Oral Immunization in Experimental Salmonellosis

Transcription

1 INFECTnON AND IMMUNITY, Aug. 197, p Copyright ( 197 American Society for Microbiology Vol. 2, No. 2 Printed in U.S.A. Oral Immunization in Experimental Salmonellosis II. Characteristics of the Immune Response to Temperature-Sensitive Mutants Given by Oral and Parenteral Routes K. J. FAHEY AND G. N. COOPER Commonwealth Serum Laboratories, Parkville, Victoria 352, Australia, and Department of Medical Microbiology, University of New South Wales, Kensington, New South Wales 233, Australia Received for publication 13 April 197 A temperature-sensitive mutant of Salmonella enteritidis, selected because of its inability to proliferate normally at 37 C, has been used as a living vaccine in mice. When given parenterally or orally, it confers a high degree of resistance against otherwise lethal S. enteritidis infections given intraperitoneally or by mouth. In contrast to most other effective living Salmonella vaccines, the temperature-sensitive mutant survives for only short periods in mouse tissues. Although the vaccine provides protection against S. typhimurium infection, possibly because of antigenic relationships between the immunizing and challenge organisms, it is ineffective against the intracellular infection caused by Listeria monocytogenes. A study of the kinetics of S. enteritidis infection in the liver and spleen of normal and immunized mice has suggested that immunity is dependent upon development of a secondary immunological response which arises approximately 7 days after introduction of the challenge infection. Although humoral antibody production forms part of this secondary response, it is not necessarily responsible for control of the infection. Over recent years, considerable interest has mutant activity, especially those relating to their developed concerning the mechanisms by which behavior in mouse tissues and the influence they immunity to Salmonella infection is achieved in have on the resistance of mice to infections caused animals and man. The conflict between those by other organisms. who hold that specific antibody is the major determining factor (11, 12) and those who relate MATERIALS AND METHODS resistance to a specifically acquired, yet broadly Bacterial strains. The virulent strain of S. enteritidis var. dansyz and the temperature-sensitive mutant effective, hyperactivity of the reticuloendothelial (RE) system (19, 25) is yet to be resolved. Irrespective of the mechanism, it is now generally viously (9). A streptomycin-resistant mutant TSF19 derived from it have been described pre- agreed that best protection in experimental (TSF19 strr) was isolated by plating 19 organisms Salmonella infection is induced when animals are of TSF19 on nutrient agar (NA) containing 1,g of exposed to nonlethal doses of viable streptomycin per ml, incubating at 28 C for 24 organisms hr, and selecting colonies for further characterization. The StrR mutant used in these studies was belonging to the same or related species. We have previously described the development resistant to 2,g of streptomycin, but in all other of temperature-sensitive mutants (TSF) of respects it appeared identical to the strain from which S. enteritidis; these agents do not cause overt it was derived. infection when given intraperitoneally (ip) or S. typhimurium (strain C5) was obtained from by mouth yet protect mice against lethal doses of D. Rowley, University of Adelaide. For the strain of virulent S. enteritidis given by the same routes mice used in these experiments, the LD5o of this organism was 1.8 X 13 organisms by the ip route. (9). This method of assessment is not the only, nor possibly the best, criterion of an effective Listeria monocytogenies (strain NB 64) was obtained from V. Ackerman of the National Biological Standards Laboratory, Canberra; the LD5o of this strain, immunizing agent against Salmonella infection. More importantly, it provides no information when given ip to mice, was 3.16 X 16 organisms. on the mechanisms which may be responsible All organisms were maintained at 4 C on Dorset for protection. In an attempt to seek more definitive answers to these questions, we have con- on a reciprocal shaker in a water bath held at 37 C egg medium. Cultures for inoculation were prepared in nutrient broth (NB) by incubation for 2 to 3 hr sidered a number of other parameters of TSF or, for the TSF mutants, at 28 C. 183

