Collateral effects of deletion of nlpD on rpoS and rpoS-dependent genes. Reply.

In their Letter to the Editor, Tsunoi et al. (1) discuss a very interesting and relevant aspect of our published work (2) and suggest topics for further investigation. This published study identifies the E. coli NlpD protein as a potent regulator of gene expression in human cells. This effect is attributed to the inhibition of RNA polymerase II phosphorylation (Pol II), mediated, in part, by effects of NlpD on the Pol II RPB1 subunit and PAF1C (2, 3). The discovery of the NlpD effect was made possible by the isolation of a loss-of-function mutant (SN25) from a human carrier of the parent strain E. coli 83972, and a mutation strategy was devised to introduce the nlpD point mutation from SN25 into the parent strain and to complement SN25 with a fully functional nlpD gene cluster. To address whether NlpD exerts its effects on host cells independently or in synergy with sigma38 (σ³⁸), the 5′nlpD coding sequence was deleted in E. coli 83972, while leaving the rpoS promoter intact. The partial nlpD deletion mutant lost the ability to inhibit PAF1C and Pol II phosphorylation.

In the E. coli genome, nlpD and rpoS form an operon. σ³⁸ expression is regulated, in part, from a promoter located within nlpD. Elegant studies addressing the role of σ³⁸ as a global regulator of bacterial gene expression are referenced in the letter by Tsunoi et al. (1). Global analyses of the rpoS regulon in E. coli K-12 have identified more than 1000 genes (4, 5). As pointed out by Dr. Iwase and colleagues, effects on σ³⁸ are therefore likely to affect bacterial phenotypes, under a variety of conditions. A loss of σ³⁸ expression changes metabolic activity or virulence in uropathogenic E. coli (6). Our preliminary comparison of E. coli 83972 and SN25 gene expression profiles indicates that typical σ³⁸-dependent stress response genes are downregulated in SN25 relative to the 83972 wild-type strain.

Importantly, the effects of NlpD on human cells were characterized using recombinant NlpD protein rather than whole bacteria. Therapeutic efficacy was demonstrated as inhibition of inflammation and accelerated bacterial clearance from infected tissues in a murine urinary tract infection model (Figure 1). In contrast to NlpD, recombinant σ³⁸ displayed no affinity for RPB1 or PAF1C in pull-down experiments. Instead, the data suggest that, by inhibiting TBP binding, σ³⁸ itself may act as an additional regulator of gene expression in infected hosts (2).

Figure 1

Recombinant NlpD treatment accelerated bacterial clearance in CY17-infected C57BL/6 mice (A) and attenuated inflammation (B). Data are presented as mean ± SEM (n = 5 mice). Two-way ANOVA with Sidak’s multiple comparison tests, *P < 0.05, **P < 0.01, ***P < 0.001.

Thanks to Drs. Tsunoi, Iyoda, and Iwase for initiating this interesting discussion.