OUP user menu

Salmonella's long-term relationship with its host

Thomas Ruby, Laura McLaughlin, Smita Gopinath, Denise Monack
DOI: http://dx.doi.org/10.1111/j.1574-6976.2012.00332.x 600-615 First published online: 1 May 2012


Host-adapted strains of Salmonella enterica cause systemic infections and have the ability to persist systemically for long periods of time and pose significant public-health problems. Multidrug-resistant S. enterica serovar Typhi (S. Typhi) and nontyphoidal Salmonella (NTS) are on the increase and are often associated with HIV infection. Chronically infected hosts are often asymptomatic and transmit disease to naïve hosts via fecal shedding of bacteria, thereby serving as a critical reservoir for disease. Salmonella utilizes multiple ways to evade and modulate host innate and adaptive immune responses in order to persist in the presence of a robust immune response. Survival in macrophages and modulation of immune cells migration allow Salmonella to evade various immune responses. The ability of Salmonella to persist depends on a balance between immune responses that lead to the clearance of the pathogen and avoidance of damage to host tissues.

  • Salmonella
  • persistence
  • systemic infection
  • chronic carrier


Salmonella are enteric bacteria that are a major cause of infectious diseases throughout the world. These bacteria infect both humans and other animals and are a common cause of zoonotic disease. Salmonella serovars are responsible for human diseases ranging from gastroenteritis to systemic infections. Systemic Salmonella infection is usually host-dependent, and Salmonella enterica servar Typhi (S. Typhi) causes only systemic infection – typhoid fever – in humans.

Salmonella typhi and Salmonella enterica serovar Paratyphi (S. Paratyphi) are important human pathogens of immense concern to public health and with considerable economic impact. They are endemic in regions of the world where drinking water quality and sewage treatment facilities are poor (House et al., ; Young et al., ) and infections remain difficult to treat by antibiotic therapy because of the increasing frequency of resistant bacteria (Wain et al., ; Kingsley et al., ).

A significant percentage (1–6%) of patients with typhoid become chronic carriers of S. Typhi, as do many people who have never had a clinical history of typhoid fever (Ledingham & Arkwright, ; Stokes & Clarke, ; Levine et al., ). These individuals shed bacteria in their stools and urine for periods of time that range from a year to a lifetime, without any apparent signs of disease (Vogelsang & Boe, ). Typhoid carriers are of special concern from a public-health viewpoint as they are the reservoirs for the spread of infection and disease. Two well-known cases of typhoid carriers are ‘Typhoid' Mary Malon who was responsible for the death of 26 people in the US and an unnamed milk carrier in South England in the late 1800s (Mortimer, ). More recent cases have been studied in greater details, showing that typhoid carriers remain an important public-health concern today (Sloan et al., ; Merselis et al., ; Levine et al., ; Lin et al., ).

From the bacterial perspective, the ability of host-adapted strains to cause a persistent infection is important for microbial survival and transmission as these bacteria act as the reservoir. Salmonella typhi is carried for years – even in the presence of an immune response – and chronic carriers of S. typhi have high levels of circulating serum antibodies to the Vi antigen and to flagellar antigens (Bao et al., ; House et al., ). Investigating host–pathogen interactions during the chronic carrier state in salmonellosis should provide insights into bacterial survival mechanisms, as well as provide information that could be used to develop new approaches for the treatment for typhoid and other persistent microbial infections.

Salmonella infection

In natural infection, Salmonella are typically acquired from the environment by oral ingestion of contaminated water/food or by contact with a carrier. Following ingestion in sufficient numbers, a proportion of the inoculum survives the low pH environment of the stomach to enter the small intestine where infection can be established.

There is evidence that some Salmonella have a preference to exploit the microfold (M) cells that are specialized epithelial cells that sample the antigenic content of the gut (Jensen et al., ; Jepson & Clark, ; Halle et al., ). After Salmonella spp. penetrate the epithelial barrier, they preferentially infect phagocytes within the lamina propria. In Salmonella gastroenteritis, the infection is usually self-limiting and does not proceed beyond the lamina propria. However, in host-adapted salmonellosis, such as typhoid fever, Salmonella-infected phagocytes gain access to the lymphatics and bloodstream, allowing the bacteria to spread to the liver and the spleen (Vazquez-Torres et al., ), and can persist in the gall bladder and bone marrow (Sinnott & Teall, ; Wain et al., ; Fig. ).

Figure 1

Persistent Salmonella infection. Schematic representation of persistent infection with S. enterica serovar Typhi in humans. Bacteria enter the Peyer's patches of the intestinal tract mucosal surface by invading M cells. This is followed by inflammation and phagocytosis of bacteria by neutrophils and macrophages and recruitment of T and B cells. In systemic salmonellosis, such as typhoid fever, Salmonella may target specific types of host cells, such as DC and/or macrophages (Mp) that favor dissemination through the lymphatics and blood stream to the MLNs and to deeper tissues. This then leads to transport to the spleen, bone marrow, liver, and gall bladder. Bacteria can persist in the MLNs, bone marrow, and gall bladder for life, and periodic reseeding of the mucosal surface via the bile ducts and/or the MLNs of the small intestine occur, and shedding can take place from the mucosal surface.

