Article Links:

• Interview: The Path to a Salmonella Research Career

Links to Other ASM Pages:

• Table of Contents
• ASM News Issues

The Path to Salmonella

Multiple horizontally-acquired genes are responsible for the specific virulence properties of Salmonella

Eduardo A. Groisman and Howard Ochman

The salmonellae constitute a major threat to our health and welfare. These gram-negative bacteria are responsible for 16 million annual cases of typhoid fever and 1.3 billion annual cases of gastroenteritis, which together result in more 3 million annual deaths. In addition, the salmonellae cause disease in a number of animal species. But, what are the genetic origins of virulence in this enteric species?

Diversity within the Salmonellae

There is a single species in the genus Salmonella--Salmonella enterica--that consists of over 2,300 serotypes. Salmonella serotypes are defined by the flagellar (or H) and lipopolysaccharide (or O) antigens; however, these serological classifications do not portray the underlying genetic relatedness of strains within the species. Based on DNA-DNA hybridization and multilocus enzyme electrophoresis, S. enterica has been split into eight subspecific groups, and most strains isolated from humans belong to subspecies I. This subspecies includes S. typhi, a human-adapted serotype that is the causal agent of typhoid fever, S. typhimurium, which is responsible for gastroenteritis in humans and provokes a typhoid-like disease in mice, and the bacteremia-causing S. choleraesuis.

By all appearances, S. enterica is a very different organism from Escherichia coli, which is typically a benign constituent of the mammalian intestinal flora. Yet, most molecular and genetic criteria reveal that these two bacterial species are remarkably similar: the chromosomes of E. coli K-12 and S. enterica serovar Typhimurium are the same size, and the order, orientation, and spacing of genes are nearly identical. In fact, the broad colinearity of their linkage maps has led to the view that these bacterial chromosomes are evolutionarily conserved. Given the overall similarities in genetic organization, what is the source of the differences that rendered E. coli a commensal and Salmonella a pathogen?

The Common Ancestor of E. coli and Salmonella

It has been estimated that E. coli and Salmonella diverged from a common ancestor some 100 million years ago. Based on ecological and genetic considerations, this time frame coincides with the origin of mammals--and, hence, homeothermy, the mammalian colon, the milk sugar lactose--which provided a new niche allowing the initial divergence between Salmonella and E. coli. But what were the characteristics of the ancestral organism that gave rise to two species having very different lifestyles? It is more likely that the common ancestor to E. coli and Salmonella was a commensal organism that became a pathogen upon acquisition of virulence gene clusters--as opposed to a pathogen that lost essential virulence determinants.

Consistent with the notion that Salmonella originated from a commensal organism, many of the genes implicated in Salmonella virulence are also present in nonpathogenic strains of E. coli. These genes encode enzymes responsible for the biosynthesis of nutrients that are scarce within host tissues, transcriptional and post-transcriptional regulatory factors, proteins necessary for the repair of damaged DNA, and other products necessary for defense against host microbicidal mechanisms. Furthermore, adhesion to cells of the urinary tract by uropathogenic strains of E. coli is mediated by P fimbriae, and the genes encoding these surface organelles are present in commensal strains of E. coli, where they increase persistence in the colon. Moreover, the biosynthetic genes argC and guaA are present in benign forms of E. coli, yet they are both required for the survival and replication of uropathogenic strains of E. coli during extraintestinal infections.

The presence of determinants that promote interactions with animal cells would seem to predispose E. coli, which are normal constituents of the human intestinal flora, to become a pathogen upon the incorporation of virulence gene clusters. And indeed, a laboratory strain of E. coli will produce the attachment and effacing lesions typically associated with enteropathogenic E. coli (i.e., EPEC) if provided the 35-kb EPEC-specific region responsible for this phenotype.

What Makes Salmonella a Pathogen

The traits that distinguish E. coli from S. enterica are associated with genes that are present in only one of the two species (as opposed to genes that are present in both organisms but differentially expressed or exhibiting allelic differences). For example, the ability to use lactose as a sole carbon source is encoded by the lac operon, which resides in a segment of the E. coli chromosome not present in the Salmonellas genome. In fact, the incorporation of the lac operon and the ensuing capacity to ferment lactose may well have been responsible for the successful commensal relationship between E. coli and its mammalian hosts. Similarly, the ability of Salmonella to transport citrate for use as a carbon source is conferred by a region of the chromosome that is unique to this genus.

