A genomic surveillance framework and genotyping tool for Klebsiella pneumoniae and its related species complex

Abstract

Klebsiella pneumoniae is a leading cause of antimicrobial-resistant (AMR) healthcare-associated infections, neonatal sepsis and community-acquired liver abscess, and is associated with chronic intestinal diseases. Its diversity and complex population structure pose challenges for analysis and interpretation of K. pneumoniae genome data. Here we introduce Kleborate, a tool for analysing genomes of K. pneumoniae and its associated species complex, which consolidates interrogation of key features of proven clinical importance. Kleborate provides a framework to support genomic surveillance and epidemiology in research, clinical and public health settings. To demonstrate its utility we apply Kleborate to analyse publicly available Klebsiella genomes, including clinical isolates from a pan-European study of carbapenemase-producing Klebsiella, highlighting global trends in AMR and virulence as examples of what could be achieved by applying this genomic framework within more systematic genomic surveillance efforts. We also demonstrate the application of Kleborate to detect and type K. pneumoniae from gut metagenomes.

Fig. 1. Relationships between Kleborate virulence and resistance scores and the prevalence of key virulence and antimicrobial resistance (AMR).

Data shown summarize Kleborate results for non-redundant set of 9705 publicly available K. pneumoniae genomes (Supplementary Data 2). a Barplot shows mean virulence score per year (right y-axis), line plots show the prevalence of individual virulence loci per year (left y-axis). Ybt yersiniabactin, clb colibactin, iuc aerobactin, iro salmochelin, rmpADC hypermucoidy rmp locus, rmpA2 rmpA2 gene. Spearman correlation coefficients (r_s) between the mean virulence score and prevalence of each locus are noted. P values from two-sided statistical testing: ybt = 2.2 × 10⁻¹⁶, clb = 0.74, iuc = 0.0002, iro = 0.68, rmpADC = 0.005, rmpA2 = 0.002. b Barplot shows mean resistance score per year (right y-axis) and line plots show the prevalence of carbapenemases (carb), acquired colistin resistance genes (col gene), mutations in MgrB/PmrB (col mut) and genes conferring resistance to extended-spectrum β-lactams (ESBL) (left y-axis). Spearman correlation coefficients (r_s) between mean resistance score and prevalence of each resistance type are noted. P values: carb = 8.3 × 10⁻⁶, col gene = 0.06, col mut = 0.05, ESBL = 5.56 × 10⁻⁵. c Mean number of acquired AMR genes and classes, over time. Spearman correlation coefficients with mean resistance score are noted. P values: AMR genes = 0.001, AMR classes = 2.2 × 10⁻¹⁶. d Histograms showing total number of acquired AMR classes predicted per genome, stratified by resistance score. e Histograms showing a total number of acquired AMR genes detected per genome, stratified by resistance score. Spearman correlation coefficients are shown in a-c; significance levels are indicated with asterisks: *p < 0.01, **p < 0.001.

Fig. 2. Kleborate genotyping results for European K. pneumoniae surveillance isolates.

Data shown summarize Kleborate results for 927 carbapenem-non-susceptible and 697 carbapenem-susceptible K. pneumoniae genomes from the EuSCAPE study (data included in Supplementary Data 2). a Geographical and lineage distribution of carbapenemase genes. Each circle represents a genome, coloured by carbapenemase (see inset legend). Barplots summarize the number of genomes from each K. pneumoniae lineage (top) and country (right), coloured by carbapenemase. b–c Cumulative prevalence of (b) capsule (K) locus and (c) O antigen locus types, for carbapenem-non-susceptible (meropenem MIC > 2) isolates, ordered by overall prevalence. Thick line indicates curve for whole data set; others give results separately for different United Nations geographical regions (see inset legend).

Fig. 3. Distribution of meropenem MIC, stratified by Kleborate-detected carbapenemase genes and OmpK35/36 porin mutations, for European K. pneumoniae surveillance isolates.

a Data shown summarize Kleborate results for 1490 K. pneumoniae genomes from the EuSCAPE study (data included in Supplementary Data 2). Each circle represents the reported meropenem MIC for an isolate, coloured by type of porin mutation/s identified by Kleborate from the corresponding genome assembly (colour key in inset legend, the prevalence of each genotype across 1490 genomes is indicated in brackets). Isolates are stratified by carbapenemase gene (enzymes labelled on x-axis) and OmpK mutations^, reported by Kleborate. Wt, full-length OmpK35 and OmpK36 with no GD/TD insertion in the OmpK36 β-strand loop; mut, otherwise; Δ, missing/truncated. Dashed lines indicate EUCAST breakpoints for clinical resistance (red, MIC > 8) and non-susceptibility (black, MIC > 2). For each boxplot, the length of the box corresponds to the interquartile range with the centre line corresponding to the median, and the whiskers represent the minimum and maximum values. b Mean meropenem MIC values for the 1490 EuSCAPE isolates, grouped by the combination of porin gene status and the presence of carbapenemase genes. Porin status is expressed as: Δ, missing/truncated; loop, GD or TD insertion in the OmpK36 β-strand loop; wt, otherwise.