Recently, S. Typhi was found in biofilms detected on gallstones from 5% of patients enrolled for gallbladder removal in Mexico City (Crawford et al., ). Furthermore, gallstones have been shown to play a significant role in Salmonella persistence, as induced gallstone formation in mice resulted in enhanced fecal shedding and enhanced colonization of gall bladder tissue (Crawford et al., ).

Mouse model of Salmonella persistence

A mouse model of persistent infection is characterized by sporadic excretion of bacteria in stools and long-term carriage of S. enterica serovar Typhimurium (S. Typhimurium) in low numbers within classical granulomatous lesions, which arise in the spleen, liver, and mesenteric lymph nodes (MLN; Monack et al., ). The data obtained from studies using persistently infected mice indicate that the most common site of chronic carriage of S. Typhimurium is the MLNs. Indeed, this is often the only site from which viable Salmonella can be recovered. Interestingly, surgical removal of the MLN results in increased numbers of S. Typhimurium bacteria reaching systemic sites early after infection, suggesting that the MLN provide a vital barrier, shielding systemic compartment dissemination of S. Typhimurium (Voedisch et al., ; Griffin & McSorley, ).

The genetic background of the mouse strains plays a pivotal role in determining Salmonella persistence (Hormaeche, ). Nramp1 is a genetic locus originally identified as being a critical factor in host defense against intracellular pathogens including Leishmania, Mycobacteria, and Salmonella. The Nramp1 gene codes for an ion transporter and is expressed primarily in macrophages and dendritic cells (DC). Commonly used mouse strains such as C57BL/6 carry two point mutations (G105A, G169D) resulting in increased susceptibility to intracellular pathogens (Vidal et al., ; Govoni et al., ). As a result, chronic infection models in C57BL/6 mice require attenuated Salmonella strains. A commonly used strain is the Salmonella auxotrophic mutant aroAaroD which is unable to make the required aromatic amino acids such as p-amino benzoic acid. These amino acids cannot be obtained from mammalian tissue, and as a result, this strain is avirulent in mice. A paper describing this attenuated strain as a live vaccine was first published in 1980 (Hoiseth & Stocker, ). Since then, it has been used to study the adaptive immune response to Salmonella in an intravenous model of infection (Salazar-Gonzalez et al., ; Srinivasan et al., ; Jackson et al., ).

Wild-type or virulent S. Typhimurium strains only produce a chronic infection in Nramp1wt/wt mice or their F1 hybrids (Monack et al., ; Johanns et al., ). Oral inoculation of up to 108 CFU of wild-type S. Typhimurium results in a chronic infection where bacteria can be found in the MLN up to 1 year postinoculation (Monack et al., ). Table  details the Nramp1 status of some commonly used mouse strains (Vidal et al., ; Malo et al., ; Govoni et al., ). Figure  outlines the different Salmonella infection models and the relative bacterial loads in the spleen and highlights the utility of the Nramp1 wild-type mice for studying host–pathogen interactions during long-term Salmonella infections. Possible mechanisms underlying the importance of Nramp1 to persistence are discussed later in this review.

Figure 2

Models of Salmonella persistent infection. Range of host genetic background, Salmonella strains, and route of infection in different models of persistent Salmonella infection. The data were taken from the specific references mentioned in table. Strain SL1344 SseD:aphT was first described in Klein (2000). Routes of infection are indicated as follows: IV, intravenous; IP, intra-peritoneal; OG, oro-gastric; sm, streptomycin treatment.

View this table:
Table 1

Nramp1 status of some commonly used mouse strains

Mouse strainNramp1 status
C57BL/6 Nramp1 mt/mt
BALB/c Nramp1 mt/mt
DBA Nramp1 wt/wt
C3H/HeJ Nramp1 wt/wt
CBA Nramp1 wt/wt
129 SvJ Nramp1 wt/wt
  • Mt, mutated allele; wt, wild-type allele.

Salmonella Typhi infection in people is often accompanied by diarrheal symptoms that have typically been difficult to model in mice. In a mouse model of infection with S. Typhimurium, antibiotic treatment before infection results in increased occurrence of diarrheal symptoms including gut inflammation and increased gastrointestinal bacterial burden in both the chronic model (Lawley et al., ) and the acute infection model (Barthel et al., ; Endt et al., ). In addition, Salmonella has been found in the gastrointestinal tract and systemic sites following infection of antibiotic-treated mice (Barthel et al., ).

Finally, we should consider the humanized mouse model of S. Typhi infection (Song et al., ) where Rag2−/−Υc−/− mice are irradiated at birth and engrafted with human fetal liver stem cells. Infection with S. Typhi results in a chronic infection with the human cells mounting both an innate and adaptive immune response. One month postinoculation, S. Typhi was found in the spleen, liver, and more variably, in the gall bladder. Within these tissues, the bacteria were mainly found in the human hematopoietic (CD45+) cells (Song et al., ). While this model might be useful for testing Salmonella vaccines, the inability of human cytokines to act on mouse tissue resulted in a lack of systemic pathology associated with chronic infection.

In addition, these models have been used to study the role of Salmonella effectors. A chronic infection state is beneficial to the bacteria as it ensures a long-lasting niche and subsequent transmission to a new host. Salmonella plays an active role in the establishment of such a hospitable environment. The bacteria manipulate the host responses through the secretion of effector proteins into the host cells. These effectors can stimulate or repress signaling cascades, sense the intracellular environment, and modify this environment in order to establish an intracellular niche.