Because species-specific genes were known to be responsible for some of the unique metabolic capabilities of a species, we hypothesized that the characterization of Salmonella-specific regions would help define the molecular basis for virulence traits in this intracellular pathogen. About a decade ago, we began to examine the virulence role of DNA sequences that are restricted to Salmonella. As our starting material, we employed a set of five clones that Renee Fitts (then at Integrated Genetics, Framingham, Mass.) had isolated to detect Salmonella in food and tissues.

Two of the clones originally isolated by Fitts are part of large Salmonella-specific regions that confer invasion of and survival within host mammalian cells--traits that are essential to Salmonella when it infects animals. After an animal acquires Salmonella, typically by consuming contaminated water or food, the bacteria must endure the acid pH of the stomach before encountering, adhering to, and entering the cells that line the intestinal epithelium. Those invasive microorganisms destined to cause systemic disease, such as typhoid fever, must survive in blood and replicate in the macrophages of liver and spleen. Meanwhile, those microorganisms causing chronic infections hide in the gallbladder of infected individuals.

Role of Salmonella Pathogenicity Islands in Virulence

Pathogenicity islands are chromosomal clusters of virulence genes found in pathogenic organisms but absent or sporadically present in related nonpathogenic species. Pathogenicity islands are often found adjacent to tRNA genes suggesting transfer into these broadly conserved sites. In addition, the G+C content of pathogenicity islands is usually different from that of the rest of the chromosome, which can be taken as additional evidence of their extraneous origins. To date, five pathogenicity islands (as well as several small islets) have been identified in Salmonella enterica, and additional virulence regions are likely to be uncovered upon completion of the ongoing Salmonella genome projects.

The best characterized pathogenicity island in Salmonella is SPI-1, a 40-kb segment that encodes genes enabling the microorganism to invade non-phagocytic cells. Unlike the other pathogenicity islands of Salmonella enterica identified to date, SPI-1 is not located immediately adjacent to a tRNA gene.

SPI-1 contains 31 genes encoding the components of a type III secretion system designated Inv/Spa, the effector proteins secreted by this system and the corresponding effector chaperones, transcriptional regulatory factors, and a few proteins of unknown function. (Type III secretion systems are export apparatuses used by gram-negative species to deliver proteins into the cytosol of eukaryotic cells.)

Apart from its role in eukaryotic cell invasion, the Inv/Spa system has been also implicated in macrophage apoptosis in vitro. While SPI-1 genes may participate in stages of infection beyond entry into the intestinal epithelial cells, mutants defective in the Inv/Spa system are attenuated in mice when inoculated orally but fully virulent if inoculated intraperitoneally. This differential behavior suggests that SPI-1 is not required during systemic infection and that its primary role is to mediate bacterial entry into the host's intestine.

The biochemical functions of the effector proteins delivered via the Inv/Spa system are beginning to be elucidated: SipA is an actin-binding protein, SopB is an inositol phosphate phosphatase, SopE activates the small GTP-binding proteins CDC42 and Rac by promoting GDP/GTP exchange, and SptP is both a GTPase-activating protein for Rac-1 and Cdc42 and a protein tyrosine phosphatase. Although mutants defective in each of these proteins show only minor defects in eukaryotic cell entry, a picture is now emerging on how these proteins contribute to Salmonella invasion of mammalian cells.

Salmonella and Shigella Deploy Similar Systems for Host Cell Invasion

Figure 1

The organization of genes encoding the type III secretion system within SPI-1 is very similar to that of the type III secretion genes which are responsible for eukaryotic cell invasion and macrophage apoptosis in Shigella flexneri. The Shigella type III system--known as Mxi/Spa--is encoded on a large virulence plasmid, attesting to the mobile nature of these sequences. But despite similarities in the Salmonella and Shigella invasion gene clusters, the Shigella plasmid is an unlikely source of SPI-1 genes. SPI-1 is ancestral to all strains of Salmonella, implying these sequences were incorporated early in the evolution of Salmonella, whereas Shigella is a recently derived pathogen. Conversely, the Shigella mxi/spa genes, which are extremely AT rich, could not have derived from the Salmonella homologs, which are only moderately AT rich.

In addition, these virulence regions were subject to extensive rearrangements before they were acquired independently by Salmonella and Shigella. For example, in Salmonella, the sipABCD genes specifying exported proteins are immediately downstream of the genes encoding the Inv/Spa secretion apparatus. But, on the Shigella virulence plasmid, the transport apparatus and effector proteins are separated by at least 10 genes, many of which are unique to Shigella. One of the striking properties of the Inv/Spa system of Salmonella is its capacity to translocate effector proteins encoded outside the SPI-1 pathogenicity island: the sopB gene is part of the SPI-5 pathogenicity island and the sopE gene is encoded within a bacteriophage. Coincidentally, the genes encoding the Shigella homologs of Salmonella SopB, as well as its specific chaperone, are located on the virulence plasmid.

A Second Type III Secretion System in Salmonella enterica

Salmonella is unique among gram-negative pathogens in that it harbors two distinct type III secretion systems, each contributing to distinct stages of the infectious process. The second type III system, designated Spi/Ssa and encoded within the SPI-2 pathogenicity island, differs from the Inv/Spa system in genetic organization, phylogenetic distribution and function. Unlike the Inv/Spa system encoded within SPI-1, Spi/Ssa is essential for causing systemic disease, and mutants defective in this system are highly attenuated in both orally and intraperitoneally inoculated animals. Specifically, the SPI-2 island appears to mediate bacterial replication, rather than survival, within host macrophages.

The SPI-2 pathogenicity island is a 40-kb segment located immediately adjacent to the valV tRNA gene on the Salmonella chromosome. The 32 genes constituting SPI-2 encode components of the Spi/Ssa secretion system, the putative effectors of the Spi/Ssa system, a two-component regulatory system, and several proteins whose functions are as yet unknown. The type III system within SPI-2 is only expressed during bacterial growth in host cells, and not when Salmonella is cultured under laboratory conditions. One of the proteins secreted by the Spi/Ssa system, SpiC, inhibits fusion of the Salmonella-containing phagosome with lysosomes, and thereby allows bacterial growth inside macrophages.

Salmonella remains within a membrane-bound vacuole in both epithelial cells and macrophages; thus, Spi/Ssa is the first example of a type III secretion system that functions intracellularly, delivering virulence proteins from within a phagosome and into the cytosol of the eukaryotic host cell. This differs from typical type III systems, such as Inv/Spa, that mediate the translocation of bacterial proteins by microorganisms located extracellularly.

The SPI-1 pathogenicity island is ancestral to Salmonella because it is present in all eight subspecies that comprise S. enterica. In contrast, SPI-2 sequences have not been detected in the most divergent subspecific group of Salmonella. These results imply that acquisition of SPI-2 occurred later in the evolution of Salmonella, conferring the ability to cause systemic disease upon organisms that were already capable of invading host tissues.

A Pathogenicity Island at the Salmonella selC Locus

The selC tRNA gene serves as the site of insertion of two different pathogenicity islands in enteropathogenic and uropathogenic strains of E. coli and as the attachment site for the retron phage f R73. The repeated use of this site for the acquisition of foreign DNA prompted us to examine whether the selC locus of S. enterica harbors horizontally acquired sequences; and this led to the identification of a 17-kb pathogenicity island, which was designated SPI-3.

The SPI-3 pathogenicity island is 17 kb in length and harbors 10 genes organized into six transcriptional units. They include the mgtCB operon, encoding the macrophage survival protein MgtC and the Mg²⁺ transporter MgtB, as well as a putative transcriptional regulator and proteins of unknown function. The SPI-3-encoded mgtC gene is required for virulence in mice. The MgtC protein may mediate Mg²⁺ uptake since its structure is suggestive of a membrane-bound protein and mgtC mutants are unable to grow in low Mg²⁺ media. SPI-3 has a mosaic structure, and the distributions of the genes within the SPI-3 island of typhimurium are variable among the salmonellae.

Other Pathogenicity Islands and Islets in Salmonella

Recent studies have uncovered at least two additional pathogenicity islands in Salmonella. One of them, SPI-4, is a 27-kb region situated next to a putative tRNA gene. SPI-4-hybridizing sequences have been detected in several host-specific serovars of S. enterica, and one of the 18 genes within SPI-4 is known to be required for intramacrophage survival.

The other of these pathogenicity islands, SPI-5, is a 7.5-kb region that maps to the serT tRNA locus. This pathogenicity island encodes six genes, four of which have been implicated in cattle enteritis by S. enterica serovar Dublin. However, recent experiments suggest that SPI-5 does not play a role in systemic disease.

