Overview of ECL8 TraDIS data: mapped insertions, insertion index profile and example gene insertion plots.

Insertions are illustrated on the (A) Chromosome and (B) Plasmid, respectively. The outermost track displays the length of the ECL8 genome in base pairs. The subsequent two inner tracks correspond to coding sequences (CDS) on the sense (blue) and antisense (green) DNA strands, respectively. Putative essential CDSs are highlighted in red. The inner-most track (black) corresponds to the location and read frequency of transposon sequences mapped successfully to the K. pneumoniae ECL8 genome. Plot generated using DNAPlotter. (C) Gene insertion index scores (IIS) of the K. pneumoniae ECL8 TraDIS library mapped in order of genomic annotation of the K. pneumoniae ECL8 (left) chromosome and (right) plasmid. Example transposon insertion profiles categorised into essential, non-essential and unclear: (D) an essential gene – bamD, an essential outer membrane factor for β-barrel protein assembly; (E) a non-essential gene – int_A1, a redundant (several copies) integrase required for bacteriophage integration into the host genome; (F) an “unclear” gene – polA an essential gene in prokaryotes required for DNA replication but showed requirement for the N-terminal 5’-3’ exonuclese domain; and (G) an insertion free region suggestive of an unannotated ORF. Transposon insertion sites are illustrated in black and capped at a maximum read depth of 1.

Summary of transposon-containing sequence reads and unique insertion points (UIPs) mapped to the K. pneumoniae ECL8

K. pneumoniae ECL8 plasmid-borne genes computationally deemed essential.

Directional insertion bias of transposon (Tn) into the cps operon.

(A) Schematic representing the Tn orientation and effect on downstream transcription. The utilized Tn transposon is flanked by terminator sequences (purple) is shown inserted in gene A (yellow) of a hypothetical two-gene transcription unit AB in the forward or reverse orientation. In the forward orientation transcription of gene B (pink) is expected to occur from the promoters of 5’ of gene A or the internal Tn5 but polycistronic mRNA differs in length due to the attenuation by the terminators. (B) Transposon insertions mapping to the K2 capsular operon of K. pneumoniae ECL8. Transposon insertions are configured in the forward orientation (green), and reverse orientation (blue) and insertion densities are capped at a maximum read depth of 50. The operon structure of the K2 capsular genes consisting of three promoters driving the expression of three unidirectional polycistronic transcripts is depicted.

Phylogenetic context of K. pneumoniae ECL8 and comparison to previously reported essential gene lists.

(A) Phylogenetic tree and ANI analysis generated using the IPGA webserver demonstrating the phylogenetic context of K. pneumoniae ECL8 compared to other previously published K. pneumoniae strains or isolates belonging to “21 ‘common’ lineages” of nosocomial origin by other groups. The average nucleotide identity (ANI) is a similarity index metric between a given pair of genomes applicable to prokaryotic organisms. A cutoff score of >95% indicates typically indicates they belong to the same species. (B) Tabular comparison of K. pneumoniae strains listing genome size and number of unique insertion points mapped (C) Venn diagram depicting the shared and unique genes required for growth in nutrient rich media (i.e. LB). Complete list of genomes, gene comparisons and exclusions lists can be found in Dataset S3.

Overview and validation of ECL8 fitness-factors for growth in pooled human urine.

(A) Schematic of the experimental design used to identify genes that provide a fitness advantage for K. pneumoniae ECL8 growth in pooled human urine. The K. pneumoniae ECL8 library was inoculated into either 50 mL of LB or 50 mL of urine and incubated at 37°C with 180 RPM shaking for 12 h. The library was passaged into 50 mL of fresh LB or pooled human urine at an initial OD600 of 0.05 two subsequent times. A 1 mL sample normalized to an OD600 of 1 from each culture was processed for genome extraction and multiplexed sequencing using an Illumina MiSeq. (B) Log2 fold change (Log2FC) of the read count for each K. pneumoniae ECL8 gene when passaged in urine relative to an LB control. Genes highlighted in red satisfy a stringent applied threshold (Log2FC >-2, Q-value ≤0.05). The Q-value is the P-value that has been adjusted for the false discovery rate for each gene. For brevity, only genes with a Log2FC ≥0 are illustrated. Selected transposon insertion profiles of genes identified as advantageous for growth in urine: (C) fepB, fepD, fepG (D) exbB, exbD (E) sodA and (F) ompA. These genes exhibited a significant loss of transposon insertions following growth in urine (red) in comparison to LB broth (blue). A 5-kb genomic region including the gene is illustrated. Reads are capped at a maximum depth of 1. (G) The fitness of gene replacement mutants relative to WT K. pneumoniae ECL8 in either LB medium or urine. The relative competitive index of single-gene replacement mutants after 12 h passages ×3 in either LB medium or urine. A relative fitness of one would indicate comparable fitness to WT. The mean is plotted (± 1 SD).

Growth of the K. pneumoniae TraDIS library following passaging in LB and urine and schematic diagrams of enterobactin synthesis, secretion and uptake.

