Salmonella Species

Classification and Taxonomy

Salmonella is a genus of the family of Enterobacteriaceae. Before 1983 the existence of multiple Salmonella spp. was taxonomically accepted. Currently, as a result of experiments indicating a high degree of DNA similarity, the genus Salmonella is separated into two species: Salmonella enterica , which contains six subspecies (I, II, IIIa, IIIb, IV, and VI), and Salmonella bongori , which was formerly subspecies V. S. enterica subspecies I contains almost all the serotypes pathogenic for humans, except for the uncommon human infections with subspecies IIIa and IIIb, which were formerly designated by the genus Arizonae.

Members of the seven Salmonella spp. can be serotyped into one of more than 2500 serotypes (serovars) according to antigenically diverse surface structures: somatic O antigens (the carbohydrate component of lipopolysaccharide [LPS]) and flagellar (H) antigens ( Table 223.1 ). The name usually refers to the location where the Salmonella serotype was first isolated. According to the current Salmonella nomenclature system in use at Centers for Disease Control and Prevention (CDC) and World Health Organization laboratories, the full taxonomic designation Salmonella enterica subsp. enterica serotype Typhimurium can be shortened to Salmonella serotype Typhimurium or Salmonella Typhimurium. The authors have chosen to use the abbreviated form in this chapter and will omit the “serotype,” for example, designating “Salmonella serotype Typhimurium” as “Salmonella Typhimurium.”

The Genome

The genome sequences of ≈8000 S. enterica strains, including S. Typhi; S. Paratyphi A, B, and C; and numerous nontyphoidal serotypes, are available in GenBank. The salmonellae genomes contain approximately 4.7 to 5.2 million base pairs, with approximately 4500 to 5400 coding sequences. Comparing sequence diversity by multilocus sequence typing suggests S . Typhi emerged from the S. enterica common ancestor around 50,000 years ago. S . Typhi and S . Paratyphi A are closely related to each other but not to other S . enterica serotypes, and their host restriction to humans is related to loss of gene function through pseudogene formation and gene deletion. Next-generation sequencing combined with traditional epidemiologic investigation permits a greater understanding of salmonellae evolution and spread. For example, whole-genome sequencing found that two closely related highly invasive strains of S . Typhimurium have recently emerged (late 20th century) and spread across sub-Saharan Africa temporally and geospatially, associated with the human immunodeficiency virus (HIV) pandemic, likely facilitated by the rapid expansion and mobility of a susceptible host population. Of interest, these strains have undergone some genome reduction, similar to what has been seen in S. Typhi, possibly as a result of greater restriction to human hosts. However, in contrast to the speculation that these strains may have resulted in greater virulence or propensity to bacteremia, if anything these strains appear less fit for resistance to innate immunity, and most usually cause simple gastrointestinal illness.

Microbiology

Salmonellae are gram-negative, non–spore-forming, facultatively anaerobic bacilli that measure 2 to 3 by 0.4 to 0.6 µm in size. Like other Enterobacteriaceae, they produce acid on glucose fermentation, reduce nitrates, and do not produce cytochrome oxidase. All organisms, except S. Gallinarum-Pullorum, are motile as a result of peritrichous flagella, and most do not ferment lactose. However, approximately 1% of organisms can ferment lactose and therefore may not be detected if only MacConkey agar or other semiselective media are used to identify Salmonella based on colorimetric assay for fermentation of lactose. The differential metabolism of sugars can be used to distinguish many Salmonella serotypes; serotype Typhi is the only organism that does not produce gas on sugar fermentation.

Freshly passed stool is preferred for the isolation of Salmonella and should be plated directly onto agar plates. Low-selective media, such as MacConkey agar and deoxycholate agar, and intermediate-selective media, such as Salmonella-Shigella, xylose-lysine-deoxycholate ( Fig. 223.1 ), or Hektoen enteric agar, are widely used to screen for both Salmonella and Shigella spp. Selective chromogenic media, such as CHROMagar Salmonella (DRG International, Springfield, NJ) , are more specific than other selective media, reduce the need for confirmatory testing and time to identification, and increasingly are used for the primary isolation and presumptive identification of Salmonella from clinical stool specimens. Stool specimens can be directly inoculated into selenite enrichment broth before plating on primary media to facilitate the recovery of low numbers of organisms. Highly Salmonella -selective media, such as selenite with brilliant green, should be reserved for use in stool cultures of suspected carriers and under special circumstances, such as outbreaks. Bismuth sulfite agar, which contains an indicator of hydrogen sulfite production and does not contain lactose, is preferred for the isolation of S. Typhi and can be used for the detection of the 1% of Salmonella strains (including most Salmonella serogroup C strains) that ferment lactose. After primary isolation, possible Salmonella isolates can be tested in commercial identification systems or inoculated into screening media, such as triple-sugar–iron and lysine-iron agar.

