Kaptive 2.0: updated capsule and lipopolysaccharide locus typing for the Klebsiella pneumoniae species complex

Abstract

The outer polysaccharide capsule and lipopolysaccharide (LPS) antigens are key targets for novel control strategies targeting Klebsiella pneumoniae and related taxa from the K. pneumoniae species complex (KpSC), including vaccines, phage and monoclonal antibody therapies. Given the importance and growing interest in these highly diverse surface antigens, we had previously developed Kaptive, a tool for rapidly identifying and typing capsule (K) and outer LPS (O) loci from whole genome sequence data. Here, we report two significant updates, now freely available in Kaptive 2.0 (https://github.com/katholt/kaptive): (i) the addition of 16 novel K locus sequences to the K locus reference database following an extensive search of >17 000 KpSC genomes; and (ii) enhanced O locus typing to enable prediction of the clinically relevant O2 antigen (sub)types, for which the genetic determinants have been recently described. We applied Kaptive 2.0 to a curated dataset of >12 000 public KpSC genomes to explore for the first time, to the best of our knowledge, the distribution of predicted O (sub)types across species, sampling niches and clones, which highlighted key differences in the distributions that warrant further investigation. As the uptake of genomic surveillance approaches continues to expand globally, the application of Kaptive 2.0 will generate novel insights essential for the design of effective KpSC control strategies.

Keywords: K locus; Klebsiella capsule K-locus genomic surveillance; Klebsiella pneumoniae; O locus.

Fig. 1.

Process of identification of novel K loci from 17 752 KpSC. Candidate genomes were identified as those with Kaptive K locus confidence calls ‘Low’ or ‘None’, and were iteratively filtered to remove: (i) fragmented or low quality locus sequences including 37 assembled using only Oxford Nanopore Technologies (ONT) data; (ii) true matches to existing K loci; (iii) IS and/or deletion variants of existing loci or other putative novel loci (see Methods for full details).

Fig. 2.

Genetic structure of novel K loci identified in this study. Coding sequences are represented by arrows, labelled by their gene name where applicable, and coloured by homology (sequence identity ≥30 %). Coding sequences predicted to encode hypothetical proteins are represented by arrows with a dashed outline. Shading between the K loci represents regions of similarity between coding sequences as identified by clinker [53], and the level of similarity indicated in the key.

Fig. 3.

Distribution of O loci and predicted O1/O2 (sub)types by species. (a) Heatmap showing the proportion of genomes of each species harbouring each distinct O locus. Total sample sizes for each species are indicated below the x-axis labels. (b) Bar graph showing the number of genomes of each species predicted to express O1 and O2 antigens. Only species for which at least one genome was predicted to express an O1 or O2 antigen are shown (total genomes indicated below the x-axis labels). Stars indicate the position of bars of size 1. Kp, K. pneumoniae ; Kvv, K. variicola subsp. variicola ; Kqs, Klebsiella quasipneumoniae subsp. similipneumoniae ; Kqq, K. quasipneumoniae subsp. quasipneumoniae ; Kvt, K. variicola subsp. tropica ; Ka, Klebsiella africana .

Fig. 4.

Distribution of predicted O types by isolate source. Bars show the proportion of isolates for each of 12 selected sources of interest that were predicted to encode each O antigen (as indicated in the key). The total numbers of isolates representing each source are indicated below the x-axis.

Fig. 5.

Distribution of predicted O types among common K. pneumoniae clones. Bars show the proportion of genomes predicted to express each O (sub)type within each clone (coloured as per the key). Sample sizes are indicated and only clones for which N>50 are shown. Globally distributed multi-drug resistant (MDR) and hypervirulent (Hv) clones are indicated by grey circles as per the key (clones as described previously [1], additionally ST340 and ST437 belong to clonal group 258, ST16 belongs to clonal group 20).