Correction

In Figure 5 panel B of this article, we reported what was believed to be a mutant of the Pseudomonas aeruginosa (PAO1) PcrG protein capable of trans-complementing an lcrG mutation. After isolation of PcrG mutants that could trans-complement lcrG, we inadvertently compared our mutated pcrG sequences to the pcrG sequence from P. aeruginosa strain 388 instead of strain PAO1. This error resulted in the incorrect characterization of a PcrG variant described in this paper (PcrG F42L). In fact, the residue at position 42 of PcrG from PAO1 is a leucine, not a phenylalanine. Therefore, we did not have a PcrG mutant, but had a wildtype clone.

This error changes our interpretation of the experiment that tested the trans-complementation of an lcrG strain of Yersinia pestis with PcrG. We reported that our clone of pcrG on plasmid pJM132 could not complement an lcrG strain of Y. pestis (Fig. 5). Discovery of the sequencing error led us to re-sequence several of our pcrG constructs. We found that our original subclone, pJM132, contained a deletion of an adenine residue between the ribosome-binding site and the initiating ATG of pcrG. Our complementing "mutant" on pJM133 (Fig. 5) contains the deleted adenine. This result suggests that the failure of pJM132 to complement is due to an expression problem. Therefore, we now conclude that pcrG from P. aeruginosa PAO1 complements an lcrG strain of Y. pestis. We apologize for the error.