Global genomic profiling of Klebsiella pneumoniae: A spatio-temporal population structure analysis

Klebsiella pneumoniae is an important clinical bacterial pathogen that has hypervirulent and multidrug-resistant variants. Uniform Manifold Approximation and Projection (UMAP) was used to cluster genomes of 16 797 K. pneumoniae strains collected, based on core genome distance, in over 100 countries during the period 1937 to 2021. A total of 60 high-density genetic clusters of strains representing the major epidemic strains were identified among these strains. Using UMAP bedding, the relationship between genetic cluster, capsular polysaccharide (KL) types and sequence type (ST) of the strains was clearly demonstrated, with some important STs, such as ST11 and ST258, found to contain multiple clusters. Strains within the same cluster often exhibited significant diverse features, such as originating from different areas and being isolated in different years, as well as carriage of different resistance and virulence genes. These data enable the routes of evolution of the globally prevalent K. pneumoniae strains to be traced. Alarmingly, carbapenem-resistant K. pneumoniae strains accounted for 51.7% of the test strains and worldwide transmission was observed. Carbapenem-resistant and hypervirulent K. pneumoniae strains are mainly reported in China; however, these strains are increasingly reported in other parts of the world. Also identified in this study were several key genetic loci that facilitate development of a new K. pneumoniae typing method to differentiate between high- and low-risk strains. In particular, the acrR, ompK35 and hha genes were predicted to play a key role in expression of the resistance and virulence phenotypes.