2 184 FAHEY AND COOPER INFEC. IMMUN. Animals. Male mice of an outbred Swiss-albino stock weighing 16 to 2 g were used. These animals were supplied by the Commonwealth Serum Laboratories, Melbourne. Infection techniques. Appropriate dilutions of the cultures were prepared in NB; they were held in an ice bath to ensure that no multiplication occurred when the injections were given. By using the Miles et al. technique (2), viable counts of the suspensions were made before and after the injections to determine the dose of organisms given to each mouse. For ip injection,.5-ml volumes were used whereas.1-ml volumes were used for intra-foot pad injections (ifp) which were given in one hind foot of each animal. The techniques used in preparing inocula for oral administration of the virulent and TSF strains have been described previously (9). Detection of organisms in mouse tissues. Mice were killed by exsanguination from the axillary veins under ether anesthesia. The liver and spleen of each animal were homogenized together by passing through a fine stainless-steel mesh into 2-ml volumes of sterile distilled water. In some experiments, a length of intestine containing a portion of jejunum, ileum, and caecum was passed through the mesh into 2-ml volumes of sterile water. For quantitative counts, the Miles et al. (2) technique was used, dilutions being prepared in sterile water and plated on Mac- Conkey agar or MacConkey agar containing 1,g of streptomycin per ml. Depending on the experiment, these plates were incubated for 24 hr at 37 or 28 C before colonies were counted. When only qualitative estimates were required, the tissue homogenates were transferred to 1-ml volumes of Rappaport's medium (23) incubated at 28 or 37 C for 48 hr, and subcultured on MacConkey agar for isolation of colonies at 28 or 37 C. Organisms were confirmed as S. enteritidis by slide agglutination of selected colonies. Serun antibody estimations. Blood was collected from the retro-orbital venous plexus of mice;.2-ml volumes were transferred to capillary tubes which were centrifuged at 8 X g for 1 min. The serum (ca..1 ml) was collected from these tubes, diluted 1:2 in a 1% solution of fresh heated normal rat serum in.85% saline, and serially diluted in twofold steps in this medium by using the Takatsy Microtitrator kit (Labor, Budapest). Equal volumes of a 2% rat red blood cell (RBC) suspension, previously sensitized by incubation for 1 hr at 37 C in a 5,ug per ml solution of S. enteritidis "haemosensitin" prepared by sodium hydroxide extraction (G. N. Cooper, Ph.D. Thesis, Univ. of Melbourne, 1958), were then added to the wells of the perspex trays. The trays were incubated at 37 C in a moist atmosphere for 1 hr. Titers were recorded as the reciprocal of the highest serum dilution which caused complete agglutination of the sensitized rat RBC. Appropriate positive and negative controls were always included. RESULTS When given by the ip route, TSF mutants have been found to induce solid resistance to lethal challenge infections of virulent S. enteritidis given by the same route; the same holds true for oral immunization and challenge (9). It was of interest to determine whether immunization by one route conferred a general immunity which could be expressed by increased survival after challenge infections given by other routes. The results of a series of experiments designed to test this are summarized in Table 1. In all instances, mice were challenged 4 weeks after receiving an immunizing dose of the mutant TSF19. Parenteral immunization with the viable TSF mutant clearly provided excellent protection against either ip or oral challenge infection. In addition, a large oral dose of the TSF mutant seemed to protect mice, not only against oral challenge but also to some extent against an otherwise lethal ip infection of S. enteritidis. Persistence of TSF mutants in mouse tissues. In vitro studies on the growth of the TSF mutants have indicated that although their growth is arrested during incubation at 37 C, this temperature is not immediately lethal. Furthermore, we have shown that "temperature-tolerant" variants occur with a frequency of 1:14 within the TSF mutant populations; these variants are capable of multiplying slowly and will produce macroscopic colonies on solid media after 24 to 48 hr of incubation at 37 C (9). It seemed important, therefore, to establish whether the TSF mutants persisted in mouse tissues for any length of time after oral or ip immunization; if this was the case, could it be attributed to selection of "temperature-tolerant" variants which possessed some capacity for multiplication at body temperature? Mice were injected ip with 5 X 16 organisms of a log phase culture of TSF19. At various intervals thereafter, groups of animals were killed by exsanguination; homogenates of spleen, liver, and intestine