Salmonella factors that contribute to persistent infection

The Salmonella factors that contribute to this pathogen's virulence have been extensively studied and reviewed elsewhere (Haraga et al., ; Ibarra & Steele-Mortimer, ; McGhie et al., ; Valdez et al., ; Table ); however, it is unclear which of these or other bacterial factors are required specifically for persistent Salmonella infection. Much effort has been concentrated on the secreted effector proteins that depend on the type III secretion systems encoded by Salmonella pathogenicity islands 1 and 2 (SPI1 and SPI2, respectively), and yet the extent to which these pathogenicity islands actually contribute to persistent infection remains uncertain. Clearly, a variety of Salmonella factors that are independent of SPI1 and SPI2 also play significant roles in persistent infection. Therefore, we will discuss both SPI1- and SPI2-dependent and independent factors that have been reported to play a role during the persistent phase of Salmonella infection. Because much of the work investigating persistent Salmonella infection has utilized S. Typhimurium, this section of the review will concentrate on the bacterial factors from this serovar (Table ).

View this table:
Table 2

Salmonella factors contributing to persistent infection

Salmonella factorAssociated functionPersistence modelReferences
SPI1Host cell invasionC. elegans, 129X1/SvJ miceErickson & Detweiler (), Gal-Mor et al. (); Ibarra & Steele-Mortimer ()
SipB, SipC, SipDType III secretion system129X1/SvJ miceErickson & Detweiler ()
SPI2Intracellular survival or dynamicsC. elegans, 129X1/SvJ miceErickson & Detweiler (), Gal-Mor et al. (), Ibarra & Steele-Mortimer (); Griffin & McSorley ()
SseK2, SseISecreted effectors129X1/SvJ mice (SseK2 in screen only)Erickson & Detweiler (); Griffin & McSorley ()
Adhesive factors
ShdA, MisLFibronectin bindingCBA/J mice, Balb/C mice with aroA mutant S. TyphimuriumHoiseth & Stocker (), Hensel et al. (), Halle et al. (), Haraga et al. (); Helaine et al. ()
StcC, StcD, SthA, BcfD, StbD, StdAFimbrial proteinsCBA/J mice, 129X1/SvJ mice (screen only)Hormaeche (); Erickson & Detweiler ()
PhoPQSensor/kinase system C. elegans Ibarra & Steele-Mortimer ()
Mig-14Antimicrobial peptide resistance129X1/SvJJones & Falkow ()
VirK129X1/SvJKillar & Eisenstein ()
PgtE129X1/SvJ (screen only)Erickson & Detweiler ()
RcsC, YdeI129Sv6Kim et al. ()
SodC1Superoxide dismutase129X1/SvJ (screen only)Erickson & Detweiler ()
SspJResistance to superoxide anionC3H/HeNKingsley et al. ()
HmpFlavohemaglobin (resistance to nitrosative stress)129X1/SvJKingsley et al. ()
AceAIsocitrate lyaseC3H/HeNMastroeni et al. ()
  • Wild-type S. Typhimurium strains were used for all persistence models unless noted otherwise.

  • ‘Screen only' denotes S. Typhimurium factors whose evidence for contribution to persistent Salmonella infection in 129X1/SvJ mice is limited to negative selection in the Salmonella persistence screen (Erickson & Detweiler, ).

SPI1 and SPI2

The pathogenicity islands SPI1 and SPI2 encode two distinct type III secretion systems that each translocates a specific group of bacterial effector proteins into host cells. SPI1-dependent translocation of bacterial invasion factors into host epithelial cells enables S. Typhimurium to penetrate the small intestines and Peyer's patches (Galan, ), accomplishing an initial step of S. Typhimurium dissemination from the gastrointestinal tract to host systemic sites (including the MLN, spleen, liver, and bone marrow). In order to search for Salmonella factors that contribute to later stages of infection (persistent infection), a negative selection screen was undertaken by our group (Lawley et al., ) utilizing a transposon-mutagenized library of S. Typhimurium and the Nramp1r wild-type mouse strain 129X1/SvJ (Lawley et al., ). Although the mice were inoculated intraparatoneally (IP) instead of orogastrically (a route that bypasses the significant bottleneck that occurs during dissemination from the gastrointestinal tract to systemic sites (Meynell & Stocker, ; Mecsas et al., ), the screen revealed that SPI1 was necessary to sustain a persistent infection for at least 1 month postinfection (Lawley et al., ). The SPI1 genes confirmed to be required for persistent infection in this study included the invasion and translocation effectors SipB, SipC, and SipD (Lawley et al., ); however, it is possible that additional SPI1 effectors also could contribute to persistent infection. In light of the fact that S. Typhimurium is extruded from dying cells in the epithelia throughout an infection (Knodler et al., ), it is likely that the ability of S. Typhimurium to continuously re-invade epithelial tissues is necessary to sustain a persistent infection. In support of this notion, SPI1 is also required for a persistently infected mouse to transmit S. Typhimurium to naïve cage mates (Lawley et al., ).

