Journal of Bacteriology, March 2007, p. 1489-1495, Vol. 189, No. 5
0021-9193/07/$08.00+0 doi:10.1128/JB.01730-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
| MINIREVIEW |

Jennifer L. Bishop,1,2,
Guntram A. Grassl,1,
and
B. Brett Finlay1,2*
Michael Smith Laboratories, University of British Columbia, Vancouver, BC, V6T 1Z4, Canada,1 Department of Microbiology and Immunology, University of British Columbia, Vancouver, BC, V6T 1Z4, Canada2
|
|
|---|
Typhoidal Salmonella serovars, such as Salmonella enterica serovars Typhi and Paratyphi, cause systemic illness that leads to an estimated 20 million cases and 200,000 deaths worldwide each year (5). In Southeast Asian countries, such as Vietnam, the prevalence of typhoidal salmonellosis is high, and patients often suffer from recurrent or relapsed infections (5, 38). Why the immune system is unable to mount a lasting protective response against typhoidal Salmonella species remains unknown. The work of Lucinda Thompson (Stanford University, Stanford, CA) aimed to address this issue by examining transcriptional signatures in the peripheral blood of serovar Typhi-infected patients from Vietnam. She described how salmonellosis patients in the acute, chronic, or convalescent stage of infection displayed distinct and temporal transcriptional profiles of immune genes. For example, profiles from acute typhoid cases had large upregulation of genes associated with increased neutrophil activity, while antiproliferative genes were downregulated. Over time, the acute signature approached that of uninfected individuals; however, some patients retained transcription profiles that indicated a more long-term effect of typhoid infection.
One of the foremost obstacles to administering effective treatment against Salmonella infection is antibiotic resistance among disease-causing strains (37). Studies conducted in the 1990s revealed significant increases in resistance to the most frequent antibiotic used against Salmonella, nalidixic acid (48). As the clinical outcome for patients infected with resistant strains is poor (13), new therapies must be sought out. In her talk, "Clinical Aspects of Enteric Fever," Christine Dolecek (University of Oxford, Oxford, United Kingdom) described how her group is developing a new evidence-based approach to testing anti-Salmonella drugs in clinical trials in Vietnam. Their primary goal is to make local clinical drug trials standardized, reproducible, and easily executed. In addition, information obtained from trials should be accessible, and sharing of information between groups should be encouraged. With these principles in mind, clinical trials are under way to determine drug efficacy and to establish dosing regimes for azithromycin and gatifloxacinnew-generation antibiotics effective against drug-resistant Salmonella (37).
Antibiotic resistance is also of critical concern in African countries, where multidrug-resistant nontyphoidal salmonellosis is one of the most common causes of bacteremia in children (13). Nontyphoidal salmonellosis is caused by Salmonella enterica serovars Typhimurium, Enteritidis, Newport, and Heidelberg and typically presents as self-limiting gastroenteritis, although in immunocompromised individuals, serious complications can ensue (20, 43). In recently published work, Sam Kariuki (KEMRI, Nairobi, Kenya) and colleagues presented data showing trends of drug resistance in nontyphoidal Salmonella (NTS) isolated from children in Kenya (25, 26). At the meeting, Kariuki described these results, showing that while drug resistance is on the rise in urban populations, NTS isolated from patients in rural areas was decreasingly resistant to amoxicillin and cotrimoxazole. In addition, Kariuki described epidemiological studies aimed at tracking the prevalence and presentation of nontyphoidal salmonellosis in various populations to determine a reservoir for these infections. The data highlighted the influence of socioeconomic factors in disease, as invasive NTS incidence was significantly higher in children from slum populations.