were incubated in Rappaport's medium at 28 C, and subcultures were subsequently made on MacConkey agar, incubating at 28 C for 24 hr. Any lactose-negative colonies were tested for biochemical activity and capacity to agglutinate with S. enteritidis antiserum; they were also plated on MacConkey agar and incubated at 37 and 42 C for 24 hr to confirm that they were temperature sensitive. As shown in Table 2, the TSF mutant persisted in the majority of animals for a period of 2 weeks; at later times it was isolated from only an occasional animal. In general, the organism was found more regularly in the liver than the spleen; on no occasion was it isolated from the intestine. Because the organisms isolated from the tissues did not grow at 37 C, it seemed that the preponderance of organisms were temperature-

3 VOL. 2, 197 TSF MUTANTS OF S. ENTERITIDIS 185 TABLE 1. Resistance to oral or ip infection with Salmonella enteritidis in mice immunized 4 weeks earlier by oral, ip, or ifp injection of TSFJ9 No. of micementm Route of challengeinfection ;No. of organisms Route of No. of TSF19 in surviving/no. Per cent t in Meantme challengea immunizationb immunizing dose of mice survival to death infectedcdas Intraperitoneal 3 X 14 ip 16 16/16 1 (2) ifp 16 19/19 1 Oral 2.3 X 1' 8/ Nil Nil /2 8.5 Oral 8.4 X 18 ip 16 2/2 1 (5) ifp 16 2/2 1 Oral 1.6 X 11 15/ Nil Nil 3/ a Figures in parentheses indicate approximate numbers of LD5 in challenge dose. b Abbreviations are: ip, intraperitoneal; ifp, intra-foot pad injections. c Survival for 28 days after challenge. sensitive mutants and not "tolerant variants" of it. Further confirmation of this was obtained from animals killed 1 or 21 days after injection. Tissues from these animals were homogenized, placed in 1-ml volumes of NB, and incubated for 18 hr at 28 C. These cultures were then diluted serially in 1-fold steps, and samples of each dilution were plated on NA plates which were incubated at 28 or 37 C to assess the frequency of "temperature-tolerant" variants in the population. In all instances, this was found to be of the order of 1:14, i.e., identical to that of the parent TSF mutant. In a second experiment, mice were given an oral dose of approximately 2 X 11 organisms of the mutant TSF19 suspended in sterile milk. Groups of animals were killed at various times, and cultures of their spleens, livers, and intestines were prepared as before. The TSF mutants persisted only for a short period in the intestinal tissues (Table 3); on no occasion was the organism isolated from the livers or spleens. As before, no preferential selection of "temperaturetolerant" variants in the intestine could be demonstrated. Course of infection in immune and nonimmune mice. Collins et al. (8) have previously stressed that quantitative estimation of viable organisms in the liver and spleen over a period of some days provides sounder evidence of resistance to Salmonella infection than do figures based on progressive mortality and mean time to death after lethal challenge. Therefore, in the following experiment, the number of organisms present in the livers and spleens of immunized and normal mice after challenge infection with the virulent S. enteritidis was assessed. The experiment also included a further attempt to demonstrate the TABLE 2. Persistence of TSFJ9 in mouse tissues after intraperitoneal injection of S X 16 organisms Time after injection days No. of mice yielding positive cultures/ no. of mice examined 3/4 4/4 4/4 3/4 3/4 1/4 1/4 /4 1/6 No. of mice with positive cultures in Spleen Liver Intestine TABLE 3. Persistence of a TSF mutant in mouse tissues after oral administration of 2 X 11 organisms No. of mice No. of mice with positive Time after yielding posi- cultures in oral tive cultures/ infection no. of mice examined Spleen Liver Intestine days 1 6/ / /6 1 7 /6 1 /6 14 /6 presence of the TSF mutant in these organs at the time of challenge. Mice were immunized with 8 x 16f organisms of a log phase culture of the TSF19 StrR mutant