During a persistent Salmonella infection, the bacteria survive within a privileged niche inside of the Salmonella containing vacuole (SCV) within phagocytes such as macrophages (Monack et al., b). Recently, Detweiler and colleagues showed that during persistent infection, S. Typhimurium preferentially grows inside of hemophagocytic cells, a specific macrophage subtype (Silva-Herzog & Detweiler, ). Survival inside of these cells required both SPI1 and SPI2 (Silva-Herzog & Detweiler, ). SPI2 is expressed by intracellular S. Typhimurium contained within the SCV and is required to translocate a group of bacterial effectors into the host cell cytosol in order to maintain intracellular survival (Jones & Falkow, ; Ochman et al., ; Cirillo et al., ; Hensel et al., ). In addition, the Salmonella persistence screen described above also identified several SPI2 transposon mutants, indicating these SPI2 mutants were not able to establish a persistent infection (Lawley et al., ). Most of these mutants were required at both early (1 week postinfection) and late (2–4 weeks postinfection) stages of infection, which is expected of bacterial factors required for intracellular survival. However, two SPI2-dependent effectors, SseK2 and SseI (SrfH), were unique, in that they were only required for late (persistent) stages of infection (Lawley et al., ). Although SseI is not required for intracellular survival of S. Typhimurium, SseI is continuously required to maintain a persistent infection in mice (McLaughlin et al., ). Instead of directly supporting intracellular survival of Salmonella, SseI translocation into phagocytes suppresses the ability of these cells to adhere (Worley et al., ) and migrate efficiently, thereby disrupting their ability to effectively communicate with other cells of the host's immune system (McLaughlin et al., ). SseI inhibits migration in part by associating with the host protein IQGAP1, an important regulator of the cytoskeleton and cell migration (McLaughlin et al., ; Kim et al., b). Although phagocyte carriage of intracellular S. Typhimurium from the gastrointestinal tract to systemic sites is required for establishing a systemic infection (Bogunovic et al., ; Voedisch et al., ), these results show that the speed of such dissemination may influence the host's adaptive immune response in such a way that controls whether a persistent Salmonella infection is permitted to occur. Indeed, mice infected with sseI mutant S. Typhimurium had increased numbers of DC and CD4+ T cells in the their spleens 12 days postinfection as compared to mice infected with wild-type S. Typhimurium (McLaughlin et al., ). Although SseK2 contributes to virulence during the early stages of infection of 129/svlmJ mice (Brown et al., ), the function of SseK2 (as well as other SPI2-dependent factors) during persistent Salmonella infection remains unknown and is the subject of future studies. SPI2 is also required for the transmission of S. Typhimurium from persistently infected mice to naïve cage mates (Lawley et al., ). While both SPI1 and SPI2 certainly contribute to persistent S. Typhimurium infection, it should also be noted that SPI1 and SPI2 are not absolutely required to establish a persistent infection at systemic sites because SPI1- and SPI2-defective mutants of S. Typhimurium (sipB and ssaV, respectively) still colonize and persist in systemic tissues, albeit at lower levels (Lawley et al., ).

Salmonella factors required for intestinal persistence

Sustained colonization of the gastrointestinal tract is another important aspect of persistent Salmonella infection because it is necessary for live bacteria to be passed in the feces and eventually transmitted to other potential hosts (Lawley et al., ). Using the CBA mouse strain (Nrampr, S. Typhimurium-resistant), several Salmonella adherence factors were shown to be important for intestinal persistence of S. Typhimurium. Baumler and colleagues discovered that the fibronectin-binding factors ShdA (Kingsley et al., , , b) and MisL (Dorsey et al., ) both contribute significantly to persistence in the gastrointestinal tract and fecal shedding of live bacteria. This group later showed that several fimbrial operons (encoding outer surface structures with possible adhesive functions), including lpf, bcf, stb, stc, std, and sth, were also required for long-term intestinal carriage and fecal shedding (Weening et al., ). In the Salmonella persistence screen described above, several genes from these operons were negatively selected, including stcC, stcD, sthA, bcfD, stbD, and stdA (bolded genes were negatively selected in both screen replicates), indicating a link between intestinal persistence and systemic persistence (Lawley et al., ). For example, the antivirulence modulator ZirTS is expressed only by S. Typhimurium colonizing the gastrointestinal tract (as well as in the feces of persistently infected mice), and such expression negatively impacts systemic colonization in Nramp1r mice (Gal-Mor et al., ). How ZirTS expression affects systemic persistence requires further investigation.

Salmonella Typhimurium factors that protect against host-derived antimicrobial peptides also contribute to persistent Salmonella infection. For example, Alegado & Tan () recently showed that several Salmonella factors were required for S. Typhimurium carriage in the intestines of Caenorhabditis elegans, including SPI2, pSLT (virulence plasmid), and PhoPQ (a sensor/kinase system that senses antimicrobial peptides (Bader et al., ), acidic pH (Bearson et al., ), and changes in metal cation concentration (Groisman et al., ). The PhoPQ regulon encompasses genes that protect against antimicrobial peptides such as polymixin B, C18G, and CRAMP, such as pgtE and mig-14 (Guina et al., ; Brodsky et al., ; Alegado & Tan, ). PgtE, a modifier of the bacterial outer surface membrane (Guina et al., ), was also identified in one of the Salmonella persistence screens carried out by Lawley et al. (). Mig-14 (Brodsky et al., ), along with several other Salmonella factors providing resistance toward antimicrobial peptides [VirK (Detweiler et al., ), RcsC (Detweiler et al., ; Erickson & Detweiler, ), and YdeI (Erickson & Detweiler, )], also contribute to persistent Salmonella infection. These results demonstrate that Salmonella resistance toward antimicrobial peptides of the host immune response is critical for Salmonella to maintain a persistent infection.

