Rapid single-colony whole-genome sequencing of bacterial pathogens

Abstract

Objectives: As a result of the introduction of rapid benchtop sequencers, the time required to subculture a bacterial pathogen to extract sufficient DNA for library preparation can now exceed the time to sequence said DNA. We have eliminated this rate-limiting step by developing a protocol to generate DNA libraries for whole-genome sequencing directly from single bacterial colonies grown on primary culture plates.

Methods: We developed our protocol using single colonies of 17 bacterial pathogens responsible for severe human infection that were grown using standard diagnostic media and incubation conditions. We then applied this method to four clinical scenarios that currently require time-consuming reference laboratory tests: full identification and genotyping of salmonellae; identification of blaNDM-1, a highly transmissible carbapenemase resistance gene, in Klebsiella pneumoniae; detection of genes encoding staphylococcal toxins associated with specific disease syndromes; and monitoring of vaccine targets to detect vaccine escape in Neisseria meningitidis.

Results: We validated our single-colony whole-genome sequencing protocol for all 40 combinations of pathogen and selective, non-selective or indicator media tested in this study. Moreover, we demonstrated the clinical value of this method compared with current reference laboratory tests.

Conclusions: This advance will facilitate the implementation of whole-genome sequencing into diagnostic and public health microbiology.

Keywords: antibiotic resistance; infectious diseases; typing.

Figure 1.

Unrooted maximum-likelihood tree of the clinical Salmonella isolate in the context of 21 different S. enterica serovars (e.g. Dublin) and S. enterica subsp. arizonae (with Salmonella Enteritidis PT4 as reference). This not only showed that the clinical isolate belonged to the serovar Enteritidis, which matched the reference laboratory results, but simultaneously provided genome-level discrimination between isolates (i.e. the study isolate was only distantly related to the other previously sequenced Enteritidis isolates). Thus single-colony WGS of salmonellae could not only replace serotyping but all currently used epidemiological techniques.

Figure 2.

Heat maps showing the presence (black) or absence (grey) of resistance or virulence determinants. (a) Carbapenemase genes for the clinical K. pneumoniae isolate. (b) Toxome for MRSA reference strain.