4 186 FAHEY AND COOPER INFEC. IMMUN. by the ip route. Four weeks later, these mice and a group of normal animals of the same age were injected ip with 5 X 14 organisms of a log phase culture of S. enteritidis var. danysz. This strain was sensitive to streptomycin. At various times after challenge, randomly selected groups of four mice were used for assessing the number of organisms present in the livers and spleens. By using the Miles et al. technique (2), samples of dilutions of the liver-spleen homogenates were plated on MacConkey agar and MacConkey Agar containing 1 Ag of streptomycin per ml, the latter being incubated at 28 C to allow development of any colonies of the TSF19 StrR mutant. In addition, samples of the tissue homogenates were transferred to Rappaport's medium and incubated at 28 C for 24 hr; subcultures were then made on MacConkey agar (incubated at 37 C) and MacConkey-streptomycin agar (incubated at 28 and 37 C) to detect the presence of organisms of the challenge and immunizing strains which were in too few numbers to be measured quantitatively. The results of this experiment are summarized in Table 4 and Fig. 1. The TSF mutant was not regularly isolated from the liver and spleens of the immunized mice; it was detected in only 5 of the 28 animals examined. In no instance was the number of the mutant cells sufficiently large (<2 per liver and spleen) to be counted by the quantitative technique. When the results for individual mice were assessed, no correlation could be found between the number of viable organisms of the challenge strain present in the liver and spleen and the presence (or absence) of the TSF strain in these tissues. Again, there was no evidence to suggest that selection of "temperature-tolerant" variants had occurred. Although it is known that after ip injection of virulent salmonellae considerable multiplication of the organism occurs in the peritoneal cavity, a generalized septicemia also occurs. As a result, the liver and spleen serve as major target organs (22). Our results amply confirm this, for, over a period of 7 days, a 1,-fold increase in the number of virulent organisms was observed in these organs of normal mice (Fig. 1); the number present about the time of death of the animals was similar to those recorded by others for infections given by the ip, intravenous (iv), or other routes (3, 26). The course of infection in these organs of mice immunized with the TSF mutant was different in several respects. First, it seemed that 1 day after infection the number of organisms was at least 1-fold lower than in control mice; although the subsequent rate of multiplication was similar in normal and im- TABLE 4. Detection of immunizing and challenge strains of Salmonella enteritidis in liver and spleens of mice Time after challenge infectiona No. of mice yielding positive cultures from liver and spleen/no. of mice examined McA(37 C)b McA-S(37 C) McA-S(28 C) days 1 4/4 /4 1/4 3 4/4 /4 3/4 5 4/4 /4 1/4 6 4/4 /4 /4 8 4/4 /4 /4 1 4/4 /4 /4 13 4/4 /4 /4 a Mice immunized ip with 8 X 16 organisms of TSF19 StrR 4 weeks earlier. b Column headings refer to media and temperature of incubation of subcultures made from Rappaport's medium. McA, MacConkey Agar; McA-S, MacConkey Agar containing 1,g of streptomycin per ml. Organisms isolated on McA at 37 C are S. enteritidis var. danysz. Organisms isolated on McA-S at 37 C are "temperature-tolerant" variants of TSF19 StrR. Organisms isolated on McA-S at 28 C are TSF19 StrR. munized mice over a period of 4 to 5 days, it ceased abruptly in the latter animals at about the 6th or 7th day of infection. Over the succeeding 7 days, the numbers declined by a factor of almost 1,-fold. Protective effects of TSF immunization against other infections. The studies of Mackaness and his colleagues (2, 8) have indicated that immunity induced by avirulent strains of Salmonella is not specific but may also be effective against intracellular infections caused by unrelated organisms. In the following experiments, groups of mice were immunized with 5 X 16 organisms of a log phase culture of TSF19 by the ip route. These animals were challenged 4 weeks later by ip injection of log phase cultures of S. typhimurium or L. monocytogenes. The results indicate that the TSF mutant is capable of protecting mice against an otherwise lethal infection caused by the antigenically related S. typhimurium but has no observable effect against Listeria infection (Table 5). Serum antibody responses in mice. Ten mice were injected with 5 X 16 organisms of TSF19 by the ip route; 4 weeks later all animals were given a similar second dose of the mutant. Using the hemagglutination technique, titrations for S. enteritidis antibody were made on serum samples obtained from each animal after the first and second injections. As shown in Fig. 2a,