Human immunodeficiency virus (HIV)/AIDS patients in Africa are another high-risk population for contracting NTS infections. The mortality rate for HIV patients infected with NTS can be as high as 60%, and bacterial recrudescence occurs in up to 45% of these individuals (11, 24). In her talk, Melita Gordon (Liverpool University, Liverpool, United Kingdom) described how her group is investigating the intracellular persistence of NTS in macrophages derived from HIV patients. Her results show that, while macrophages from HIV patients show no defects in bacterial internalization or killing, macrophages primed with gamma interferon hyperproduce tumor necrosis factor alpha, interleukin 10 (IL-10), and IL-12 in response to NTS. Gordon's previously published results demonstrating the importance of gamma interferon in human monocyte-derived macrophages support her current findings and highlight the importance of using human cells for in vitro studies (12). Currently, Gordon's group is conducting further research to determine the mechanisms by which NTS-infected macrophages produce altered cytokine profiles in the context of HIV infection.
Molecular genotyping of NTS strains is fundamental to tracking disease-associated and drug-resistant strains in various populations. Currently, serotyping techniques and pulsed-field gel electrophoresis are used; however, the accuracy of these techniques is controversial (55). At the meeting, Mark Achtman (Max-Planck Institute for Infection Biology, Munich, Germany) addressed this issue and highlighted the difficulties associated with molecular typing of Salmonella isolates. According to Achtman, serotyping libraries are unreliable and no distinction can be made between serotypes isolated from different disease pathologies and hosts. Multilocus sequence typing (MLST) was described by Achtman as an alternative option to more accurately define Salmonella isolates. While MLST of closely related Salmonella isolates that share housekeeping genes can be problematic, the technique can be fine tuned to discriminate between sequences derived from stable temperate phages, which are more uniquely scattered throughout the genome (45). Strain discrimination also is more accurate when MLST is used in combination with pulsed-field gel electrophoresis or other typing techniques (17). While Achtman did point out that MLST should ultimately be replaced with even more sensitive techniques, such as small nucleic polymorphism identification, MLST has longevity in epidemiological research and is invaluable in studies such as those ongoing in Kenya and Malawi.
Paul Wigley (University of Liverpool, Wirral, United Kingdom) presented data on the roles of Salmonella pathogenicity island 1 (SPI-1) and SPI-2 in avian serovar Typhimurium infection. In chicks older than 4 days, infection by serovar Typhimurium results in chronic gastrointestinal colonization that remains until slaughter (42 to 49 days). Infection of newly hatched chicks leads to a severe systemic infection with high mortality. Using this model, Wigley found that while an SPI-1 mutant caused severe intestinal pathology, an SPI-2 mutant caused little or no intestinal pathology, had reduced bacterial loads in the gut, and was attenuated for systemic infection. This research has identified a novel role for SPI-2 in the systemic infection of newly hatched chicks. Furthermore, this model will allow researchers to elucidate the particular bacterial effectors and host mechanisms involved.
Andreas Baumler's group (University of California, Davis) has been interested in investigating why, in humans, serovar Typhimurium induces such a massive neutrophil influx into the intestine while serovar Typhi does not. Using the calf ileal-loop model of enterocolitis, he showed that serovar Typhimurium invades the follicle-associated epithelium at the tips of adsorptive villi, leading to flagellin-dependent chemokine production. IL-23 p19 levels were elevated after 2 hours, leading to IL-17 expression, which in turn stimulated GRO
expression. GRO
is a potent chemoattractant for neutrophils, and the massive influx of neutrophils helped to contain the infection but also added to the destructive pathology of the gut tissue. Similar to serovar Typhi infection of humans, and in contrast to their results with serovar Typhimurium, they observed little chemokine production and no neutrophil infiltration upon serovar Typhi infection of ileal loops. Baumler suggested that this difference is, at least in part, due to serovar Typhi's expression of the Vi antigen, as a serovar Typhi SPI-7 mutant (a Vi mutant) elicited elevated levels of the chemoattractant IL-8 in human monocytic THP-1 cells. In addition, expression of the Vi antigen in serovar Typhimurium led to reduced IL-17 production in ileal loops. Thus, it seems likely that the Vi antigen counterbalances pathogen-mediated activation of host responses, enabling serovar Typhi to evade Toll-like receptor-mediated innate responses (44).
