. 2023 A trans-kingdom T6SS effector induces the fragmentation of the mitochondrial network and activates innate immune receptor NLRX1 to promote infection. Nat Commun 14 871. (https://doi.org/10.1038/s41467-023-36629-3)
Schneditz G , Rentner J , Roier S , et al. 2014 Enterotoxicity of a nonribosomal peptide causes antibiotic-associated colitis. Proc Natl Acad Sci U S A 111 13181–13186. (https://doi.org/10.1073/pnas.1403274111)
Shao Y , Forster SC , Tsaliki E , et al. 2019 Stunted microbiota and opportunistic pathogen colonization in caesarean-section birth. Nature 574 117–121. (https://doi.org/10.1038/s41586-019-1560-1)
Shi Q , Quan J , Lan P , et al. 2020 Prevalence and characteristics of pks gene cluster harbouring Klebsiella pneumoniae from bloodstream infection in China. Epidemiol Infect 148 e69. (https://doi.org/10.1017/S0950268820000655)
Shibu P , McCuaig F , McCartney AL , et al. 2021 Improved molecular characterization of the Klebsiella oxytoca complex reveals the prevalence of the kleboxymycin biosynthetic gene cluster. Microb Genomics 7 000592. (https://doi.org/10.1099/mgen.0.000592)
Storey D , McNally A , Åstrand M , et al. 2020 Klebsiella pneumoniae type VI secretion system-mediated microbial competition is PhoPQ controlled and reactive oxygen species dependent. PLoS Pathog 16 e1007969. (https://doi.org/10.1371/journal.ppat.1007969)
Strakova N , Korena K & Karpiskova R 2021 Klebsiella pneumoniae producing bacterial toxin colibactin as a risk of colorectal cancer development – a systematic review. Toxicon 197 126–135. (https://doi.org/10.1016/j.toxicon.2021.04.007)
Subramanian S , Geng H , Wu L , et al. 2024 Microbiota regulates neonatal disease tolerance to virus-evoked necrotizing enterocolitis by shaping the STAT1-NLRC5 axis in the intestinal epithelium. Cell Host Microbe 32 1805–1821.e10. (https://doi.org/10.1016/j.chom.2024.08.013)
Tarracchini C , Fontana F , Mancabelli L , et al. 2022 Gut microbe metabolism of small molecules supports human development across the early stages of life. Front Microbiol 13 1006721. (https://doi.org/10.3389/fmicb.2022.1006721)
Thurmond RL , Gelfand EW & Dunford PJ 2008 The role of histamine H1 and H4 receptors in allergic inflammation: the search for new antihistamines. Nat Rev Drug Discov 7 41–53. (https://doi.org/10.1038/nrd2465)
Tse H , Gu Q , Sze K-H , et al. 2017 A tricyclic pyrrolobenzodiazepine produced by Klebsiella oxytoca is associated with cytotoxicity in antibiotic-associated hemorrhagic colitis. J Biol Chem 292 19503–19520. (https://doi.org/10.1074/jbc.M117.791558)
Turocy T & Crawford JM 2025 Bacterial small molecule metabolites implicated in gastrointestinal cancer development. Nat Rev Microbiol 23 106–121. (https://doi.org/10.1038/s41579-024-01103-4)
Unterhauser K , Pöltl L , Schneditz G , et al. 2019 Klebsiella oxytoca enterotoxins tilimycin and tilivalline have distinct host DNA-damaging and microtubule-stabilizing activities. Proc Natl Acad Sci U S A 116 3774–3783. (https://doi.org/10.1073/pnas.1819154116)
Vanuytsel T , Bercik P & Boeckxstaens G 2023 Understanding neuroimmune interactions in disorders of gut-brain interaction: from functional to immune-mediated disorders. Gut 72 787–798. (https://doi.org/10.1136/gutjnl-2020-320633)
von Tesmar A , Hoffmann M , Abou Fayad A , et al. 2018 Biosynthesis of the Klebsiella oxytoca pathogenicity factor tilivalline: heterologous expression, in vitro biosynthesis, and inhibitor development. ACS Chem Biol 13 812–819. (https://doi.org/10.1021/acschembio.7b00990)
Wami H , Wallenstein A , Sauer D , et al. 2021 Insights into evolution and coexistence of the colibactin- and yersiniabactin secondary metabolite determinants in enterobacterial populations. Microb Genomics 7 000577. (https://doi.org/10.1099/mgen.0.000577)
Wang C , Mao J , Redfield S , et al. 2014 Systemic distribution, subcellular localization and differential expression of sphingosine-1-phosphate receptors in benign and malignant human tissues. Exp Mol Pathol 97 259–265. (https://doi.org/10.1016/j.yexmp.2014.07.013)
Wang Z , Liu Y , Liu P , et al. 2023 Genomic and clinical characterization of Klebsiella pneumoniae carrying the pks island. Front Microbiol 14 1189120. (https://doi.org/10.3389/fmicb.2023.1189120)
Woodcroft BJ , Aroney STN , Zhao R , et al. 2024 SingleM and sandpiper: robust microbial taxonomic profiles from metagenomic data. (https://doi.org/10.1101/2024.01.30.578060)
Wyres KL , Lam MMC & Holt KE 2020 Population genomics of Klebsiella pneumoniae. Nat Rev Microbiol 18 344–359. (https://doi.org/10.1038/s41579-019-0315-1)
Yang J , Long H , Hu Y , et al. 2022 Klebsiella oxytoca complex: update on taxonomy, antimicrobial resistance, and virulence. Clin Microbiol Rev 35 e0000621. (https://doi.org/10.1128/CMR.00006-21)
Yuan J , Chen C , Cui J , et al. 2019 Fatty liver disease caused by high-alcohol-producing Klebsiella pneumoniae. Cell Metab 30 675–688.e7. (https://doi.org/10.1016/j.cmet.2019.08.018)
Zechner EL & Kienesberger S 2024 Microbiota-derived small molecule genotoxins: host interactions and ecological impact in the gut ecosystem. Gut Microbes 16 2430423. (https://doi.org/10.1080/19490976.2024.2430423)