Direct detection of enteric pathogens from stool specimens by DNA-based syndrome panels is increasingly used by clinical laboratories to allow providers to rapidly identify the cause of gastroenteritis. To ensure that outbreaks of similar organisms are detected and investigated, all specimens that test positive for nontyphoidal Salmonella (NTS) by culture-independent diagnostic testing and for which isolate submission is requested or required under public health reporting rules should be cultured in the clinical laboratory or at a public health laboratory.

Isolates with typical biochemical profiles for Salmonella should be serogrouped with commercially available polyvalent antisera or sent to a reference or public health laboratory for complete serogrouping. Salmonellae are serogrouped according to their polysaccharide O (somatic) antigens, Vi (capsular) antigens, and H (flagellar) antigens according to the Kauffman-White scheme. The Vi antigen is a heat-labile capsular homopolymer of N -acetylgalactosaminouronic acid that is used for the identification of S. Typhi strains and on occasion other Salmonella serotypes by slide agglutination. In S. Typhi and S. Paratyphi C the polysaccharide Vi antigen can inhibit O-antigen agglutination because it is so abundant, and boiling is required to inactivate Vi antigen and to detect O antigen. Most antigenic variability occurs in the O antigen, which is composed of chains of oligosaccharide attached to a core oligosaccharide that is linked covalently to lipid A.

Although serotyping of all surface antigens can be used for formal identification, most laboratories perform a few simple agglutination reactions that differentiate specific O antigens into serogroups, designated as groups A, B, C ₁ , C ₂ , D, and E Salmonella. Strains in these six serogroups cause approximately 99% of Salmonella infections in humans and warm-blooded animals. Although this grouping is useful in epidemiologic studies and can be used to confirm genus identification, it cannot identify whether the organism is likely to cause enteric fever because considerable cross-reactivity occurs among serogroups. For example, S. Enteritidis, and S. Typhi are both group D, and S. Typhimurium and S. Paratyphi B are both group B.

Genotyping methods frequently are used for epidemiologic purposes to differentiate strains of common Salmonella serotypes. These methods include ribotyping, pulsed-field gel electrophoresis, insertion sequences analysis, polymerase chain reaction–based fingerprinting, multilocus sequence typing, and increasingly whole-genome sequencing.

Epidemiology

In many countries the incidence of human Salmonella infections has increased markedly in recent decades, although good population-based surveillance data are mostly lacking, especially from sub-Saharan Africa. In the United States NTS species cause an estimated 1.2 million cases of foodborne illness each year, second only to noroviruses, and are associated with an estimated hospitalization rate of 2.7% and death rate of 0.5%. In the United States the incidence rate of NTS infection has remained relatively stable in the last 20 years and continues to be driven largely by S. Typhimurium and S. Enteritidis ( Fig. 223.2 ). In 2016 the incidence rate of salmonellosis (16.60/100,000 population) was second only to Campylobacter (17.43/100,000 population) among nine potentially foodborne diseases under active surveillance. In comparison, during 2010–14, reported NTS incidence rates ranged between 21.4 to 25.7 per 100,000 population in the European Union. Globally, S. Typhimurium (43.5%) and S. Enteritidis (17.1%) are the most common Salmonella serotypes, with large differences observed in the serotype distribution between regions but lesser differences between countries within the same region. The incidence of NTS infection is highest during the rainy season in tropical climates and during the warmer months in temperate climates, coinciding with the peak in foodborne outbreaks.