5 VOL. 2, 197 TSF MUTANTS OF S. ENTERITIDIS 187 the antibody response after the first injection was slow to develop and of a very low order. Only 6 of the 1 mice had detectable levels of antibody at any time during the 28-day period. Nonetheless, the first injection obviously induced immunological memory, for all animals responded to the second injection and relatively high levels of antibody were found; the peak of the secondary response occurred at day 7. Attempts to demonstrate an antibody response after oral administration of 3 x 19 or 1' organisms were unsuccessful. In a final experiment, 1 mice were immunized ip with 3.5 X 16 organisms of the TSF19 mutant. Four weeks later, these animals were chalz..> 8-7 ' o z 3 2 Jo I.. 3 s DAYS AFTER CHALLENGE FIG. 1. Growth of S. enteritidis var. dansyz in livers and spleens of normal mice and mice immunized 4 weeks earlier with 8 X 16 organisms of the mutant TSF 19, StrR by the ip route. I I lenged ip with 15 organisms (5 LD5) of a log phase culture of S. enteritidis var. danysz. A group of normal mice were injected with the same dose of the virulent strain. The serum antibody titers were measured after the first and second injections. In the case of mice given the virulent strain alone, no antibody response was detected over the first 7 days; thereafter, the animals became so severely ill that blood samples could not be collected. All of these animals died between the 1th and 13th days. One of the mice injected with the TSF mutant died seven days after challenge; for this reason, it was not included in the series. As shown in Fig. 2b, the antibody response after injection of the TSF mutant was minimal in the remaining nine animals. However, after challenge with the virulent strain, all animals developed significant serum antibody titers which reached their peak approximately 7 days later. DISCUSSION It is generally conceded that the present forms of killed bacterial vaccines used for prophylaxis against enteric infections such as typhoid, cholera, and bacillary dysentery do not guarantee complete or lasting protection in a substantial proportion of individuals. One of the major obstacles to developing more effective vaccines is the lack of suitable experimental models which reproduce the major features of these diseases and thus might be used as a basis for assessing protective capacity of immunizing agents. In the case of typhoid fever, the emphasis has shifted to natural Salmonella infection in rodents as the most suitable alternative for judging the potential value of various forms of prophylactic agents and, more importantly, defining the important mechanisms of immunity which operate in the! enteric fevers. How valid the extrapolation to human typhoid might be has never clearly been stated. In recent years, as a result of studies on TABLE 5. Protective effects of TSFJ9 against intraperitoneal (ip) infections of Salmonella typhimurium and Listeria monocytogenes given 4 weeks after ip immunization No. of mice Mean time to Challenge infection No. in of challengea organisms Mice immunized with surviving/no. of Per mice. survival cent death (days) infected (as Salmonella typhimurium C5 4.5 X 14 TSFl9 24/ (25) Controls /2 1.5 Listeria monocytogenes NB X 17 TSFl9 2/ (3) Controls /2 3.5 a Figures in parentheses indicate approximate LD5 in challenge infection. b Survival 28 days after challenge.

6 1XX FAHEY AND COOPER INFEC. IMMUN. i 5 x lo orgs TSF 19 IP. z 4 X F 123 >' 32 C z I' S 16 orgs TSF 19 I.P. 2j 1 1x orgs SAL.ENTERITIDIS'DANSYZ IP DAYS AFTER INJECTION FIG. 2. (A) Serum antibody titers in mice after ip injection of the TSFI9 mutant. (B) Serum antibody titers in mice given one ip injection of TSFJ9 followed 4 weeks later by ip injection of S. enteritidis var. dansyz. rodents, some attention has been given to the use of streptomycin-dependent and other mutants (1, 15, 27) which may be given in viable form and have been found to protect animals from lethal Salmonella infections. The temperature sensitive mutants of S. enteritidis which we have described (9) offer an alternative approach to achieving excellent anti-salmonella immunity. There is now general agreement that control of Salmonella infection in the rodent is ultimately dependent upon destruction of the organisms by phagocytic cells of the RE system. Unfortunately, the more important issue concerning the mediating mechanisms, humoral antibody (12), cytophilic antibody (16, 24), or less specific cell hyperfunction with associated development of delayed hypersensitivity (7, 17), remains a matter of contention. Undoubtedly, as Blanden et al. (2) have pointed out, the significance of any particular mechanism may be dependent upon the route used for infecting the animals. Thus, although the ip route may highlight the role of specific antibody (1, 26), the iv route emphasizes the role of cellular mechanisms (2). Irrespective of the views one may have on the importance of either mechanism, there seems no good reason at this moment to exclude the possibility that it may be necessary for both mechanisms to act in a complementary