Wolf-Dietrich Hardt (Institut für Mikrobiologie, Zurich, Switzerland) pioneered the streptomycin-pretreated-mouse model and has used it to explore the factors that contribute to triggering Salmonella-induced acute intestinal inflammation in vivo. Using both steptomycin-pretreated mice and a novel calf ileal-loop model, Bryan Coburn, from Brett Finlay's group (University of British Columbia, Vancouver, Canada), presented his data describing a novel role for SPI-2 in intestinal disease (4). Hardt's group is now investigating the role of dendritic cells (DCs) in the streptomycin-pretreated-mouse model using the recently described method of depleting CD11c+ DCs with diphtheria toxin (23). Upon infection with wild-type serovar Typhimurium, CD11c+ DC-deficient mice developed intestinal inflammation; however, no inflammation was detected when mice were infected with an SPI-1 mutant. Furthermore, in this model, the SPI-1 mutant colonized the gut but not systemic sites of infection. In contrast, delaying depletion of CD11c+ DCs until 18 and 40 h after infection with the SPI-1 mutant led to normal levels of bacterial uptake to systemic sites and intestinal pathology. Thus, Hardt concluded that DCs are crucial for the uptake of SPI-1 mutant bacteria but are not responsible for the development of pathology in the streptomycin-pretreated-mouse model.
Duncan Maskell and Pietro Mastroeni (both from the University of Cambridge, Cambridge, United Kingdom) study the in vivo pathogenesis of serovar Typhimurium and how it impacts immunity. Maskell demonstrated that, in contrast to infections in tissue culture, where salmonellae grow to high numbers within macrophages, they observed only one or two bacteria per phagocyte in vivo. Using the mouse typhoid model, he showed that within tissues, bacteria are in foci that grow in size and number as the infection spreads, and these foci are spatially and functionally independent (49). This work highlights the potential disparity between in vitro and in vivo findings and emphasizes the need to understand how Salmonella escapes and spreads within infection foci.
Mastroeni demonstrated that coinfection with wild-type serovar Typhimurium and an attenuated strain given as late as 2 days postinfection exacerbates the growth of wild-type bacteria in the mouse typhoid model. They believe this effect was mediated by IL-10, as levels were elevated in serum shortly after intravenous infection. Mastroeni also suggested that during the spread of the bacteria to new foci, bacteria will be present extracellularly, leading to opsonization and targeting to Fc receptors. Using bone marrow-derived macrophages, he showed that opsonization with immune serum increased the oxidative burst but did not influence phagolysosome fusion. This effect was mediated mainly through Fc
RI. Fc
RI/II/III triple-knockout mice were still susceptible even after immunization with attenuated bacteria that induced otherwise protective Th1 and antibody responses.
Serovar Typhimurium colonizes, replicates, and persists in the gut of Caenorhabditis elegans, making C. elegans an attractive model system to study both the host and bacterial factors that are required for persistent infection (1). Accordingly, Rosanna Alegado, from Man-Wah Tan's group (Stanford University, Stanford, CA), screened a serovar Typhimurium transposon library for mutants that failed to persist in the nematode gut. She identified 18 genes, 4 of which have known roles in the murine typhoid model of infection and 14 of which have no known role in an animal model. One class of mutants identified in the screen displayed sensitivity to osmotic stress and antimicrobials. Using these mutants, persistence defects could be rescued when antimicrobials were knocked down in the worm. Another class of mutations showed defects in biofilm growth, implying that biofilm formation is important for persistence in the worm.