Unlike S. Typhi and S. Paratyphi, whose only reservoir is humans, NTS can be acquired from multiple animal reservoirs. Transmission of NTS to humans can occur by many routes, including consumption of food animal products, especially eggs, poultry, undercooked ground meat, dairy products, fresh produce contaminated with animal waste, contact with animals or their environment, and contaminated water. During the 1980s and 1990s, S. Enteritidis associated with shell eggs emerged as the predominant Salmonella serotype and source of foodborne disease in the United States and some other countries. In the United States the rate of reported S. Enteritidis isolates increased from 0.6 per 100,000 population in 1976 to a high of 3.9 per 100,000 in 1994. As a result of intensive surveillance and control effort, including egg farm management practices, such as rodent control and vaccination of young hens, egg quality-assurance programs on farms, egg refrigeration during storage and transport, and consumer education, the incidence of S. Enteritidis infection has declined in the United States and other developed countries (see Fig. 223.2 ). However, outbreaks of S. Enteritidis infection associated with shell eggs continue to occur. In 2010 a national outbreak of S. Enteritidis infection resulted in more than 1900 reported illnesses and the recall of 500 million eggs. Infection localizes to the ovaries and upper oviduct tissue and is transmitted to the forming egg before shell deposition, resulting in contamination of the albumen and yolk. Although cooking eggs until all liquid yolk is solidified kills S. Enteritidis, the use of pasteurized egg products remains the safest alternative for institutions and the general public.

Transmission of S. Enteritidis from farm to farm may be facilitated by contaminated chicken manure, insects, and rodents and by ingestion of feed contaminated with mouse droppings because S. Enteritidis strains cultured from the spleens of mice caught on farms have enhanced ability to contaminate eggs. The loss of cross-immunity resulting from culling chickens infected with S. Gallinarum and S. Pullorum in the United States and United Kingdom also may have contributed to the emergence of S. Enteritidis.

Salmonella live in the intestines of most food animals, and contamination of raw poultry and meat products can occur during slaughter and processing. Retail ground poultry and meat are at high risk of contamination with Salmonella , including with antimicrobial-resistant strains. In 2013 18.0% of ground chicken, 15.0% of ground turkey, and 1.6% of ground beef specimens sampled by the US Department of Agriculture tested positive for Salmonella. Although raw chicken carcasses and other meats are less commonly contaminated with Salmonella than is ground poultry, cross-contamination of food items from handling of raw chicken and inadequate hand hygiene are risks for sporadic salmonellosis in the home. There is considerable mismatch between animal and human Salmonella serotypes, suggesting that the risk of transmission is not equal for all food products and serotypes.

Changes in food consumption and the rapid growth of international trade in agricultural food products and increasing use of manufacturing technologies have facilitated the dissemination of new Salmonella serotypes associated with fresh fruits and vegetables. Human or animal feces may contaminate the surface of fruits and vegetables and may not be removed by washing. Recent multistate foodborne outbreaks of salmonellosis in the United States associated with fresh produce include papayas—multiple serotypes, cantaloupe—multiple serotypes, pistachios— S . Montevideo, cucumbers— S . Poona, alfalfa sprouts—multiple serotypes, bean sprouts— S . Enteritidis, and tomatoes—multiple serotypes. Tomatoes can internalize Salmonella when immersed in water, and contamination on the tomato or melon surface can be transferred to the interior when it is cut. Sprout seeds can become contaminated before sprouting, and soaking seeds with 20,000 parts per million calcium hypochlorite or other disinfectant can reduce but does not eliminate the risk of sprout-associated illness. Recent salmonellae outbreaks have been associated with peanut products, including peanut butter and paste used as a food additive. Salmonellae appear capable of colonizing and adhering to the nut and can be present in raw nuts and, if inadequate processing and or roasting has occurred, in nut-related products.

Manufactured food items pose an enormous potential hazard of foodborne salmonellosis in developed countries because of their centralized production and wide-scale distribution. Both pasteurized and unpasteurized milk and milk products, including ice cream and powdered infant formula, have been recognized as sources of Salmonella infections.

Salmonellosis associated with exotic pets is a resurgent public health problem, especially from exposure to reptiles, including turtles, iguanas, lizards, and snakes, and from amphibians such as aquatic frogs. Of all Salmonella serotypes, 40% have been cultured predominantly from reptiles and are rarely found in other animals or humans. Based on extrapolation from population-based surveillance, 6% of all sporadic Salmonella infections and 11% among persons younger than 21 years are attributable to contact with reptiles or amphibians. The recognition of pet turtle–associated salmonellosis led to the banning of shipment of small pet turtles in the United States in 1975 and in several countries, but small turtles continue to be sold illegally and pose a health risk, especially to children. Exposure to pet birds, live poultry, such as chicks and ducklings, pet rodents and hedgehogs, dogs and cats, and to pet food and pet treats made from animal parts are other reported sources of human salmonellosis, including infection with multidrug-resistant (MDR) strains.