Salmonella factors coping with oxidative and nitrosative stress

During a persistent Salmonella infection, bacteria are predominantly found within the SCV inside of immune cells such as macrophages (Monack et al., b). In addition to antimicrobial peptides, these cells produce reactive oxygen species (ROS, including superoxide anion, hydrogen peroxide, and hydroxyl radical) and reactive nitrogen species (RNS, including nitric oxide and peroxynitrite) in order to control the invading pathogen (Vazquez-Torres & Fang, ). In fact, Nramp1r mice that are deficient in inducible nitric oxide synthase (iNOS) cannot be persistently infected with Salmonella and die during the acute phase (Mastroeni et al., b).

Thus, several S. Typhimurium factors that specifically cope with ROS and RNS also contribute to the ability of Salmonella to persist within the host. sodC1 (superoxide dismutase) encodes an enzyme required for the detoxification of superoxide anion and is required for S. Typhimurium virulence in vivo (Fang et al., ; Krishnakumar et al., ). This sodC1 gene was also identified in the negative selection screen described previously (Lawley et al., ). In addition, a superoxide-sensitive mutant of S. Typhimurium (sspJ) is attenuated at systemic sites (including the MLN and liver) 19 days postinfection (van Diepen et al., ). hmp (flavohemoglobin) catalyzes the detoxification of nitric oxide and is required to maintain a persistent Salmonella infection in mice 98 days postinfection (Bang et al., ). Finally, SPI2 has been shown to play a role in spatially distancing the SCV from producers of ROS (NADPH oxidase) and RNS (iNOS) (Vazquez-Torres et al., b, Gallois et al., ; Chakravortty et al., ), although the significance of this particular function of SPI2 during a persistent infection needs more investigation (Aussel et al., ).

Ion transport and metabolism

Nramp1 encodes a phagocyte-specific metal cation transporter (such as Mn2+ and Fe2+) that localizes to the SCV and extrudes essential metal ions away from the bacteria creating a more hostile environment within the SCV (Boyer et al., ; Courville et al., ). Correspondingly, the Salmonella metal ion transporters, MntH, SitABCD, and FeoB were shown to be required for intracellular replication and virulence in Nramp1r mice (Boyer et al., ; Zaharik et al., ). Whether these transporters specifically contribute to persistent infection requires further testing in Salmonella persistence mouse models.

In addition to manganese and iron, magnesium and zinc are critical for intracellular Salmonella growth and the zinc transporter ZnuABC is necessary for virulence in Nramp1r mice (Ammendola et al., ). Salmonella Typhimurium has three transporters for magnesium (MgtA, MgtB, and CorA), all of which are important for virulence in mice (Blanc-Potard & Groisman, ; Boyer et al., ; Papp-Wallace et al., ). mgtC, a gene within the same operon as mgtB, is also required for virulence in Nramp1r mice and was identified in both replicates of the Salmonella persistence screen described above (Blanc-Potard & Groisman, ; Gunzel et al., ; Lawley et al., ). However, the exact biochemical activity of MgtC remains elusive (Gunzel et al., ).

Alternatively, certain metabolic pathways have increased importance during persistent Salmonella infection. Isocitrate lyase (AceA) catalyzes the first step in the glyoxylate shunt pathway, enabling the bacteria to utilize fatty acids as a carbon source. AceA is required specifically during the persistent stage of Salmonella infection. The aceA mutant is attenuated by 12 days postinfection and becomes more attenuated at later time points, demonstrating that Salmonella becomes increasingly dependent on fatty acid metabolism during a persistent infection (Fang et al., ).

The equilibrium of a chronic carrier state resides in the balance between the capacity of the bacteria to overcome the host responses, and the strength of the immune system and its ability to clear the bacteria. Salmonella is able to take advantage of different cell types of the immune system to ensure its intracellular replication and dissemination throughout the body.

Immune response to Salmonella persistence

When a pathogenic microorganism first infects its host, there is usually a dramatic activation of the innate and adaptive immune responses, which can result in disease symptoms. If the pathogen and the host survive this initial interaction, the adaptive immune system usually clears the invading offender. However, some pathogenic bacteria are capable of maintaining infections in mammalian hosts even in the presence of inflammation, specific antimicrobial mechanisms, and a robust adaptive immune response, and can therefore be described as giving rise to persistent infection (Young et al., ; Rhen et al., ; Monack et al., b). The persistence of the bacteria after a normal infection implies that an ‘immune status-quo' has been established after the acute response.

The host as well as the bacteria adopts various mechanisms, which bring about an ‘immune equilibrium'. Pathogens have acquired many mechanisms such as antigenic variation (Saunders, ; Reeves, ) and antigenic imitation, inhibition of synthesis of host proteins, inactivation of humoral immune components, and hiding in sites inaccessible to the immune system to avoid immune recognition and ensure their multiplication, survival, and persistence in the host even in the presence of an active immune response (Young et al., ). These responses can therefore be described as giving rise to persistent infections (Balaram et al., ).

Role of macrophages in S. Typhimurium persistence

For Salmonella, as for many other facultative intracellular pathogens, the key to a successful infection lies in the outcome of their encounter with the host's macrophages. Salmonella can escape killing by these phagocytes and survive and multiply within them, giving rise to chronic infections.

