Regulation of Salmonella enterica serovar typhimurium invasion genes by csrA

Abstract

Penetration of intestinal epithelial cells by Salmonella enterica serovar Typhimurium requires the expression of invasion genes, found in Salmonella pathogenicity island 1 (SPI1), that encode components of a type III secretion apparatus. These genes are controlled in a complex manner by regulators within SPI1, including HilA and InvF, and those outside SPI1, such as the two-component regulators PhoP/PhoQ and BarA/SirA. We report here that epithelial cell invasion requires the serovar Typhimurium homologue of Escherichia coli csrA, which encodes a regulator that alters the stability of specific mRNA targets. A deletion mutant of csrA was unable to efficiently invade cultured epithelial cells and showed reduced expression of four tested SPI1 genes, hilA, invF, sipC, and prgH. Overexpression of csrA from an induced araBAD promoter also negatively affected the expression of these genes, indicating that CsrA can act as both a positive and a negative regulator of SPI1 genes and suggesting that the bacterium must tightly control the level or activity of CsrA to achieve maximal invasion. We found that CsrA affected hilA, a regulator of the other three genes we tested, probably by controlling one or more genetic elements that regulate hilA. We also found that both the loss and the overexpression of csrA reduced the expression of two regulators of hilA, hilC and hilD, suggesting that csrA exerts its control of hilA through one or both of these regulators. We further found, however, that CsrA could affect the expression of both invF and sipC independent of its effects on hilA. One additional striking phenotype of the csrA mutant, not observed in a comparable E. coli mutant, was its slow growth. Phenotypic revertants that had normal growth rates, while maintaining the csrA mutation, were common. These suppressed strains, however, did not recover the ability to invade cultured cells, indicating that the csrA-mediated loss of invasion cannot be attributed simply to poor growth and that the growth and invasion deficits of the csrA mutant arise from effects of CsrA on different targets.

FIG. 1

Effect of csrA on bacterial growth. Strains were grown with aeration in LB broth buffered with HEPES to pH 8. The csrA⁺ plasmid used was pCA132. Values represent mean and standard error for strains tested in triplicate.

FIG. 2

Effect of csrA overexpression on invasion gene expression. csrA was expressed under the control of the araBAD promoter on plasmid pCA114 in the ara9 strain background. Arabinose or glucose was added to a final concentration of 0.5% to induce or repress, respectively, csrA expression. Values represent mean and standard error for strains tested in triplicate.

FIG. 3

Control of hilA expression by csrA. pLS31 and pLS79 carry transcriptional lacZY fusions to hilA. pLS31 has positions −497 to +420 of hilA, and pLS79 has positions −39 to +420 of hilA. pCA132 expresses csrA, and pCA71 expresses csrB. Values represent mean and standard error for strains tested in triplicate.

FIG. 4

Regulation of hilD and hilC by csrA. Northern analysis was performed using equal amounts of total bacterial RNA prepared from each of four strains and probes generated from hilD or hilC. pCA114 carries csrA under the control of the araBAD promoter on pBAD18. Strains with both plasmids were grown with 0.5% arabinose added to the medium for promoter induction.

FIG. 5

A model for regulation of SPI1 genes by CsrA. Under conditions conducive to invasion, active CsrA concentration is kept low, either by decreased production or by binding to CsrB. At this concentration, CsrA preferentially degrades a repressor (heavy dashed line) that affects hilC and hilD and independently invF. For simplicity, positive control by CsrA is shown as resulting from the message degradation of a single repressor. Alternatively, CsrA might have more than one target in the invasion gene regulation pathway. Under invasion-repressing conditions, CsrA concentration increases, degrading the message of an activator that works through hilD and/or hilC, thus reducing the expression of invasion genes. Solid arrows represent known routes of regulation; dashed arrows show the hypothetical route of SPI1 gene control by CsrA.