Multidrug resistance among human NTS isolates is increasing in both developing and developed countries. A diversity of transferable resistance plasmids have been identified from MDR NTS strains and contribute to the conjugative transfer of resistance between enteric bacterial species. Of particular concern has been the worldwide emergence in the 1990s of a distinct strain of MDR S. Typhimurium, characterized as definitive phage type 104 (DT104), that is resistant to at least five antimicrobials—ampicillin, chloramphenicol, streptomycin, sulfonamides, and tetracyclines (R-type ACSSuT). All DT104 strains contain a chromosome- and integron-encoded β-lactamase (PSE-1) that appears to have been acquired from plasmids in Pseudomonas spp. The DT104 strain has broad host reservoirs, and its widespread clonal dissemination in domestic livestock, especially among beef and dairy cattle, likely was promoted by use of antimicrobials on farms for therapeutic uses and for growth enhancement. In the United States the proportion of NTS with the ACSSuT phenotype has been decreasing since the early 2000s ( Fig. 223.3 ). In 2014 resistance to at least ACSSuT was reported in 3.1% of NTS, including 14.5% of S. Typhimurium and 9.9% of S. Heidelberg isolates. Acquisition of S. Typhimurium DT104 is associated with exposure to ill farm animals and to a variety of meat products, including raw or undercooked ground beef. Infection with DT104 is associated with increased risk of bloodstream infection and hospitalization compared with infection with susceptible strains, likely reflecting inappropriate empirical antimicrobial therapy.

Outbreaks and sporadic cases of NTS resistant to third-generation cephalosporins have been reported, and international travel and adoption may have contributed to the global spread. Resistance is most commonly mediated by a transferable plasmid containing the ampC ( bla _CMY ) gene, although other extended-spectrum β-lactamases have been described. The bla _CMY gene is probably acquired by horizontal genetic transfer from Escherichia coli strains in food-producing animals and linked to the widespread use of the veterinary cephalosporin ceftiofur. In 2014 2.1% of NTS isolates from humans in the United States were ceftriaxone resistant (minimum inhibitory concentration [MIC] ≥ 4 µg/mL). Ceftriaxone resistance is more common among NTS isolated from blood than stool and is associated with invasive infection and high case-fatality rates among African children, although most cases have concomitant malnutrition, HIV, and malaria.

An MDR strain of S. Newport (MDR-AmpC), with decreased susceptibility to ceftriaxone (MIC > 2 µg/mL) and resistance to eight other human antimicrobials and ceftiofur, has emerged in the United States. In 2014 MDR-AmpC was detected in 1.2% of all US NTS and 3.0% of S. Newport isolates. Risk factors for infection with MDR-AmpC S. Newport include consumption of uncooked ground beef, runny eggs or omelets, and recent exposure to an antimicrobial to which the strain is resistant. More recently, carbapenemase-producing NTS have been reported in Europe, North Africa, and southern Asia.

Over the last decade, strains of NTS with decreased susceptibility to ciprofloxacin (MIC 0.12–0.5 µg/mL) or ciprofloxacin resistance (MIC ≥1 µg/mL) have emerged and have been associated with delayed response and treatment failure. In 2014 4.3% of NTS isolates in the United States had a ciprofloxacin MIC ≥0.12 (see Fig. 223.3 ), and the proportion is higher in Europe (≈6%). These strains have diverse resistance mechanisms, including single and multiple mutations in the DNA gyrase genes gyrA and gyrB and mutations in the chromosomally encoded quinolone resistance-determining region and plasmid-encoded quinolone resistance genes that are not reliably detected by nalidixic acid susceptibility testing or standard ciprofloxacin disk diffusion. Because commercial test systems do not contain ciprofloxacin concentrations sufficiently low to allow use of this breakpoint, laboratories need to determine the ciprofloxacin MIC by Etest or another alternative method.

In Taiwan in 2000 a high-level ciprofloxacin-resistant strain of S. Choleraesuis caused a large outbreak of invasive infections that was linked to the use of enrofloxacin in swine feed. On the basis of increased prevalence of nalidixic acid–resistant Salmonella and fluoroquinolone-resistant Campylobacter spp. in humans, the US Food and Drug Administration withdrew approval of the use of fluoroquinolones in poultry in 2005.