It has been well established that macrophages serve as host cells for Salmonella in vitro or in acute infection models (Fields et al., ; Groisman & Saier, ; Gulig et al., ). Similarly, persistent Salmonella reside within MOMA-2+ macrophages of chronically infected mice (Monack, 2004). The fate of macrophages that are persistently infected with Salmonella is not known, nor is it clear how the bacteria infect new host cells over time. It is possible that bacteria persist within macrophages for the lifetime of the host cell and then infect a new macrophage. However, S. Typhimurium is able to induce host cell death in vivo (Richter-Dahlfors et al., ; Monack et al., ), providing a potential mechanism by which Salmonella can escape from an infected cell to infect neighboring cells.

In a similar model of chronic infection, Salmonella was found within hemophagocytic macrophages, cells that have engulfed white and red blood cells. In vitro assays showed that these cells might represent a survival niche for Salmonella (Nix et al., ). However, it is not clear whether these cells are actively targeted by bacteria or whether the presence of Salmonella is owing to the engulfment of previously infected cells. Interestingly, studies of the intracellular replication of Salmonella revealed that the vast majority of infected cells contain very few intracellular organisms (Sheppard et al., ; Monack et al., ). It appears that active repression of growth rate by a bacteria-directed mechanism may be critical for Salmonella virulence (Tierrez & García-del Portillo, ). More recently, a study suggested that Salmonella enters a dormant-like stage upon entry into macrophages (Helaine et al., ). These studies are in line with the average number of bacterial cells found within macrophages (3–4) in a chronic infection model (Monack et al., ). However, the replicative state of Salmonella in persistently infected cells remains to be elucidated.

Intracellular trafficking of the Salmonella phagosome and the interactions between the bacteria and host factors have been analyzed in epithelial cell models or in tissue culture macrophages from susceptible mice (Haraga et al., ). However, the trafficking and gene expression patterns of persistent intracellular Salmonella have not yet been investigated. About 80% of the bacteria present in persistently infected MLN are found within MOMA-2 expressing macrophages (Monack et al., ). The remaining 20% could be extracellular or in unidentified cells. Other in vivo models using susceptible mouse strains have shown that Salmonella colonizes predominantly macrophages in the liver and CD18+ cells (macrophages and/or DCs) in the blood (Richter-Dahlfors et al., ; Vazquez-Torres & Fang, ). Phagocytes also account for the majority of cells containing intracellular bacteria in the spleen (Salcedo et al., ). Infection of neutrophils and DCs by Salmonella has also been documented in spleen and MLN (Dunlap et al., ; Yrlid et al., ; Cheminay et al., ) in acute models of infection.

Dendritic cells

DCs are involved in the dissemination of the bacteria from the intestine to the systemic organs. Salmonella-infected DCs are able to migrate to lymphoid tissue, as shown in an adoptive transfer mouse model (Zhao et al., ). CD103+ CD11b+ DCs are the main DC population that transports Salmonella from the intestinal tract to the MLN early after infection, and the migration of infected DCs is partially dependent on Toll-like receptor 5 (TLR5) and the chemokine receptor CCR7 (Bogunovic et al., ; Uematsu & Akira, ; Voedisch et al., ). DC migration does not appear to be necessary for Salmonella dissemination from the MLN to systemic organs (Voedisch et al., ). However, these observations have been made only 2 days after infection in a susceptible mouse model, which limits the conclusions one can make regarding DC-dependent Salmonella dissemination during chronic infections, especially because Salmonella is usually not detected in MLN from resistant mice until 4–5 days postinfection (unpublished data from our group). In fact, our group showed that DC migration was regulated by the Salmonella SPI2 effector SseI. SseI plays a role in blocking the migration of host immune cells and consequently attenuates the host's ability to clear systemic bacteria (McLaughlin et al., ).

DCs are also involved in antigen presentation to T cells. Uptake of Salmonella by DC and the activation of DC immune functions, including antigen presentation, are regulated by Salmonella LPS and bacterial effectors from both SPI1 and 2 T3SS (Halici et al., ; Zenk et al., ; Bueno et al., ). Moreover, the reduced intracellular proliferation of Salmonella within DCs limits antigen presentation and thus the development of a rapid T-cell response (Albaghdadi et al., ). The limited intracellular replication and the control of T-cell responses, through antigen presentation, are likely to be important factors for the establishment of a persistent infection, as they allow survival of both the host and the bacteria. Interestingly, a subset of infected DCs can migrate from the lamina propria into the intestinal lumen (Arques et al., ). This finding could be important regarding the chronic infection model, as this migration pattern could be involved in the continuous presence of Salmonella in the gut lumen. This hallmark of the chronic infection model is related to the shedding of bacteria in stools of infected mice, which is thought to be responsible for the transmission of Salmonella to other animals (Lawley et al., ).

Adaptive immune response to Salmonella chronic infection

The initial invasion of gut-associated lymphoid tissues by Salmonella induces a massive inflammatory response, characterized by the recruitment of neutrophils, DCs, inflammatory monocytes, and macrophages (Halle et al., ; Rydstrom & Wick, ). All of these cell types are important for containing the initial bacterial invasion and thus play a role in the establishment of a persistent infection (Griffin & McSorley, ). These early events, as well as the cells and molecules involved have been extensively studied with regard to the acute phase of infection; however, the potential role of these cell types in the immune status of bacterial carriers is scarce (Balaram et al., ).

