In a mouse infection model, Salmonella enterica serovar Typhimurium can reside and proliferate within phagocytes in deeper tissues (35,47) and it can follow up to induce macrophage death by a possible necrosis (19). This model has been studied as a model of typhoid fever infection in humans caused by S. enterica serovars Typhi and Paratyphi that remains a significant problem in developing countries (46). Three vaccines against typhoid fever are presently in use. Two are injected, and one is orally administered. The two injected vaccines consist of killed whole bacteria and capsular material (Vi antigen) (51). The one live oral vaccine strain, serovar Typhi Ty21a, is an attenuated strain of mutations in galE (a gene in the galactose utilization pathway that affects O-antigen production) (10) and rpoS ( S ) (48, 49). Generally, an oral vaccine is cheaper, easier, and safer to administer than an injected vaccine. In human field trials, the oral vaccination with serovar Typhi Ty21a proved to be safe and effective, but its efficacy varied considerably from trial to trial (9, 51). To develop more promising vaccine strains, researchers have carried out human trials to test various serovar Typhi vaccine candidate strains that harbor precise mutations in specific genes (29 In the mouse model, at least three distinct genetic loci (ity, lps, and xid) affect the ability of the animal to successfully resist systemic infections by serovar Typhimurium. The ity s (designating susceptible response) allele of BALB/c or C57BL/6 mice and the lps d (designating defective response) allele of C3H/HeJ mice are associated with increased susceptibility to infection (30,62) and decreased susceptibility to endotoxin lethality (44), respectively. The ity phenotype is linked to Nramp1 (61), which encodes a macrophage-specific phosphoglycoprotein that is required by the phagosomal membrane during phagocytosis (15,61). The increased susceptibility to infection is associated with a nonconservative amino acid sub-

We found that oral immunization with flagellum-defective mutant strains of Salmonella enterica serovar Typhimurium with the ClpXP-deficient background protected mice against oral challenge with the virulent strain. These data indicate that Salmonella flagellin is not a dominant protective antigen in oral immunization with attenuated live vaccine strains.

We evaluated the efficacy of CS2022 (the Lon protease-deficient mutant strain of Salmonella enterica serovar Typhimurium) as a candidate live oral vaccine strain against subsequent oral challenge with a virulent strain administered to BALB/c and C57BL/6 mice. CS2022 persistently resided in the spleen, mesenteric lymph nodes, Peyer's patches, and cecum of both strains of mice after a single oral inoculation with 1 x 10(8) colony-forming units. Finally, CS2022 almost disappeared from each tissue sample by week 12 in BALB/c mice, whereas CS2022 still resided in each tissue type at week 12 after inoculation of C57BL/6 mice. A significant increase in the serovar Typhimurium lipopolysaccharide-specific secretory immunoglobulin A (s-IgA), as measured for one of the mucosal immune responses, was detected in bile and intestinal samples of both strains of immunized mice at week 4 after immunization. In addition, the expression of gamma interferon mRNA in the spleens of both strains of immunized mice, especially those of C57BL/6 mice, was significantly increased at week 4 after immunization and was boosted during the following 5 days after the challenge was administered to the mice. Furthermore, peritoneal macrophages isolated from immunized mice at week 4 after immunization exhibited an increase in intracellular killing activity against both virulent and avirulent Salmonella. The present results suggested that salmonellae-specific s-IgA on the mucosal surfaces induced by immunization with CS2022 generally prevented mice from succumbing to an oral challenge with a virulent strain. Simultaneously, CS2022 promoted the protective immunity associated with macrophages in both strains of mice.

No abstract