Although health care–associated salmonellosis is infrequent, such infections have been associated with MDR strains, sustained transmission, and substantial morbidity and mortality. The most frequently reported route of transmission of NTS in health care facility–associated outbreaks is foodborne. Although less common, transmission of Salmonella from patients to health care providers has been associated with phlebotomy, handling soiled linen, noncompliance with barrier precautions, and fecally incontinent residents. However, the risk of transmission from health care providers to patients appears to be low if infection control measures, including hand hygiene, correct use of personal protective equipment, and routine disinfection of patient-care equipment, are observed. In contrast, the risk of nosocomial transmission to neonates and infants from acutely or chronically infected family members appears higher. Neonates are at high risk for fecal-oral transmission of Salmonella because of relative gastric achlorhydria and the buffering capacity of ingested breast milk and formula. High-iron infant formula may further increase the risk of infant salmonellosis compared with breastfeeding. Contaminated enteral feeding and crowding also have been associated with nosocomial transmission among pediatric patients. Control of outbreaks in daycare centers may be difficult because of the need for frequent diaper changing and the higher rate and longer duration of convalescent carriage seen in the preschool-age group.

Residents of nursing homes are at increased risk of foodborne salmonellosis and more severe morbidity and mortality because of poor infection control compliance and presence of comorbid illnesses, acid-suppressing medications, and waning immunity.

Pathogenesis

Salmonella infections begin with the ingestion of bacteria in contaminated food or water. Estimates of the infectious dose vary substantially and depend on the method of determination. In studies involving administration of laboratory Salmonella strains to healthy human volunteers, the median dose required to produce disease was approximately 10 ⁶ bacteria. In contrast, investigations of point-source outbreaks suggest that as few as 200 bacteria may produce nontyphoidal gastroenteritis in many of those exposed and that the ingested dose is an important determinant of incubation period and disease severity. Discrepancies in these results may stem from use of strains attenuated by in vitro passage in the challenge experiments and from variation in disease susceptibility in the general population. Gastric acidity represents the initial barrier to Salmonella colonization and conditions or medications, including antacids, H2 blockers, and proton pump inhibitors, that increase gastric pH increase susceptibility to infection. On exposure to acid in vitro, salmonellae display an adaptive acid tolerance response that probably facilitates bacterial survival in the stomach and passage to the small intestine.

Interactions With Intestinal Epithelium and Induction of Enteritis

Salmonellae must evade host antimicrobial factors secreted into the intestinal lumen, including antimicrobial peptides, bile salts, and secretory immunoglobulin A, and traverse a protective mucous barrier before encountering intestinal epithelial cells. Salmonellae express an array of distinct fimbriae that contribute to tight adherence to intestinal epithelial cells in culture. It is necessary to delete multiple fimbriae synthesis genes to prevent infection in animal models, suggesting that functional redundancy exists. Microscopy reveals that salmonellae invade intestinal epithelial cells by a morphologically distinct process termed bacteria-mediated endocytosis ( Fig. 223.4 ). Shortly after bacteria adhere to the apical epithelial surface, profound cytoskeletal rearrangements occur in the host cell, disrupting the normal epithelial brush border and inducing formation of membrane ruffles that reach out and enclose adherent bacteria in large vesicles. This process resembles the membrane ruffling and macropinocytosis induced in many cell types by growth factors and is functionally distinct from receptor-mediated endocytosis, the mechanism by which many other pathogens enter nonphagocytic cells. After the bacteria internalize, a fraction of the Salmonella -containing vesicles transcytose to the basolateral membrane, and the apical epithelial brush border reconstitutes. The epithelial cell type that serves as the principal portal for Salmonella invasion remains uncertain. In the mouse enteric fever model salmonellae preferentially adhere to and enter the specialized microfold cells (M cells) that overlie lymphoid tissue within Peyer patches. In bovine and rabbit models of enteritis, however, salmonellae do not appear to interact preferentially with M cells but, instead, adhere to and invade intestinal enterocytes diffusely. It is possible that M cells are the principal portal of entry in the enteric fever syndrome and that generalized invasion of enterocytes plays a greater role in the enteritis induced by NTS serotypes.