fashion if fully effective protection is to be achieved. The TSF mutant used here possesses most of the characteristics of the virulent strain from which it was derived; the only significant phenotypic differences were its inability to synthesize flagella protein and to form cross septa (and thus multiply normally) at the body temperature of the host. Though some increase in bacterial cell material might occur over a period of time, the temperature restriction probably places a severe limitation on the capacity of the mutant to disseminate widely through the tissues of the host and to remain viable for long periods in these tissues. In these respects it differs from (and may have some advantages over) other mutants, avirulent strains, or nonpathogenic species of salmonellae which have been used in similar studies. Several general conclusions can be made concerning the nature of the immunity conferred upon mice exposed to relatively small numbers of the TSF mutants. As pointed out earlier, the immunity of S. enteritidis infection develops within a period of 1 to 2 weeks and is long-lived. It is of a generalized nature; ip or ifp injection of the mutant provides excellent protection against otherwise lethal ip or oral infection. It is significant that oral administration of the mutant is not only effective against oral infection but also, to a lesser extent, against the stringent test of an ip challenge infection. The fact that 4% of orally immunized animals survived this challenge indicates that limited invasion of the intestinal tissues by the mutant had probably occurred and was sufficient to induce a generalized immune response.

7 VOL. 2, 197 TSF MUTANTS OF S. ENTERITIDIS 189 When avirulent or mutant strains of salmonellae are used as immunizing agents, they disseminate throughout the RE tissues of mice and exist there in detectable (often countable) numbers for long periods (2, 3, 8). Mitsuhashi et al. (21), by using a diffusion chamber technique, demonstrated that though humoral antibody might appear, the development of immunity to challenge infection depended upon prolonged interaction of the living vaccine with host tissues; Collins et al. (3, 8) have further claimed that there is a direct correlation between the capacity of such agents to confer protection against S. enteritidis and their ability to establish a permanent infection in the liver and spleen of the host. The results obtained with the TSF mutant indicate that this may not be essential. For obvious reasons, it is very much easier to demonstrate the presence of an organism in mouse tissues than it is to prove conclusively its absence. Rappaport's medium, found by others to allow isolation of from one to five viable salmonellae from fecal material (23), was used in an attempt to enrich selectively the growth of any TSF mutants present in the mouse liver and spleen. With this technique, the mutant was found in the tissues of all mice examined as long as 2 weeks after immunization; thereafter, it was found in only a small proportion of animals, even though immunity was maintained for at least 16 weeks (9). The TSF mutant cultures contain organisms described as "temperature-tolerant" variants which occur with a frequency of ca. 1:14. Although capable of slow multiplication at 37 C, there was no evidence to suggest that selective proliferation of these variants had occurred in vivo nor that they survived indefinitely in the mouse tissues. Collins (3) has pointed out that, irrespective of the route of inoculation, death is finally determined by the multiplication of organisms in the liver and spleen and occurs when "toxic" levels are reached. It was of interest to follow quantitatively the course of infection in normal and TSF-immunized mice after ip injection of virulent organisms. By using a streptomycinresistant TSF mutant, it was also possible to determine whether the immunizing strain was present in the tissues; as mentioned earlier, it was found in less than 2% of the animals and then in numbers less than 2 per liver and spleen. Although the method used for homogenizing livers and spleens of mice may be open to criticism on the grounds of incomplete disruption of tissue cells, the fact that in normal mice the rate of multiplication of virulent organisms and the numbers obtained before death were similar to that found by others who have used more sophisticated techniques (5) supports the belief that our results adequately reflect the course of S. enteritidis infection in these tissues. Furthermore, in unpublished experiments, one of us (G.N.C.) has found that when homogenates prepared by the present method were