H-NS mediates the silencing of laterally acquired genes in bacteria

Abstract

Histone-like nucleoid structuring protein (H-NS) is a modular protein that is associated with the bacterial nucleoid. We used chromatin immunoprecipitation to determine the binding sites of H-NS and RNA polymerase on the Salmonella enterica serovar Typhimurium chromosome. We found that H-NS does not bind to actively transcribed genes and does not co-localize with RNA polymerase. This shows that H-NS principally silences gene expression by restricting the access of RNA polymerase to the DNA. H-NS had previously been shown to preferentially bind to curved DNA in vitro. In fact, at the genomic level we discovered that the level of H-NS binding correlates better with the AT-content of DNA. This is likely to have evolutionary consequences because we show that H-NS binds to many Salmonella genes acquired by lateral gene transfer, and functions as a gene silencer. The removal of H-NS from the cell causes un-controlled expression of several Salmonella pathogenicity islands, and we demonstrate that this has deleterious consequences for bacterial fitness. Our discovery of this novel role for H-NS may have implications for the acquisition of foreign genes by enteric bacteria.

Figure 1. H-NS Chip and Transcriptomic Data in Context of the Genome Atlas for the S. typhimurium Chromosome

The data from the RNAP and H-NS ChIP-on-chip experiments are shown using coloured concentric circles which are mapped onto the Salmonella chromosome. In addition, we show our previously published transcriptome data for the JH4000 hns-deletion mutant [2] and four DNA structural measures (http://www.cbs.dtu.dk/services/GenomeAtlas [15]). The order of the concentric circles is indicated in the legend. Numbers on the inside of the innermost circle are the location relative to position zero measured in millions of base-pairs (Mbp). The origin of replication, predicted terminus region, location of rRNAs and the SPI1 and SPI2 pathogenicity islands are indicated at the edge of the outermost circle.

Figure 2. Correlation between Binding of H-NS, RNAP, and Level of Gene Expression

(A) Frequency distribution of relative gene expression levels for S. typhimurium. The results are shown for all S. typhimurium genes (filled circle) and for genes found to co-localize with H-NS (filled triangle) or RNAP (open circle) in the ChIP-on-chip experiment. Relative gene expression levels for S. typhimurium strain LT2 were taken from Ono et al [2]. (B and C) Scatter plot correlating binding of RNAP or H-NS to gene expression levels. The relative expression level of a gene (y-axis) is plotted against the corresponding level of binding to RNAP or H-NS (x-axis) as measured in the ChIP-on-chip experiments (see Materials and Methods).

Figure 3. Distribution of H-NS and RNAP over Segments of the S. typhimurium Genome

(A–C) Graphs showing H-NS (filled circle) and RNAP binding (open circle) over the rrnA locus, proU operon, and leuO gene. (D and E) H-NS and RNA polymerase binding over the Salmonella pathogenicity islands 1 and 2 (SPI1, SPI2). Genes within SPI1 or SPI2 are represented by black arrows. The boundaries of SPI1 and SPI2 were defined from the Horizontal Gene Transfer Database http://www.tinet.org/~debb/HGT [25].

Figure 4. Correlation between AT-Content, Curvature and H-NS, and RNAP Binding to DNA

For each oligonucleotide on the microarrays we calculated its average AT-content (A) and curvature (B). To minimize bias that might have been introduced by the algorithms used to design the oligonucleotides, flanking sequences (50 bp) of the oligonucleotides were included in the analysis. The results of each oligonucleotide were plotted against the level of H-NS binding (grey circles). Window-averaged results (window size: 101) are also shown (solid line). (C and D) AT-content and curvature profiles were calculated from putative promoter sequences aligned by over-laying every S. typhimurium gene at the translational start. The profiles have been calculated for all the S. typhimurium genes (black), and for those subsets of genes bound by H-NS (green) or RNAP (red).

Figure 5. H-NS Confers a Growth Advantage for Wild-Type Salmonella by Silencing Genes Acquired through Lateral Gene Transfer

The growth of the five S. typhimurium strains were followed in LB, as described in Materials and Methods. The data shown are representative of three experiments, and discussed in the text.