The OD600 of the K. pneumoniae TraDIS library following 12 h of growth (P1) and two sequential 12 h passages (P2 and P3). The library was passaged into fresh medium (A) LB or (B) urine to an initial OD600 of 0.05. To determine the effect of iron supplementation and depletion, urine was supplemented with exogenous iron (C) 100 μM FeSO4 or an iron chelator (D) 100 μM 2,2-dipyridyl. The average OD600 of three biological replicates for each time point is plotted (±) 1 SD. (E) Area under curve comparative analysis (OD600 vs. time) of K. pneumoniae ECL8 and mutants grown in LB, urine or urine supplemented with 100 μM FeSO4 grown for 16 hours with 180 rpm shaking. The average of three biological replicates is plotted for each condition with error bars representing SD. (F) Simplified schematic of the enterobactin synthesis pathway. YbdZ, a co-factor of EntF for the terminal steps for enterobactin synthesis, depicted in light red had a Log2FC sequence read value of -1.58 suggesting this gene conferred an overall fitness advantage for growth in urine. (G) Schematic representation of enterobactin secretion and uptake. The TonB transport system is present in Gram-negative bacteria and is required to transport Fe-bound enterobactin through the outer (OM) and inner membrane (IM) to the cytosol where it can be utilized. Based on Log2FC sequence read value, loss of TolC (blue) was beneficial for growth, relative to an LB control. Loss of proteins, colored in red, had Log2FC sequence read values <-2 suggesting they confer a fitness advantage for grown in urine. Proteins depicted in grey were genes that had Log2FC values that ranged from -1 to 1 exposed to urine relative to an LB control. Genes depicted in black were essential and had no determinable Log2FC value.

Overview and validation of ECL8 genes that increase resistance to complement-mediated killing.

(A) The experimental methodology utilized for screening the TraDIS library in human serum and a heat-inactivated serum control. K. pneumoniae ECL8 (2x108 cells) of the mutant library was inoculated into either 1 mL of human serum or 1 mL of heat-inactivated human serum and incubated for 90 min. Following exposure to serum, cells were grown to an OD600 of 1 in LB medium to enrich for viable mutants. A 1 mL sample normalized to an OD600 of 1 from each culture was processed for genome extraction and multiplexed sequencing using an Illumina MiSeq. (B) Log2FC for each gene of the K. pneumoniae ECL8 TraDIS library when incubated in pooled human serum relative to a heat-inactivated serum control. Selected genes highlighted in red are amongst the total of 144 genes that satisfy a stringent applied threshold (Log2FC ≥-4, Q-value ≤0.05). For brevity, only genes with a Log2FC ≥0 are illustrated. Inset: transposon insertion profile of wbbY, gene with the highest fold Log2FC, flanked by wbbZ and a transposable element at its 3’. Transposon insertions following exposure to serum and a heat-inactivated serum control are illustrated in red and blue, respectively. Transposon reads have been capped at a maximum of 10. (C) Transposon insertion profiles of genes within the: LPS, O-antigen and the ECA biosynthesis operons. Genes in red font had a significantly (Log2FC = ≥-4, Q-Value = ≤0.05) decreased fitness when disrupted with a transposon following exposure to serum for 90 minute (red), relative to a heat-inactivated serum control (blue). Operons are not drawn to scale and reads capped at a maximum read depth of 1. (D) Growth profile of WT K. pneumoniae ECL8 and Δwbby::aph in LB broth. Mean is plotted (± 1 SD). (E) Serum killing assay of WT K. pneumoniae ECL8 and Δwbby::aph. Mean is plotted (± 1 SD). (F) LPS profiles of WT K. pneumoniae ECL8 and Δwbby::aph. Overnight cultures of each strain were normalised to an OD600 of 1. The LPS was separated on 4-12% Bis-Tris gels and was visualized by silver staining using the SilverQuest kit (Invitrogen).

Overview and validation of ECL8 fitness-factors for survival in pooled human serum.

(A) The experimental methodology utilized for screening the TraDIS library to identify genetic factors that increase resistance to human serum. K. pneumoniae ECL8 (2×108 cells) of the mutant library was inoculated into 1 mL of human serum and incubated for 180 min and compared to before serum exposure input control. The output pool was washed with PBS and plated onto LB agar supplemented with kanamycin. Following overnight growth, ∼150,000 colonies were recovered and pooled for sequencing. Comparative analysis using AlbaTraDIS software depicting genes with (B) decreased insertions suggesting a loss of fitness or (C) increased insertions suggesting a gain of fitness to serum exposure. (D) Transposon and read count insertion profiles of hns locus: red illustrating pooled mutant serum exposed for 180 min and blue denoting the before serum exposure input control. (E) Growth profile of WT K. pneumoniae ECL8 and Δhns::aph in LB broth. Mean is plotted (± 1 SD), where n=3. (F) Serum killing assay of WT K. pneumoniae ECL8 and Δhns::aph. Mean is plotted (± 1 SD), where n=3.