Salmonellae encode a type III secretion system (T3SS) within Salmonella pathogenicity island 1 (the SPI-1 T3SS), which is required for bacteria-mediated endocytosis and intestinal epithelial invasion. T3SSs are complex macromolecular machines that have evolved to subvert host cell function through the translocation of virulence proteins directly from the bacterial cytoplasm into the host cell (see Chapter 1 for an overview). Salmonella mutants lacking a functional SPI-1 T3SS do not invade epithelial cells in tissue culture and are severely attenuated in animal models of infection after oral administration. In the past decade, considerable attention has focused on identifying the virulence proteins translocated into epithelial cells by the SPI-1 T3SS and delineating the host cell processes these proteins target. At least five translocated proteins are essential for efficient invasion of cultured epithelial cells, although invasion in animal tissues may be more complicated and diverse.

Two SPI-1 translocated proteins, SipC and SipA, promote membrane ruffling and Salmonella invasion through direct interactions with the actin cytoskeleton. The SipC protein inserts into the host cell plasma membrane and forms part of a protein complex that allows translocation of additional SPI-1 virulence proteins directly into the host cell cytoplasm. SipC also directly nucleates actin polymerization at the site of Salmonella attachment and stimulates actin filament bundling. The SipA protein further enhances actin polymerization through stabilization of actin filaments and reduction of the critical concentration for polymerization. SipA mutants invade epithelial cells less efficiently than wild-type bacteria and induce disorganized, diffuse ruffling in host cells, in contrast to the localized ruffling induced around wild-type bacteria.

Additional SPI-1 translocated proteins contribute to Salmonella invasion by targeting members of the Rho family of monomeric guanosine triphosphate (GTP)-binding proteins (G proteins). Rho family members, including Cdc42, Rac, and Rho, regulate the structure and dynamics of the actin cytoskeleton and are required for formation of the membrane ruffles that mediate Salmonella internalization. The SPI-1 translocated proteins SopE and SopE2 directly activate Rac1 and Cdc42 in vitro by acting as guanosine diphosphate/GTP exchange factors (GEFs) and induce membrane ruffling and macropinocytosis after microinjection into epithelial cells. SopB is an additional SPI-1 translocated protein that targets inositol phosphate signaling within the host cell by acting as an inositol polyphosphatase. Among other effects, this activity indirectly stimulates Rho GTPases and promotes membrane ruffling. This may be an important pathway for the bacteria to enter human cells, as recent data suggest that a polymorphism in VAC14, a regulator of phosphoinositide can alter entrance of bacteria into host cells and determines host susceptibility to S. Typhi in Viet Nam.

Recent data suggest that only Rac1 and RhoG are essential for the effects of SopE, SopE2, and SopB. Although mutation of sopB, sopE, or sopE2 alone does not impact invasion, combined deletion of these three genes leads to a severe reduction in epithelial cell invasion. Such functional redundancy among translocated proteins is an emerging theme in a variety of T3SSs. Overall, available data indicate that SipA and SipC act in concert with downstream cellular effectors of activated Rho GTPases to initiate and spatially direct the actin rearrangements that lead to Salmonella internalization.

Studies in mice indicate that salmonellae may also cross the intestinal epithelial border by an SPI-1–independent process involving host dendritic cells. These cells express tight junction proteins and can intercalate between intestinal epithelial cells and access the intestinal lumen without disrupting epithelial integrity. In this manner, dendritic cells may internalize bacteria in the intestinal lumen and subsequently carry these bacteria to distant sites as they undergo their physiologic migration to lymphoid tissues. The diversity of mechanisms used by salmonellae to cross the intestinal barrier indicates the importance of this mechanism to its lifestyle within mammals.

In addition to invasion of intestinal epithelial cells, Salmonella serotypes clinically associated with gastroenteritis induce a secretory response in intestinal epithelium and initiate recruitment and transmigration of neutrophils into the intestinal lumen. The SPI-1 T3SS is also required for these responses in tissue culture and animal models of enteritis. Specifically, Salmonella strains unable to deliver any SPI-1 virulence proteins, as a result of mutations in the secretion apparatus, fail to induce fluid secretion or neutrophil accumulation in ligated bovine ileal loops and do not cause gastroenteritis in calves. In tissue culture models of enteritis translocation of SPI-1 proteins into intestinal epithelial cells leads to synthesis and polarized secretion of inflammatory mediators and neutrophil chemokines, including interleukin-8 (IL-8).