The magnitude and duration of an immune response are modulated by the production of cytokines. CD4 T-helper (Th) cells play a central role in the production of cytokines during Salmonella infection. The most important signal for T-cell helper differentiation is the cytokine profile present at the time of T-cell stimulation. For example, a Th1 profile is induced by the IL-12 family and interferon, whereas a Th2 profile is induced by IL-10 and IL-4. Th1 is typically seen as pro-inflammatory and as inducing intracellular antimicrobial killing mechanisms, while Th2 is regarded as anti-inflammatory and responsible for the response to helminthes, allergies, and low dose antigens. The potential role of these immune profiles during persistent Salmonella infections has been studied in several mouse models of chronic infection.

The establishment of a persistent infection can be divided into two phases. First, an early resistance phase during which the immune system works to reduce the number of invading bacteria and has been characterized by a high Th1 and a low Th2 responses. Indeed, TNF-α, IL-12, IFN-g, and nitric oxide (NO) derivatives are all required for the control of Salmonella growth by the infected host in the acute stage of the infection (Butler & Girard, ; Trinchieri, ; Mastroeni, ; Mizuno et al., ; Netea et al., ; Stoycheva & Murdjeva, ; Sashinami et al., ). In addition, Blackwell et al. () demonstrate that there is an in vivo bias toward the development of a Th1 response in mice bearing the wild-type Nramp1 allele, while a Th2 response is elicited in Nramp1 mutant mice. Because Nramp1 mutant mice are susceptible to Salmonella infection, this finding underscores the importance of a Th1 response in the resistance to the acute phase of Salmonella infection. This resistance phase is crucial in setting the equilibrium between the immune response and the bacteria that will allow for the establishment of the long-term infection state.

The second step is the maintenance of this equilibrium. Although this stage may require lower Th1 response than in the early phase, IFN-γ still plays an important role in maintaining the equilibrium, as injection of an anti-IFN-γ antibody into mice persistently infected with S. Typhimurium led to the reactivation of an acute infection (Monack, 2004). In addition, persistence of Salmonella Clp-XP-deficient bacteria in Balb/c mice required IFN-γ and TNF-α (Yamamoto et al., ). Furthermore, T-bet-deficient mice that did not generate IFN-γ-producing CD4 T cells had higher bacterial loads in spleens compared to WT mice and eventually succumbed to infection with attenuated Salmonella (Ravindran et al., ). Interestingly, the T-bet-deficient mice survived the first 3 weeks of infection, which correlated with significantly higher bacterial levels compared to WT mice at 3 weeks after infection. This would suggest a role for T-bet-dependent IFN-γ production in the persistence phase of an infection, rather than in the early response to the bacterial invasion.

The balance between Th1 and Th2 levels is likely to be critical for the maintenance of a persistent microbial infection. Various studies have suggested that the pathogenesis and persistence of various chronic bacterial, viral, or parasite infections involve a Th1 to Th2 cytokine switch (Fitzgerald, ; Becker, ). Mice persistently infected with Salmonella have high antibody titers, which is consistent with a Th2-biased immune response (Monack, 2004). In addition, a Th2 response is generally associated with a higher bacterial load. In support of this, mice exposed to lead (Pb), which leads to a Th2-skewed immune response, have significantly higher bacterial burdens compared to control mice (Fernandez-Cabezudo et al., ). In addition, co-infection by Salmonella and Schistosoma, which induces a Th2-skewed host immune response (Sacco et al., ; Wynn et al., ; Schramm & Haas, ), increased Salmonella growth and prolonged the duration of Salmonella infection (Rocha et al., ; Njunda & Oyerinde, ; Bouree et al., ; Muniz-Junqueira et al., ). However, surprisingly, a study of mice deficient for Nramp1 and infected with a sublethal dose of Salmonella showed a reduced bacterial load was associated with a higher Th2 response (Caron, ). Taken together, these findings suggest that the cytokine profiles differ in acute infection compared to persistent Salmonella infections. It appears that a Th1 response is beneficial for the clearance of a high load of bacteria and at the same time conducive for the persistence of low bacterial load. Thus, the Th1/Th2 balance appears to be critical in maintaining an exquisite equilibrium between pathogen survival and host clearance mechanisms.

In addition to Th1 and Th2 T cells, there has been a great deal of interest recently in the role of Th17-producing T cells in the control of Salmonella infections. Th17 cells are defined as a distinct lineage from Th1 and Th2 cells and are characterized by the expression of cytokines such as IL-17A and F, IL-22, and IL-26. IL-17A and IL-22 are induced during some Salmonella infections, and recent work in mice, calves, and the macaque has suggested that these cells are critical for controlling local invasion by Salmonella (Raffatellu et al., ; Godinez et al., ; Liu et al., ). However, a recent study suggested that IL-17 may not be as important for the inflammatory response to S. Typhimurium in the gut (Songhet et al., ). In addition, a recent study indicated that IL-17A-producing cells had a limited role in the maintenance of the protective immunity against Salmonella (Schulz et al., ). Thus, more studies are necessary to fully address the potential role of Th17 T cells during persistent Salmonella infection. Finally, the role of another subset of effector CD4 T cells, regulatory T cells (Tregs), has been studied in a chronic model of Salmonella infection. The Tregs suppressive potency modulates the effectiveness Th1 responses during the course of Salmonella infection and thus influences the progression of persistent infection (Johanns et al., ).

