Combining Shigella Tn-seq data with gold-standard E. coli gene deletion data suggests rare transitions between essential and non-essential gene functionality

Table 1 Genes artefactually inferred as essential in Shigella

Gene(s)	Evidence for Artefactual Inference of Shigella Essentiality
acrAB, tolC, ybaB, ksgA, yebC, smpB (0.73) ^a	Affect kanamycin resistance [35–37]
dnaQ, holD, recD, xseA, ruvA (0.58), ruvB (0.60), ruvC (0.61), recB (0.58), recC (0.65)	Likely to affect the transposition process; dnaQ, holD, ruvA, and ruvB inferred as essential using Tn-seq in Salmonella [14]
priB	Deficient in plasmid maintenance [63, 64]
fhuACD, tonB	Involved in iron acquisition which is critical for growth in the iron limited media used in this study [33]
rpsT	S20 ribosomal subunit; new data indicates mutants have poor growth [42] (in conflict with Keio data)
miaA	tRNA dimethyl transferase; previous data indicates E. coli mutants have poor growth [43] (in conflict with Keio data)
ompA	Outer membrane porin; clean knockouts appear viable [65]; mutant forms are frequently lethal [44]
pitA	Metal phosphate transporter with ten transmembrane segments; transposon disruption of substrate transporters is three-fold more likely to be inferred as essential compared to clean deletion (see main text; Fig. 5)
potB	Type I ABC transporter (Putrescine / spermidine transporter); see above
sapB	Type I ABC transporter (unknown substrate); see above
ptsH, yeaC, ydhR	Short reading frames (306 bp, 272 bp, and 258 bp, respectively)

^aGene deletions of orthologous E. coli genes with growth levels less than OD600 0.75 have these levels in parentheses

ISSN: 1471-2180