subjected to further disruption on a high speed blendor, the viable counts were never increased by more than a twofold factor and in some instances were slightly lower than on the original homogenate. Statistically, there were no significant differences in counts obtained after the first and second homogenizing procedures. In immunized mice 1 day after injection, the numbers were 1-fold less than in the controls, suggesting that enhanced destruction of the challenge organism had occurred in the peritoneal cavity, liver, spleen, or in all three sites. In spite of this, the organism multiplied in the liver and spleen at the same rate as in normal animals for a period of 6 to 7 days; an abrupt change in the course of infection then occurred, resulting in a rapid reduction of viable organisms. This contrasts with the results observed when S. gallinarum was used as a living vaccine; whereas the immunizing strain remained in detectable numbers in the liver and spleen, no multiplication of S. enteritidis (given ip or iv) could be detected in these tissues (6, 8). It more closely resembles the situation found by Colfins (4) in animals which had been immunized with sublethal doses of S. enteritidis or S. typhimurium up to 15 days before challenge. The contrast in these results implies that there may be significant differences in the means by which immunity is achieved in animals, depending on whether the immunizing organism is capable of surviving in significant numbers in the tissues for long periods. We cannot deduce from the present experiments the relative contributions that humoral or cellular mechanisms made to the high level of resistance found in the TSF-immunized mice. It is possible that residual humoral antibody may have been responsible for the initial reduction in numbers presumably in the peritoneal cavity (1) as well as the liver and spleen (2, 8). However, enhanced destruction of the organisms in these initial stages did not lead to immediate control of the infection, for multiplication occurred over the succeeding 6 days. The infection then aborted abruptly at day 7, and this can only be attributed to a secondary immunological response initiated by the challenge infection. Although serum antibodies developed during the infection, there is no reason to suppose that the control was entirely due to the increased supply of opsonins; the result can also be interpreted

8 19 FAHEY AND COOPER INFEC. IMMUN. in terms of a specific recall of cellular immunity (4, 5, 17) leading to rapid intracellular destruction of the organisms. Blanden et al. (2) have found that mice immunized with sublethal doses of S. typhimurium are resistant to challenge with doses of L. monocytogenes equivalent to 2 LD5o; the immunity was related to the presence of viable organisms of the immunizing strain in the animal tissues. The present results support this; mice which were injected with the TSF mutant of S. enteritidis were susceptible to a 3 LD5 challenge dose of a less virulent strain of L. monocytogenes given at a time when few if any TSF organisms remained viable in the liver and spleen. In contrast, the TSF mutant protected animals against S. typhimurium infection, an organism which possesses at least two somatic antigens in common with it. This result does not negate the role of cellular immunity, for if protection in this instance is dependent upon a secondary or recall response, it might easily be induced by an organism which is antigenically related to but not necessarily identical with the immunizing strain. Jenkin et al. (12, 13) have claimed that humoral antibodies specific for antigen factor 5 may be important in determining resistance to S. typhimurium infection given by the ip route. As the TSF mutant does not possess this antigen, the challenge infection could not have induced a secondary response to it. It is evident that the results presented here do not decisively delineate the roles played by humoral and cellular immune mechanisms to Salmonella infection; in this respect, they are characteristic of most publications on this obtuse subject. However, they demonstrate that there are a number of features of the immunity induced by the TSF mutants which are quite different from those until now regarded as essential for resistance to Salmonella infection. Just as it is stated that the use of particular routes of challenge may overly stress the importance of one immune mechanism, it seems possible that the use of a particular type of immunizing agent may overemphasize certain prerequisites of an effective immunizing agent. In a sense, the TSF mutants might be considered to provide a desirable