Fig. 4. Summary of genome collection metadata, and Kleborate-derived virulence and antimicrobial resistance (AMR) genotypes, for all publicly available Klebsiella genomes.

Data shown summarize Kleborate results for 11,277 non-redundant Klebsiella genomes publicly available as at 17 July 2020 (Supplementary Data 2). From left to right: barplots showing source information by geographical region and sample type (coloured as per inset legend); heatmaps showing the prevalence of virulence loci (blue) and predicted AMR drug classes (red) (as per inset scale bars). Genomes are summarized by species, ordered by species complex: KpSC, K. pneumoniae species complex; KoSC, K. oxytoca species complex; and other Klebsiella. In the heatmaps, the total number of genomes in which each type of virulence/AMR determinant was detected are indicated below each column. Column names are as follows: ybt yersiniabactin, clb colibactin, iuc aerobactin, iro salmochelin, rmp hypermucoidy Rmp, rmpA2 hypermucoidy rmpA2, AGly aminoglycosides, Col colistin, Fcyn fosfomycin, Flq fluoroquinolone, Gly glycopeptide, MLS macrolides, Phe phenicols, Rif rifampin, Sul sulfonamides, Tet tetracyclines, Tgc tigecycline, Tmt trimethoprim, Bla β-lactamases, inhR β-lactamase inhibitor, ESBL extended-spectrum β-lactamases, ESBL_inhR extended-spectrum β-lactamase with resistance to β-lactamase inhibitors, Carb carbapenemase, Bla_chr intrinsic chromosomal β-lactamase, SHV mutations in SHV, Omp truncations/mutations in ompK35/ompK36, Col truncations in mgrB/pmrB conferring colistin resistance, Flq mutations in gyrA/parC conferring resistance to fluoroquinolones.

Fig. 5. Distribution of (a) resistance and (b) virulence scores among genomes belonging to the 30 most common K. pneumoniae lineages.

Data shown summarize Kleborate results for non-redundant set of 9705 publicly available K. pneumoniae genomes (Supplementary Data 2). Lineages were defined on the basis of multi-locus sequence types (STs) reported by Kleborate, and ordered from highest to lowest difference between mean virulence and mean resistance score. Minimum genome count per ST shown is 50. Ybt yersiniabactin, clb colibactin, iuc aerobactin, VP virulence plasmid, ESBL extended-spectrum β-lactamase, Carb carbapenemase, Col colistin resistance determinant.

Fig. 6. Insights from resistance and virulence scores.

Data shown summarize Kleborate results for non-redundant set of 9,705 publicly available K. pneumoniae genomes (Supplementary Data 2). a–b Mean resistance and virulence scores grouped by (a) lineage and (b) sample type. Each circle represents a single lineage (multi-locus sequence type, ST) or sample type as labelled; size indicates the number of genomes (as per inset legend); colour indicates groups per inset legend. c Heatmap showing number of genomes with each combination of resistance and virulence scores. Convergent genomes correspond to a virulence score ≥3 (carrying iuc) and resistance score of ≥1 (carrying ESBL and/or carbapenemase gene/s), as indicated by the red box. d Barplots showing lineage distribution of convergent genomes, for each combination of resistance score and virulence score. Lineages are grouped into hypervirulent (hv), multidrug resistant (MDR) and others; and coloured by ST (as per inset legend).

Fig. 7. Convergence of AMR and virulence determinants in the K. pneumoniae population, identified by Kleborate analysis of public genomes.

a Geographical and lineage distribution of convergence events. Each circle represents a unique convergence event (i.e. a monophyletic clade harbouring both ESBL/carbapenemase genes and iuc; see interactive tree at https://microreact.org/project/JDyan46yctyDh6weEUjWN, summary of events in Supplementary Data 7, assignment of genomes to events in Supplementary Data 2). Circles are scaled by the number of total genomes linked to the event and coloured to indicate whether convergence is inferred to have occurred via (i) acquisition of AMR gene/s (ESBL or carbapenemase/s) by a hypervirulent lineage, (ii) acquisition of an iuc-encoding plasmid by an AMR or non-AMR lineage, or (iii) unresolved means as per inset legend. Marginal barplots show the number of convergence events (colour blocks) and genomes (block heights) associated with each lineage (top) or geographical region (right). Lineages were defined on the basis of multi-locus sequence types (STs), number of convergence events estimated for each is labelled at the top of each bar. b Distribution of convergent genomes by location. Countries from which convergent genomes were detected are coloured on the map; circles represent the number of convergent genomes detected in each UN-defined geographical region (indicated by colour, as per inset legend), scaled and labelled with the minimum estimated number of unique convergence events specific to each region (excluding inter-regional convergence events). The total number of convergence events affecting each region, including region-specific and inter-regional convergence events, are given in brackets in the inset legend.