Several SPI-1 translocated proteins that contribute to intestinal inflammation and fluid secretion have been identified. Stimulation of Rho GTPase signaling by SopE and SopE2 also leads to activation of microtubule-associated protein kinase pathways and movement of the proinflammatory transcription factor nuclear factor kappa B (NF-κB) to its site of action in the nucleus. In addition to its role in invasion, the inositol polyphosphatase activity of SopB leads to accumulation of d -myoinositol-1,4,5,6-tetrakisphosphate in epithelial cells. The increased concentration of this compound ultimately leads to an increase in cellular basal chloride secretion, with associated fluid flux. The SPI-1 translocated proteins SopA and SopD also contribute to intestinal secretory and inflammatory responses in ligated ileal loops, but the molecular basis of these effects remains unclear. Many other effector proteins that are delivered by the T3SS apparatus may also effect these or similar pathways with different targets. Individual nontyphoidal salmonellae have a diverse complement of effector proteins; for instance, many strains do not have SopE2. The association of specific effector proteins could alter the pathogenicity of specific strains and their emergence in humans from animal reservoirs.

After Salmonella invasion, intestinal inflammation may also result from activation of the innate immune system through stimulation of proinflammatory receptors present on phagocytes and the basolateral surface of intestinal epithelia. This includes activation of Toll-like receptor 4 (TLR4) by LPS and TLR5 by bacterial flagellin. The cytosolic surveillance pathway is also activated by the translocation of flagellin into the cytoplasm by the T3SS and its recognition by the inflammasome through the IL-1β converting enzyme protease-activating factor (IPAF), or NLRC4, pathway. This pathway results in the secretion of IL-1β, an important proinflammatory cytokine. Intestinal inflammation probably contributes to fluid secretion and diarrhea through disruption of the epithelial barrier and increased water flux by an exudative mechanism. In contrast to the neutrophilic inflammation and gastroenteritis induced by NTS strains, S. Typhi induces monocytic inflammation in the human intestine and produces significantly less, if any, diarrhea. The molecular basis of this difference in the host response remains unknown. One possibility is the presence of the Vi polysaccharide capsule in most strains of S. Typhi that can prevent recognition of LPS by TLR4.

Several studies demonstrate that salmonellae also use the SPI-1 T3SS to deliver proteins that downregulate the host inflammatory response associated with Salmonella invasion. The SptP protein inactivates Rho GTPase signaling by acting as a GTPase-activating protein (RhoGAP). This directly opposes the activity of SopE and SopE2 and reduces membrane ruffling and proinflammatory signaling after bacterial invasion. In addition, the SspH1 ubiquitin ligase and AvrA proteins inhibit NF-κB activation and related host cell cytokine synthesis. These SPI-1 translocated proteins may promote bacterial persistence in the host by maintaining host cell integrity and allowing evasion of the host immune response. The presence of SPI-1 translocated proteins with opposing molecular actions (e.g., SopE and SptP) suggests that there may be temporal ordering of protein function, with initial activity of SPI-1 proteins associated with invasion and proinflammatory signaling and subsequent activity of antiinflammatory proteins. This dampening of the inflammatory response attributed to multiple bacterial effector proteins may contribute to the long period of relative asymptomatic colonization of the intestinal tract typical of NTS infection.

After inflammation is generated, an important component of salmonellae survival during gastroenteritis and its continued colonization of the intestinal tract after the resolution of disease involves the organisms’ use of the sulfur-containing compound tetrathionate as an electron acceptor to promote energy metabolism in a microaerobic environment. The intestinal microbiota generate toxic hydrogen sulfide gas through their metabolism, and intestinal epithelia detoxify this gas to thiosulfate. On the induction of inflammation by salmonellae and recruitment of neutrophils, reactive oxygen radicals convert the thiosulfate to tetrathionate, which only salmonellae can use for microaerophilic-based respiration to generate energy. This allows the organism to outcompete with commensals and effectively colonize the intestinal tract. The growth advantage to salmonellae in the host conferred by tetrathionate respiration explains the utility of tetrathionate enrichment broth in the identification of salmonellae. Of interest, this process has been lost in typhoidal salmonellae that are inefficient colonizers of the intestinal tract.