Induced by Salmonella typhimurium in - Infection and Immunity

INFECTION AND IMMUNITY, Jan. 1976, p. 172-179 Copyright © 1976 American Society for Microbiology

Vol. 13, No. 1

Printed in U.S.A.

Indigenous Microorganisms Prevent Reduction in Cecal Size Induced by Salmonella typhimurium in Vaccinated Gnotobiotic Mice GERALD W. TANNOCK' AND DWAYNE C. SAVAGE* Department of Microbiology and School of Basic Medical Sciences, University of Illinois, Urbana, Illinois 61801

Received for publication 30 June 1975

Germfree CD-1 mice challenged by the oral route with Salmonella typhimurium had ceca that were abnormal in appearance and reduced in size compared to those of germfree controls. Similarly, germfree mice injected with heat-killed S. typhimurium or gnotobiotes associated with three indigenous microbes (Lactobacillus, Bacteroides, Clostridium), and subsequently challenged with S. typhimurium also had small ceca. By contrast, gnotobiotic mice that had been both injected with the heat-killed S. typhimurium and associated with the three indigenous microbes before challenge with S. typhimurium had ceca similar in size and appearance to germfree mice. Thus, indigenous microorganisms could interfere with the mechanism by which the pathogen induced the decrease in cecal size, but could do so only in mice injected with heat-killed bacteria. This phenomenon suggests synergism between the interference effected by the indigenous bacteria and the resistance mechanisms of the animal.

Microbial interference refers to interactions by which allochthonous microorganisms are prevented by autochthonous microbes from establishing in habitats in an ecosystem. Examples of interference in which the indigenous microbiota ofthe gastrointestinal tract prevents the establishment of intestinal pathogens have been provided in experimental infections of mice with Shigella (5), Vibrio cholerae (14), and Salmonella (1, 6, 7). In Salmonella infections, little is known of the microbial types and mechanisms involved in interference. The indigenous microbes may interfere directly by competing with the pathogen for an essential resource in the ecosystem (e.g., nutrients) or by producing toxic metabolic products (10). The indigenous microbiota may also interfere indirectly by influencing host properties so that an environment unfavorable for the establishment of a pathogen is produced (10, 15). Outbreaks of salmonellosis among domestic animals are precipitated by adverse dietary and environmental conditions (8, 9). This situation can be mimicked in an experimental mouse model (17). Many factors involved in host resistance to Salmonella infection are likely to be disrupted by adverse dietary and environmental conditions (stress). One such factor is the indigenous microbiota of the gastrointesI Present address: Department of Microbiology, University of Otago, Dunedin, New Zealand.

tinal tract. Marked changes occur in the microbial populations inhabiting the gastrointestinal tract of mice subjected to adverse dietary and environmental conditions (15). In particular, the population levels of microbes associated with epithelia (lactobacilli and fusiform-shaped bacteria) are strikingly reduced in stressed mice from the levels seen in nonstressed animals, whereas the levels of coliforms and grampositive cocci increase. Increased population levels of Salmonella typhimurium coincide with the changes in the indigenous microbiota in mice previously inoculated with the pathogen and subsequently stressed (15). Investigation of the indigenous microbiota of stressed mice has, therefore, provided clues as to which microbes in the gastrointestinal tract may be important in interfering with the establishment of S. typhimurium. We have tested the ability of indigenous Lactobacillus, Bacteroides, fusiform-shaped Clostridium, Escherichia coli, and anaerobic coccus strains to interfere with the establishment of this pathogen in the gastrointestinal tract of mice. MATERIALS AND METHODS Mice. COBS, CD-1 (Charles River Breeding Laboratories, Inc., Wilmington, Mass.) male mice were purchased directly from the supplier. The mice were housed in plastic cages with Isocaps (Isocage, Carworth, New City, N.Y.) containing AbSorb-Dri (Allied Mills, Chicago, Ill.) and given Lab-