Not all pathogenicity islands are large regions encoding multiple genes or complex functions. For example, the 1.6-kb sifA is a Salmonella-specific gene required for the bacterial-promoted formation of filamentous structures in the lysosomal vacuoles of infected epithelial cells. The sifA-containing fragment interrupts the pot operon and is bounded by 14-bp direct repeats, suggesting that its incorporation involved a recombination mechanism similar to those mediating the integration of phage or transposable elements.

Coordinating Expression of Horizontally-Acquired Virulence Genes

Horizontal gene transfer offers a rapid means for bacteria to acquire new functions. Genes participating in the same general pathway are often clustered and their incorporation as a single unit can confer new abilities upon a recipient microorganism. However, this poses a problem: for acquired sequences to be of use to the recipient organism, their expression must be coordinated with that of the rest of the genome.

Many pathogenicity islands encode regulatory proteins governing the expression of genes located within the island--such as HilA and InvF in SPI-1, and SpiR/SsrA, SsrB in SPI-2. These island-encoded regulators are often themselves under the control of global regulators that are already present in the genome. This hierarchy of regulators serves to coordinate the expression of acquired sequences with that of the rest of the genome.

The best example of how global regulators affect virulence is provided by the PhoP/PhoQ two-component system, an ancestral regulator that governs the expression of genes in the SPI-1 and SPI-3 pathogenicity islands. The PhoP/PhoQ system responds to the extracellular levels of Mg²⁺. Thus, when the Mg²⁺ concentration in the environment is low, transcription of PhoP-activated genes is induced, whereas high Mg²⁺ concentrations repress the expression of PhoP-activated genes.

PhoP-activated genes are expressed to high levels within host cells--presumably due to the low Mg²⁺ concentration in the Salmonella-containing phagosomes--resulting in the expression of virulence genes, such as mgtC, which promote intramacrophage survival. A similar situation occurs in PhoP-repressed virulence genes, such as those specifying components of the Inv/Spa secretion apparatus. In this case, high concentrations of Mg²⁺, such as those in the extracellular environment, enhance the expression of PhoP-repressed genes, thereby promoting the initial steps in the invasion of epithelial cells by Salmonella.

Virulence in Salmonella demands a very large number of genes. Many of these genes were introduced into Salmonella as a result of gene transfer events and bestowed new abilities upon the microorganism. Such events suggest that the expanded capacities conferred by acquired genes can result in new pathogens and that gene transfer promotes the emergence of new bacterial species.

SUGGESTED READING

Baumler, A. J. 1997. The record of horizontal gene transfer in Salmonella. Trends Microbiol. 5:318-322.

Conner, C. P., D. M. Heithoff, S. M. Julio, R. L. Sinsheimer, and M. J. Mahan. 1998. Differential patterns of acquired virulence genes distinguish Salmonella strains. Proc. Natl. Acad. Sci. USA. 95:4641-4645.

Groisman, E. A., and H. Ochman. 1997. How Salmonella became a pathogen. Trends Microbiol. 5:343-349.

Hacker, J., G. Blum-Oehler, I. Muhldorfer, and H. Tschape. 1997. Pathogenicity islands of virulent bacteria: structure, function and impact on microbial evolution. Mol. Microbiol. 23:1089-1097.

Hardt, W. D., H. Urlaub, and J. E. Galan. 1998. A substrate of the centisome 63 type III protein secretion system of Salmonella typhimurium is encoded by a cryptic bacteriophage. Proc. Natl. Acad. Sci. USA 95:2574-2579.

Hensel, M., J. E. Shea, A. J. Baumler, C. Gleeson, F. Blattner, and D. W. Holden. 1997. Analysis of the boundaries of Salmonella pathogenicity island 2 and the corresponding chromosomal region of Escherichia coli K-12. J. Bacteriol. 179:1105-1111.

Hueck, C. J. 1998. Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol. Mol. Biol. Rev. 62:379-433.

Strauss, E. J., and S. Falkow. 1997. Microbial pathogenesis: genomics and beyond. Science 276:707-712.

Wood, M. W., M. A. Jones, P. R. Watson, S. Hedges, T. S. Wallis, and E. E. Galyov. 1998. Identification of a pathogenicity island required for Salmonella enteropathogenicity. Mol. Microbiol. 29:883-891.

Salmonella genome projects: http://genome.wustl.edu/gsc/bacterial/salmonella.shtml ; http://www.sanger.ac.uk/Projects/S.typhi/