B cells

Resolution of a primary S. Typhimurium infection in mice is a combination of innate and T-cell-mediated effects with a less significant role for B cells and antibodies, because B cell-deficient mice can clear infection. However, the Th1 response these mice develop is transient, and they are not protected from re-infection with a virulent strain (Mastroeni et al., b; Mastroeni, ). In a model of susceptible mice persistently infected with an attenuated Salmonella strain, B cells were shown to contribute to the early phase of T-cell programing via a MyD88-dependent mechanism and are required in a BCR-dependent process for the development of the memory T-cell response (Barr et al., ). Thus, TLR activation of B cells optimizes the generation of the primary Th1 response, a process that does not require Ag presentation, but relies on B cell cytokine secretion. Thus, BCR recognition and B cell Ag presentation are an absolute requirement for the development of Th1 memory cells and hence, protective immunity to Salmonella (Barr et al., ).

What is the role of the indigenous intestinal microbiota during Salmonella persistence? The interactions between the gut microbiota and the immune response are under intense study because of the effects of this relationship with the health of an individual. The microbiota has been shown to play an important role in the host susceptibility to acute Salmonella infection (Bohnhoff et al., ; Collins & Carter, ; Stecher et al., ; Sekirov et al., ). In chronically infected mice, the microbiota is required to lower the bacterial load in the gut and thus lower the fecal shedding of the bacteria. This effect seems to be because of the microbiota itself rather than by a microbiota-induced mucosal response (Endt et al., ).

Human chronic infection

Chronically infected individuals shed bacteria in their stools and urine for periods of time that range from a year to a lifetime, without any apparent signs of disease (Vogelsang & Boe, ). Typhoid carriers are of special concern from a public-health viewpoint as they are the reservoirs for the spread of infection and disease.

An association of chronic typhoid carriage and carcinoma of the gallbladder was first reported by Axelrod et al. (). Welton et al. () observed increased incidence of cancer of the hepatobiliary system in typhoid carriers; this was later confirmed by other studies (Mellemgaard & Gaarslev, ; el-Zayadi et al., ). Metabolites (mutagens and inflammation inducers) and toxins produced by the multiplying bacteria present in the gallbladder have been suggested as the causes of the mutational changes (Nath et al., ).

What is the immune state of human typhoid carriers? Optimal operation of IL-12- and IFNγ-dependent immunity and T-cell function appears to be essential for the eradication of Salmonella from reservoirs within the reticuloendothelial system (Mastroeni et al., ; Dougan et al., ). However, the human carrier state could be associated with an incapacity to develop an effective immune response (Thompson et al., ), although further studies of patients with typhoid are needed to confirm this observation. Chronic carriers likely have genetic or other more transient differences (e.g. disruption of indigenous intestinal microbiota) that predispose them to an immune state that allows for the establishment of this ‘equilibrized' state. This response can be efficient enough to survive the acute peak of infection but too weak to allow the eradication of the bacteria from the organism.


As of today, very few studies have been conducted on chronic Salmonella infections, despite the relevance of this problem for human health. As a consequence, very little is known about the biology of a persistent infection. For example, the mechanisms involved in the transmission of bacteria from a carrier to naïve host remain very poorly understood.

One of the difficulties of such studies stems from the genetic background of the mice used for persistent studies, which carry the Nramp1 wild-type (WT) allele. The most widely used mouse strains (namely C57BL/6 and Balb/c) have a mutated Nramp1 allele. As a consequence, suboptimal models have been developed using Salmonella mutant strains, or sublethal doses of the bacteria. In addition, few genetic studies can be conducted using Nramp1 WT mice, as a limited number of knockout strains are available. This issue might be overcome in the near future with the development of C57/BL6 strains carrying the wild-type version of Nramp1 (Arpaia et al., ). This resistant mouse strain will enable the study of the role of many genes in Salmonella persistence.

The functions of Salmonella effector proteins have primarily been analyzed in acute models of infection. However, the roles of many known effector proteins remain unclear. The study of these proteins in the context of a long-term infection may reveal their functions. The few studies conducted thus far have already challenged the classical views of the roles of SPI1 and SPI2 effectors. Moreover, some effectors may have a specific role during Salmonella chronic infection and thus remain unidentified. It is likely that many unknown virulence factors will be revealed with deeper analyses of Salmonella persistence. In addition, this model will shed new light on the biology of Salmonella–host interactions. The discovery of new virulence factors, or new functions of known factors, will likely reveal new molecular interactions with the host cell. It will be especially interesting to study these interactions in macrophages and DC as Salmonella has been shown to reside within these two cell types during the course of a persistent infection. These studies may reveal how Salmonella manipulates phagocytic cells in order to maintain a balance between replication and activation of danger signals.

The S. Typhimurium persistent mouse model of infection will lead to an increased understanding of the interplay between bacterial pathogens and the immune system. The Th1/Th2 balance is crucial for the maintenance of bacterial persistence and the survival of the host, as a strong Th1 response leads to the clearance of the bacteria, while Th2 skewing leads to susceptibility and death of the host. But how the equilibrium is reached and maintained remains unclear. It will be particularly interesting to understand whether bacterial virulence factors influence or manipulate the Th1/Th2 balance. Likewise, some host factors may play a role in bacterial persistence by favoring the long-term carriage of Salmonella. Identifying bacterial and host factors will likely lead to the identification of new drug targets, which could be relevant for the treatments for other chronic bacterial infections, such as Mycobacterium tuberculosis.


  • Editor: Christoph Dehio


View Abstract