The diversity of Klebsiella pneumoniae surface polysaccharides

Abstract

Klebsiella pneumoniae is considered an urgent health concern due to the emergence of multi-drug-resistant strains for which vaccination offers a potential remedy. Vaccines based on surface polysaccharides are highly promising but need to address the high diversity of surface-exposed polysaccharides, synthesized as O-antigens (lipopolysaccharide, LPS) and K-antigens (capsule polysaccharide, CPS), present in K. pneumoniae. We present a comprehensive and clinically relevant study of the diversity of O- and K-antigen biosynthesis gene clusters across a global collection of over 500 K. pneumoniae whole-genome sequences and the seroepidemiology of human isolates from different infection types. Our study defines the genetic diversity of O- and K-antigen biosynthesis cluster sequences across this collection, identifying sequences for known serotypes as well as identifying novel LPS and CPS gene clusters found in circulating contemporary isolates. Serotypes O1, O2 and O3 were most prevalent in our sample set, accounting for approximately 80 % of all infections. In contrast, K serotypes showed an order of magnitude higher diversity and differ among infection types. In addition we investigated a potential association of O or K serotypes with phylogenetic lineage, infection type and the presence of known virulence genes. K1 and K2 serotypes, which are associated with hypervirulent K. pneumoniae, were associated with a higher abundance of virulence genes and more diverse O serotypes compared to other common K serotypes.

Keywords: K antigen and O antigen; Klebsiella pneumoniae; seroepidemiology; surface polysaccharide; vaccine target.

Fig. 1.

(a) rfb gene clusters (see Fig. S1 , available in the online Supplementary Material for full list), genes are colored according to function (see Fig. S1) and above or below the line according to the coding strand and the direction of transcription (black arrows). (b) WbdD sequence alignment. Comparison of E. coli 09a (Uniprot accession number Q47592) and one representative each for the two types of K. pneumoniae O3 WbdD (O3 long (isolate 5151_3#1) and O3 short (9878_1#12)) and OL104 wbdD (5193_7#2). The alignment was colored according to the BLOSUM62 score. Domain detection was performed using InterProScan5 (Jones et al. 2014) based on the E. coli WbdD sequence.

Fig. 2.

Phylogeny of a randomly selected subset of isolates. Lineages are labeled according to the scheme of (Holt et al. 2015), subdividing the K. pneumoniae species into subspecies KpI (K. pneumoniae), KpII (K. quasipneumoniae) and KpIII (K. variicola). O and K serotypes and LPS core types, as well as the geographical origin, are denoted by colored squares as indicated.

Fig. 3.

Distribution of serotypes in human K. pneumoniae isolates in the global collection. The eight most common O-types (left bars) and ten most common K-types (right bars) are shown. Other serotypes are shown in dark grey.

Fig. 4.

(a,b) Virulence genes in K. pneumoniae serotypes. Frequency of isolates containing (a) siderophore and mucosity and (b) fimbrial adhesin genes for selected serotypes. →, Significantly overrepresented in this serotype vs all other serotypes (Fisher’s exact test, P-value <0.01). ←, Significantly underrepresented. (c) Genomic context of the K1 associated alginate lyase (blue: alginate lyase, green: thiamine pyrophosphate biosynthesis cluster, grey: other).