Report

An H-NS-like Stealth Protein Aids Horizontal DNA Transmission in Bacteria

See allHide authors and affiliations

Science  12 Jan 2007:
Vol. 315, Issue 5809, pp. 251-252
DOI: 10.1126/science.1137550

Abstract

The Sfh protein is encoded by self-transmissible plasmids involved in human typhoid and is closely related to the global regulator H-NS. We have found that Sfh provides a stealth function that allows the plasmids to be transmitted to new bacterial hosts with minimal effects on their fitness. Introducing the plasmid without the sfh gene imposes a mild H-NS phenotype and a severe loss of fitness due to titration of the cellular pool of H-NS by the A+T-rich plasmid. This stealth strategy seems to be used widely to aid horizontal DNA transmission and has important implications for bacterial evolution.

Horizontal transfer of genetic material allows bacteria to acquire new traits, such as those contributing to virulence or symbiosis, or resistance to antibiotics and other antimicrobial agents (1, 2). An unresolved issue in studies of horizontal transfer is how the newly acquired genes become embedded in the regulatory networks of the host in ways that promote the long-term interests of both the genes and the host (3). Importing even a structurally simple plasmid can impose a fitness cost on the bacterium (46), so one might anticipate that the arrival of large groups of genes could place the recipient at a competitive disadvantage.

R27 is a large plasmid that was isolated from Salmonella enterica serovar Typhimurium (S. Typhimurium) in the early 1960s (7) and has been detected more recently in S. Typhi outbreaks in India, Pakistan, and Southeast Asia (8, 9). An IncHI1 plasmid called pSf-R27 that is 55% A+T and 99.7% identical to R27 was discovered in Shigella flexneri 2a strain 2457T (3, 1012). The plasmids differ in that pSf-R27 harbors no antibiotic resistance genes. Like other R27-like plasmids, pSf-R27 encodes a protein that is 59% identical to the DNA binding protein H-NS (3, 10, 11), an abundant nucleic acid binding protein found in many Gram-negative bacteria. H-NS binds to regions of curvature in A+T-rich DNA, allowing it to act globally as a repressor of transcription (3, 13). The pSf-R27–encoded paralog, Sf h, can substitute fully for H-NS (and vice versa) in Escherichia coli and S. flexneri and has a similar preference for binding curved A+T-rich DNA sequences (10, 11).

To test the hypothesis that horizontal transfer by conjugation of the sfh gene located on pSf-R27 could alter global gene expression patterns in transconjugants, we introduced plasmid pSf-R27, with or without a functioning sf h gene, into the mouse-virulent SL1344 strain of S. Typhimurium. We observed the effect on the transcriptome, competitive fitness, and virulence of the transconjugants. Surprisingly, the introduction of pSf-R27 resulted in relatively few changes in the recipient, whereas its Δsfh derivative (pSf-R27Δsfh) exerted many unexpected effects.

The native pSf-R27 plasmid altered the transcription of a limited number of S. Typhimurium genes (Fig. 1), with just 25 showing more than a 50% reduction (table S1) and 68 showing more than a doubling in transcription (table S2). In contrast, the introduction of the sfh-deleted plasmid pSf-R27Δsfh resulted in altered expression of a much wider range of genes distributed across all categories (Fig. 1), with 119 genes down-regulated (table S3) and 323 genes up-regulated (table S4). Prominent among these were genes involved in bacterial virulence, motility, the DNA damage response, and central metabolism. The effect on genes contributing to each of these processes was confirmed by the reverse transcription polymerase chain reaction (figs. S1 and S2). Consistent with the changes seen in its transcriptome, SL1344 (pSf-R27Δsfh) was considerably more resistant to ultraviolet radiation than either of the other strains (fig. S1) and had an enhanced level of survival in macrophage (Fig. 2). It also displayed reduced motility (fig. S2). The changes in motility and virulence were reminiscent of the phenotypes of hns mutants (1416). Full motility was restored when the hns gene was expressed from a recombinant plasmid, and the cloned sf h gene also restored full motility in an hns mutant, consistent with the known interchangeability of these proteins (fig. S2).

Fig. 1.