Bile affects the expression of serovar Typhimurium genes that are important for virulence, and this has been proposed to enhance colonization and persistence within the gallbladder (40). Serovar Typhimurium forms biofilms on the surfaces of human gallstones, which may contribute to the development of the carrier state (41). Robert Crawford, from John Gunn's group (Ohio State University, Columbus), developed an in vitro assay of biofilm formation on human cholesterol, a model used to mimic the human gallstone. Not only did serovar Typhimurium form biofilms on human cholesterol, he found that bile enhanced biofilm formation in his assays. Cellulose and colanic acid are normally key constituents of serovar Typhimurium biofilms; however, Crawford found that serovar Typhimurium mutants incapable of making these exopolysaccharides (EPSs) formed robust biofilms on human cholesterol. This suggested that a novel EPS was involved in gallstone biofilms. The novel EPS was found to be encoded in the yih operon and was shown to be transcriptionally activated by bile. Production of the novel EPS was enhanced by bile in the in vitro assay of biofilm formation on human cholesterol and was detected on human gallstones.
In vitro, Salmonella replicates to high numbers within epithelial cells and macrophages, yet in vivo, infected cells usually contain only one or two bacteria (3, 32, 49). Years ago, Francisco Garcia-del Portillo noticed that Salmonella do not replicate or cause cytotoxicity in fibroblasts but simply remain in a persistent state (31). At this meeting, he presented evidence that persistence in fibroblasts occurs in vivo, with images of serovar Typhimurium targeting nonphagocytic (CD45/) cells in the lamina propria. Persistence in fibroblasts may be key to understanding why chronic Salmonella infections are not cleared by the immune system, since the host has no means of detecting bacteria hiding within these cells.
Denise Monack (Stanford University, Stanford, CA) recently described a model of serovar Typhimurium persistence in nrampr mice (33) and, together with Trevor Lawley, identified bacterial genes that contribute to long-term systemic infection (28). Their new work describes a mouse model that mimics the natural fecal-oral route of transmission in nrampr mice. They found that naïve mice, housed with mice that had been orally infected, became infected rapidly and to the same extent as the "seeder" mice. They assessed the role of SPI-1 in transmission and found that, although shed in high numbers, a
sipB mutant was incapable of being transmitted to naïve mice. This system provides a basis on which one can genetically dissect both pathogen and host factors required for horizontal transmission.
Francisco Garcia-del Portillo's group (Centro Nacional de Biotecnologia, Madrid, Spain) investigated whether the signaling pathways involved in fibroblast invasion were the same as for epithelial cell invasion. Surprisingly, they found that serovar Typhimurium lacking SPI-1 secretion (
SPI-1) was able to invade fibroblasts efficiently. Upon closer examination, they observed that
SPI-1 strains induced filopodium-like splashes on the surfaces of fibroblasts. Fibroblast invasion by
SPI-1 required tyrosine phosphorylation, phosphatidylinositol 3 kinase, MEK, actin, and microtubules. However, unlike epithelial cells, fibroblast invasion was Rac1, Cdc42, and Rho independent. This work is important, since until now, the mechanisms involved in invasion of all nonphagocytic cells were assumed to be the same.
Brad Cookson (University of Washington, Seattle) described mechanisms of Salmonella-induced pyroptosisa caspase 1-dependent proinflammatory cell death. Infection of J774 macrophages with serovar Typhimurium resulted in caspase 1-dependent DNA fragmentation, poly(ADP-ribose) polymerase activation, pore formation (1.1- to 2.4-nm diameter), and subsequent osmotic lysis. Swelling of serovar Typhimurium-infected cells could be inhibited by pretreatment with glycin (which blocks Fe fluxes) or yvad (which blocks caspase 1) or by infection with the
prgH SPI-1 mutant. Cell lysis was not required for release of the caspase 1-activated cytokines IL-1ß and IL-18, but pore formation was temporally associated with it. This work has since been published (10).