alternative to both killed and living vaccines; on one hand, they are demonstrably more effective than killed preparations but, on the other, they do not have the disadvantage of establishing a persisting infection of the RE system which, for many reasons, might be ultimately injurious to the host. ACKNOWLEDGMENTS This investigation was supported by grants from the National Health and Medical Research Council of Australia and the Division of Communicable Diseases, World Health Organization. The assistance of the Commonwealth Serum Laboratories, Melbourne, is also gratefully acknowledged. LITERATURE CITED 1. Akiyama, T., K. Maeda, and D. Ushiba Studies on immunity of experimental typhoid. Challenge of mice passively immunized with antiserum through various routes. Jap. J. Bacteriol. 17: Blanden, R. V., G. B. Mackaness, and F. M. Collins Mechanisms of acquired resistance in mouse typhoid. J. Exp. Med. 124: Collins, F. M Cross-protection against Salmonella enteritidis infection in mice. J. Bacteriol. 95: Collins, F. M Recall of immunity in mice vaccinated with Salmonella enteritidis or Salmonella typhimurium. J. Bacteriol. 95: Collins, F. M Effect of specific immune mouse serum on the growth of Salmonella enteritidis in nonvaccinated mice challenged by various routes. J. Bacteriol. 97: Collins, F. M Effect of a specific immune mouse serum on the growth of Salmonella enteritidis in mice preimmunized with living or ethyl alcohol-killed vaccines. J. Bacteriol. 97: Collins, F. M., and G. B. Mackaness Delayed hypersensitivity and Arthus reactivity in relation to host resistance in Salmonella-infected mice. J. Immunol. 11: Collins, F. M., G. B. Mackaness, and R. V. Blanden Infection-immunity in experimental salmonellosis. J. Exp. Med. 124: Fahey, K. J., and G. N. Cooper Oral immunization against experimental salmonellosis. I. Development of temperature-sensitive mutant vaccines. Infec. Immun. 1: Howard, J. G Resistance to infection with Salmonella paratyphi C in mice parasitized with a relatively avirulent strain of Salmonella typhimurium. Nature (London) 191: Jenkin, C. R The effect of opsonins on the intracellular survival of bacteria. Brit. J. Exp. Pathol. 44: Jenkin, C. R., and D. Rowley Basis for immunity to typhoid in mice and the question of "cellular immunity." Bacteriol. Rev. 27: Jenkin, C. R., and D. Rowley Partial purification of the "protective" antigen of Salmonella typhimurium and its distribution amongst various strains of bacteria. Aust. J. Exp. Biol. Med. Sci. 43: Jenkin, C. R., M. L. Karnovsky, and D. Rowley Preparation of an artificial antigen and immunity to mouse typhoid. Immunology 13: Kishimoto, Y Live-vaccine immunization in experimental typhoid with a streptomycin-dependent strain of Salmonella enteritidis. Jap. J. Bacteriol., 2: Kurashige, S., N. Osawa, M. Kawakami, and S. Mitsuhashi, Experimental salmonellosis. X. Cellular immunity and its antibody in mouse mononuclear phagocytes. J. Bacteriol. 94: Mackaness, G. B The immunological basis of acquired cellular resistance. J. Exp. Med. 12: Mackaness, G. B The relationship of delayed hypersensitivity to acquired cellular resistance. Brt. Med. Bull. 23: Mackaness, G. B., R. V. Blanden, and F. M. Collins, Host-parasite relations in mouse typhoid. J. Exp. Med. 124: Miles, A. A., S. S. Misra, and J.. Irwin The estimation of the bactericidal power of the blood. J. Hyg. 38: Mitsuhashi, S., N. Osawa, K. Saito, and S. Kurashige Cellular immunity with special reference to monocytes of mice immunized with live vaccine of Salmonella enteritidis. Tohoku J. Exp. Med. 89: rskov, J., K. A. Jensen, and K. Kobayshi Studien i.ber Breslauinfecktion der Mause speziel mit Rucksicht

9 VOL. 2, 197 TSF MUTANTS OF S. ENTERITIDIS 191 auf die Bedeutung des Retikuloendothelial gewebes. Z. Immunitaetsforsch. Exp. Ther. 55: Rappaport, F., N. Konforti, and B. Navon, A new enrichment medium for certain salmonellae. J. Clin. Pathol. 9: Rowley, D., K. J. Turner, and C. R. Jenkin The basis of immunity to mouse typhoid. 3. Cell-bound antibody. Aust. J. Exp. Biol. Med. Sci. 42: Saito, K., T. Akiyama, M. Nakano, and D. Ushiba The interaction between Salmonella enteritidis and tissue cultured macrophages derived from immunized animals. Jap. J. Microbiol. 4: Ushiba, D., K. Saito, T. Akiyama, M. Nakano, T. Sugiyama, and S. Shirono Studies on experimental typhoid. Bacterial multiplication and host cell response after infection with Salmonella enteritidis in mice immunized with live and killed vaccines. Jap. J. Microbiol. 3: Williams Smith, H The immunization of mice, calves and pigs against Salmonella dublin and Salmonella choleraesuis infections. J. Hyg. 63: