Posttranscriptional control of the Salmonella enterica flagellar hook protein FlgE

Abstract

Previous work suggested that the FlgE (flagellar hook subunit) protein in Salmonella enterica serovar Typhimurium was posttranscriptionally regulated in response to the stage of flagellar assembly. Specifically, the FlgE protein could be detected in flagellar mutants defective at the stages of assembly before or after rod assembly but not in rod assembly mutants, yet flgE mRNA levels were unaffected. To elucidate posttranscriptional mechanisms involved in the coupling of flgE gene expression to hook assembly, the RNA sequences at the 5' and 3' ends of the flgE-containing mRNA processed from the large flgBCDEFGHIJKL operon were determined by rapid amplification of cDNA ends, and secretion of the FlgE protein in different flagellar assembly mutant strains was analyzed. The sequences 5' and 3' of the flgE gene where RNA processing occurred was within 15 bases upstream of the flgD stop codon and at bases 145 to 147 downstream of the flgF start codon, respectively. The ribosome binding site of the flgD gene was found to be inhibitory to flgE translation in strains deleted for the upstream flgD gene, unless the region 15 bases upstream of the flgD stop codon was present. Secretion of FlgE into the periplasm was monitored using beta-lactamase (Bla) fusions as a periplasm-specific reporter, which conferred resistance to ampicillin when FlgE-Bla was secreted into the periplasm. Using this assay, we found that the effect of rod assembly mutants on FlgE levels was due to FlgE turnover in the periplasm and that the FliE rod component protein was required for efficient FlgE-Bla secretion.

FIG. 1.

Flow diagram of 5′ RACE. Total RNAs were randomly ligated and then reverse transcribed using a random primer. PCR amplification was performed with 5SF and gene-specific primers. The PCR products were cloned and sequenced.

FIG. 2.

RNA processing sites in the flgB-L operon. (A) Organization of genes in the flgB-L operon. The mRNA processing sites are indicated. (B) Major (large arrows) and minor (small arrows) mRNA processing sites were mapped. The stop codon of flgD is indicated in RP1. RP, RNA processing.

FIG. 3.

Motility assay of flgD-deficient strain TH8547 with pflgD (A) or pflgDE (B). pTrc99A (vector control), pflgD, or pflgDE was transformed into the wild type (LT2) or TH8547. pTrc99A and pflgD did not affect the motility of the wild type. TH8547 was not complemented with pflgD (A) but was complemented with pflgDE (B).

FIG. 4.

Construction of flgD strains and motility assay of flgD-deficient strains with pflgD or pflgDE. The motility activities (percentage of wild-type motility with pflgD or pflgDE) of flgD strains are indicated in the left two columns.

FIG. 5.

Isolation and characterization of motile revertants of pflgD/TH8547. (A) The sequences of two RBSs are shown. The sites of the mutations that restored the motility activity in TH9936, TH9938, and TH9939, respectively, are indicated by asterisks. The numbers are relative to the flgB start codon, ATG. (B) Motility assay of motile revertants. Three point mutants are indicated. (C) TH9945 and TH9946 were made by site-directed mutagenesis. Their motility activities are similar to those of TH9936 and TH9938, respectively.