From within the Salmonella-containing vacuole (SCV), serovar Typhimurium subverts host cell trafficking in order to persist inside the host cells (27). Rab family GTPases regulate membrane-trafficking events (14) and are therefore key targets for hijacking by serovar Typhimurium. John Brumell (Hospital for Sick Kids, Toronto, Canada) compared wild-type SCVs to "model phagosomes" and screened for differences in Rab GTPase acquisition using a Rab array consisting of 43 different members. Rab GTPases found to be exclusive to "model phagosomes" (e.g., Rab 8, 13, 32, and 35) were suggested to be involved in phagolysosomal fusion. Six Rab GTPases were exclusively acquired by the SCV and not by "model phagosomes." Rab9 is classically known to be involved in the recycling of mannose-6-phosphate receptors from late endosomes (14). Thus far, functional studies of the SCV-specific Rab GTPases has revealed that Rab9 is present on Salmonella-induced filaments and is required for their formation. These studies may provide clues as to how serovar Typhimurium alters host cell trafficking, enabling it to persist and replicate intracellularly.
RpoS is an RNA polymerase alternative sigma factor that is required for the survival of bacteria under various stresses, including low-nutrient conditions, high osmolarity, low pH, and oxidative pressure (18). RpoS in Salmonella is also critical for the regulation of spv genes, which are important for systemic infection in mice (9, 15). Eduardo Groisman (Washington University, St. Louis, MO) described how the PhoP-PhoQ two-component system stabilizes RpoS in S. enterica. RpoS accumulates as PhoP is activated under low-Mg2+ conditions. PhoP stabilizes RpoS in a regulatory cascade that begins with the transcriptional activation of iraP. Interaction between IraP and the target protein RssB prevents degradation of RpoS by the ClpXP protease. Thus, PhoP controls RpoS-regulated spv genes, which are essential for virulence in the mammalian host (56).
Histone-like nucleoid structuring protein (H-NS) is a bacterial nucleoid-associated protein that acts as a transcriptional repressor that alters DNA structure by recognizing and binding curved DNA (7). Ferric Fang (University of Washington, Seattle) and Jay Hinton (Institute for Food Research, Norwich, United Kingdom) both reported that H-NS has a high affinity for AT-rich DNA sequences and oligomerizes head to tail in these areas to increase binding (30, 35). Because high AT content is a hallmark of laterally transferred genes and these sequences are normally found outside of the regulatory network, H-NS acts in pathogens like Salmonella to silence newly transferred genes (xenogeneic silencing) until they are required. Specific activation of horizontally acquired virulence sequences can be achieved by antisilencing, whereby H-NS is displaced by other DNA binding proteins, like SlyA. Hinton and Fang showed that H-NS silences all SPIs and spv. Accordingly, bacteria with a mutated hns gene upregulate SPI-2 and spv genes and downregulate flagellar chemotaxis (30, 35). In fact, constitutive expression of SPI-2 in H-NS mutants has resulted in difficulties in manipulating these strains due to poor growth. Based on these results, H-NS is thought to play a major role in the evolution of virulence by enabling the acquisition and maintenance of foreign DNA.
aroC/
ssaV mutant in healthy Western subjects to individuals in developing countries is a significant challenge (52). However, new trials testing this vaccine strain in adults and children from developing countries is under way, and if successful, it could be the first single-dose oral vaccine for typhoid fever available worldwide. Development of a live oral serovar Typhi vaccine has been difficult, since little information is available regarding the mechanisms behind protective immunity and immunological memory for serovar Typhi or the interaction between the bacterium and the gut microenvironment. In his talk, Mark Sztein (University of Maryland, Baltimore) showed that B-cell production of soluble antibodies, classical and HLA-E-restricted cytotoxic-T-lymphocyte killing of infected cells, cell-mediated immunity, CD4+ and CD8+ T-cell homing to the gut, and distinct subsets of memory T cells are essential for protective immunity against serovar Typhi infection (46, 47). In addition, he underscored the importance of examining the carrier state in the development of typhoid vaccines. Importantly, we must ask whether vaccines that confer protective immunity in vaccinated individuals are also effective in preventing transmission from carriers.