The impact of plasmids pSf-R27 and pSf-R27Δsfh on the transcriptome of SL1344. Data are presented for exponential (white bars) and stationary-phase (black bars) microarray data from SL1344 versus SL1344 (pSf-R27) (A) and SL1344 (pSf-R27) versus SL1344 (pSf-R27Δsfh) (B). Gene categories are given at the left of each panel and are based on the Kyoto Encyclopedia of Genes and Genomics (KEGG). The histograms represent the percentage of genes in each category affected by the introduction of the plasmid.

Fig. 2.

The effect of plasmids pSf-R27 and pSf-R27Δsfh on bacterial host survival in J774-A.1 macrophage. J774-A.1 macrophage–like cells were infected with SL1344, SL1344 (pSf-R27), or SL1344 (pSf-R27Δsfh) as described in supporting online material. Bacteria were recovered at the time intervals shown, and the number of colony-forming units of each bacterial strain was determined.

Overall, these data suggested a correlation between the presence of the pSf-R27Δsf h plasmid and a mild H-NS phenotype, even though the hns gene on the Salmonella chromosome remained intact and its expression was unaltered (tables S1 to S4). Given the wide-ranging influence of pSf-R27Δsf h on the gene expression profile and several phenotypes of SL1344, we tested the possibility that the introduction of this plasmid might influence the competitive fitness of the bacterium. The presence of pSf-R27 had a mild negative effect on the fitness of SL1344, whereas the mutant plasmid pSf-R27Δsf h strongly reduced bacterial fitness (Table 1). H-NS is known to target A+T-rich sequences in the Salmonella genome (15, 16). The hns-like phenotype of the exconjugants harboring pSf-R27Δsf h suggested that the A+T-rich plasmid might influence the global gene expression pattern of the bacterium—and hence its fitness—by titrating H-NS, a protein that is closely related to Sfh. Both proteins bind to pSf-R27Δsf h (fig. S3) and have strikingly similar DNA structural requirements for binding (10, 11). They also bind to the regulatory region of the ssrA virulence regulatory gene (fig. S4) and other genes identified in this study as responding to the presence of pSf-R27Δsf h. We tested this hypothesis by cloning the DNA fragment encompassing the Sfh/H-NS binding region from the ssrA promoter into plasmid pUC18. Such binding regions consist of a nucleation site from which H-NS can polymerize along the DNA (13). This construct caused a strong reduction in SL1344 fitness, and the subsequent introduction of a compatible recombinant plasmid (pPD101sfh) expressing the Sf h protein restored fitness (Table 1). This showed that the reduction in fitness was due to the presence of multiple copies of the H-NS/Sfh binding region and did not require any property that was specific to the pSf-R27Δsf h plasmid.

Table 1.

Bacterial competitive fitness measurements. Relative mean fitness values shown are with reference to the common competitor SL1344 Nalr.

StrainRelative mean fitness
SL1344 1.02 ± 0.04
SL1344 (pSf-R27) 0.83 ± 0.01
SL1344 (pSf-R27Δsfh) 0.22 ± 0.04
SL1344 (pSf-R27Δsfh pPD101) 0.47 ± 0.06
SL1344 (pSf-R27Δsfh pPD101sfh) 1.30 ± 0.07
SL1344 (pUC18) 0.72 ± 0.03
SL1344 (pUCA+T) 0.47 ± 0.02
SL1344 (pUCA+T, pPD101sfh) 0.98 ± 0.02
SL1344 (pPD101sfh) 1.06 ± 0.02

We propose that sf h is a stealth gene that allows the A+T-rich pSf-R27 plasmid to enter new bacterial hosts with a minimal impact on global gene expression patterns and fitness, ensuring the future competitiveness of the new plasmid-host combination. This strategy has the effect of smoothing the horizontal transmission of genetic information within and between bacterial populations, as evidenced by the presence of hns-like genes on other plasmids (1720) and on other horizontally acquired A+T-rich DNA elements such as pathogenicity islands that are known to bind H-NS (20, 21). The positive effect on competitive fitness of adding the sf h gene to commonly used cloning vectors has obvious biotechnological implications.

Supporting Online Material

www.sciencemag.org/cgi/content/full/315/5809/251/DC1

Materials and Methods

Figs. S1 to S4

Tables S1 to S6

References

References and Notes

View Abstract

Navigate This Article