The inability of Salmonella typhi to induce a progressive fatal infection in most laboratory animals, analogous to that seen in man, dictated that most experimental studies of typhoid immunity had to be performed in the mouse typhoid model. ~Drskov and Moltke (1) carried out a classical study into the early distribution of Salmonella in the mouse gut after the ingestion of bread soaked in an S. typhimurium broth culture. However, these workers were unable to precisely delineate the primary site of bacterial invasion in these animals. Until relatively recently, surprisingly little more has been published which contributes to our understanding of the primary site of enteric infection in the mouse. Earlier studies from this laboratory indicated that although the majority of orally introduced S. enteritidis had a rapid transit time through the normal mouse intestine, a small proportion of the inoculum established itself within the walls of the small intestine and in the cecum several days before a systemic infection could be demonstrated (5, 6). However, the primary site of the infection, and particularly, the route taken by the organisms to reach the liver and spleen, was unclear. Studies in opium-treated guinea pigs or in streptomycintreated prestarved mice indicated a heavy early involvement of the lamina propria of the small intestine (7, 8). However, it was possible that the highly artificial experimental conditions used by these investigators may have contributed to this localization (5), so it was decided to follow the early distribution of an intragastric inoculum of virulent S. enteritidis in sequential sections of the normal undisturbed mouse intestinal tract. The fate of the inoculum was followed quantitatively throughout the gastrointestinal tract and the primary site of muscosal penetration was determined. Materials and MethodsAnimals.--6-to 8-wk old specific pathogen-free CD-1 (Charles River Breeding Laboratories, Inc., Wilmington, Mass.) and B6D2 (C57B1/6 X DBA/2 F1) mice of either sex were

CD-1 mice were vaccinated intragastrically or intramuscularly with one or two doses of 200,g of heat-killed Salmonella enteritidis 5694. Control mice were vaccinated with sublethal doses of living S. enteritidis Se795. The mice were challenged intragastrically with approximately 106 S. enteritidis 5694 SMR 7 to 14 days later, and the growth of the challenge population in the liver, spleen, mesenteric lymph nodes, lungs, and intestine was measured quantitatively. Mice receiving two doses of heat-killed vaccine by mouth were able to delay the systemic emergence of a gastrically introduced salmonella infection by 1 to 2 days. The corresponding liver and spleen populations were slightly lower than those seen in the normal controls. On the other hand, mice receiving the living, attenuated vaccine (either intravenously or intragastrically) developed an effective anti-salmonella immunity against subsequent reinfection.both in man and animals after oral vaccination 451

MATERIALS AND METHODS Animals. Sixto eight-week-old male and female random-bred, specific pathogen-free, CD-1 (Charles River, Wilmington, Mass.), inbred C57B1/6 (Jackson Laboratories, Bar Harbor, Me.) and B6D2 (C57Bl/6 x DBA/2 F0) mouse strains were used. The animals were maintained 10 to a cage with free access to food (Charles River Rat/Mouse Formula, Country Foods, Syracuse, N.Y.) and water. Bacteria. The two strains of Y. enterocolitica used have been described in detail elsewhere (4). The WA strain (ATCC no. 27729 and NCTC no. 10938) was recently isolated from the blood of a human patient and found to be highly pathogenic for mice (4). Strain 5819, which is avirulent for mice, was kindly supplied by the New York State Department of Health, Albany.

The in vivo growth of Salmonella paratyphi A, S. paratyphi B, S. paratyphi C, and S. typhi, as well as of an S. typhi-typhimurium hybrid, was studied in three different strains of mice. S. paratyphi A and B and S. typhi demonstrated very little growth potential in any of the intravenously infected mice, even after as many as 20 serial mouse passages. It was noted, however, that small numbers of viable S. paratyphi B and S. typhi persisted in the spleens of infected mice for up to 28 days. Salmonella paratyphi C and the S. typhi-typhimurium hybrid gave rise to progressive systemic infections beginning from very small intravenous inocula. The median lethal doses for the C57B1 strain of mouse were about five organisms. The relevance of these findings with regard to the development of an animal model for studying human typhoid fever vaccines is discussed.

No abstract