Another obstacle in the development of vaccines is the variety of typhoidal and nontyphoidal salmonelloses caused by various serovars and strains. Because vaccinology relies so heavily upon immunity to specific antigens, it is imperative that studies be designed to look at the degree of cross-protection that vaccines offer between different epidemiologically important salmonellae. In his talk, Mike Levine (University of Maryland, College Park) described work that is ongoing in Chile and Egypt, two areas where typhoid fever is often caused by serovar Paratyphi (19, 51). Two strains of serovar Paratyphi, A and B, exist in distinct distributions throughout the world. Because one in four cases of typhoid fever is caused by serovar Paratyphi rather than serovar Typhi (5), cross-protection between serovar Typhi vaccines and serovar Paratyphi would be optimal. According to Levine, current trials using the Ty21A vaccine in Chile and Egypt have shown that cross-protection can be seen in 58% of cases where individuals are infected with serovar Paratyphi B, whereas no protection is gained against serovar Paratyphi A (50). Levine suggested that the cross-protection between the serovar Typhi vaccine and serovar Paratyphi B is due to shared epitopes of O antigens and cell-mediated immune responses to protein antigens in the live vaccine.
Contamination of feed livestock, such as chickens, is a major source of Salmonella infections. As such, vaccines against Salmonella in these animals are an important step in preventing the spread of infection to humans. Inhibiting colonization of chicken intestines with virulent Salmonella via vaccination with mutant Salmonella vaccine strains was discussed by Vanessa Eeckhaut (Ghent University, Merelbeke, Belgium). Her group found that oral vaccination of broiler chicks on the day of hatching with a serovar Enteritidis hilA mutant significantly reduced colonization of the chicks by a virulent strain fed to the animals 1 day later. Importantly, the hilA mutant strain was cleared within 1 month of infection, which is critical for preventing transmission of the vaccine strain to humans during consumption. However, as levels of the vaccine strain decreased, virulent Salmonella increased. These studies demonstrated that in order to produce an effective vaccine for chickens that is based on colonization inhibition, an appropriate attenuated strain must be found which colonizes animals long enough to provide lasting inhibition of virulent Salmonella but is still cleared before slaughter.
Vaccine strains of Salmonella could also possibly be used in the control of various proinflammatory autoimmune diseases (21, 53, 54). Recently, David Pascual's group (Montana State University, Bozeman) has shown that an attenuated Salmonella vaccine strain increases the number of T-regulatory cells in an experimental mouse autoimmune encephalomyelitis model and that these cells are responsible for diminishing clinical disease (22). In addition, the vaccine, which expresses colonization factor AgI (CFA/I), is anti-inflammatory, induces the production of Th2 T cells, and reduces the production of proinflammatory cytokines from macrophages (22). Other subsets of T cells, such as cytotoxic and memory T cells, are also stimulated by Salmonella vaccines.
Holger Russmann (University of Munich, Munich, Germany) showed that mice challenged with tumor cells expressing the Listeria monocytogenes major histocompatibility class I-restricted peptide p60 and immunized with Salmonella that translocate a chimeric p60 protein via its type III secretion system have increased levels of p60-specific CD8 and memory T cells. Induction of this cell set was associated with prevention of fibrosarcoma development in the mice (36). While more studies need to be carried out to determine the precise functions of Salmonella antigens in the maintenance of protection in these model systems, these studies highlight the potential for using Salmonella-based vaccines for the treatment of various common and severe illnesses, such as inflammatory disease and cancer.
Published ahead of print on 22 December 2006. ![]()
E.C.B., J.L.B., and G.A.G. contributed equally to this work. ![]()
|
|
|---|
S (RpoS) subunit of RNA polymerase. Microbiol. Mol. Biol. Rev. 66:373-395.
neutralization. J. Exp. Med. 199:231-241.
aroC
ssaV derivatives as vehicles for delivering heterologous antigens by chromosomal integration and in vivo inducible promoters. Infect. Immun. 73:362-368.This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»