172

VOL. 13, 1976

CECAL DAMAGE IN GNOTOBIOTIC MICE

Blox (Allied Mills) and acidified water (11) ad libitum. The mice were fed sterile Charles River diet for 1 week prior to and throughout the duration of the experiments. Germfree, CD-1 (Charles River) male mice were housed in flexible plastic isolators (Standard Safety Equipment, Palatine, Ill.) and given sterile Charles River diet and sterile water ad libitum. Bacterial strains. Indigenous bacteria were isolated from gastrointestinal specimens from CD-1 mice and maintained as follows. (i) Lactobacilli were isolated on medium 10A (12) incubated at 37 C in a candle jar for 48 h. The isolates were maintained anaerobically (oxygen-free argon-3% carbon dioxide) at 37 C on supplemented brain heart infusion agar (4). (ii) Bacteroides were isolated and maintained anaerobically on supplemented brain heart infusion agar incubated at 37 C for 7 days. (iii) A fusiform-shaped Clostridium species was isolated and maintained anaerobically on Sweet E agar (15) incubated at 37 C for 7 days. (iv) E. coli was isolated on Tergitol-7 medium plus chloride tetrazolium (13) incubated aerobically at 37 C for 24 h. (v) A gram-positive coccus was isolated and maintained in the same way as was the Clostridium species. (vi) S. typhimurium strain LT2 was obtained in pure culture from L. Rothfield (Connecticut Health Centre, Farmington, Conn.). Characteristics of isolates. The isolated strains of indigenous microbes were examined for the following characteristics: (i) Gram stain reaction and morphology; (ii) production of spores on chopped meat agar slants (4) after a 3-week period of anaerobic incubation at 30 or 37 C; (iii) production of volatile and nonvolatile acids in peptone-yeast extract-glucose medium (PYG) (4), brain heart infusion, or Sweet E without rumen fluid (4, 15); and (iv) ability to colonize the gastrointestinal tract of germfree CD-1 mice. All broth media were incubated anaerobically for 2 to 3 days at 37 C. Growth of the anaerobic coccus in liquid media was enhanced by the addition of 0.1% Tween 80. Qualitative determinations of volatile and nonvolatile acids were made using the methods of Virginia Polytechnic Institute (4) and a Dohrman Anabac gas chromatograph (Dohrman-Envirotech, Mountain View, Calif.). Injection of heat-killed S. typhimurium. Groups of germfree mice were injected with a heat-killed (70 C, 2 h) brain heart infusion broth culture (37 C, 24 h) of S. typhimurium LT2. The following protocol was used: (i) intraperitoneal injection of 0.1-ml aliquots of heat-killed broth culture and Freund complete adjuvant (Difco Laboratories, Detroit, Mich.); (ii) two intraperitoneal injections of 0.2 ml of heat-killed broth culture at 2-day intervals; and (iii) intragastric gavage with 0.1 ml of heat-killed broth culture 1 day prior to challenge with viable S. typhimurium. Experimental procedure. Nine groups of germfree mice approximately 5 weeks old were treated according to the following protocols. Group A consisted of germfree mice. Group B was injected with heatkilled S. typhimurium, and associated with broth cultures ofLactobacillus (100-1), Bacteroides (116-2), and Clostridium (109-2) by contamination offood and

173

intragastric inoculation. Seven days after the final intragastric gavage with heat-killed S. typhimurium, a challenge dose of approximately 104 living S. typhimurium was given intragastrically to each mouse. The mice were killed and examined 3 days later. Group C was injected with heat-killed S. typhimurium. Group D was challenged with S. typhimurium as described for group B. Group E was associated with indigenous bacteria as described for group B. Group F was injected with heat-killed S. typhimurium and associated as in group B, plus E. coli (104-1). Group G was injected with heat-killed S. typhimurium and associated as in group B, plus anaerobic coccus (999-2). Group H consisted of germfree mice conventionalized by intragastric gavage with homogenate from cecum and colon of CD-1 mice. Group I consisted of COBS mice. Groups C, E, F, G, H, and I were challenged with S. typhimurium as described for group B. Preparation of gastrointestinal specimens. Mice were killed with chloroform 3 days after challenge with S. typhimurium. The body weight of each mouse was recorded. At autopsy, the stomach, ileum, cecum, and colon were removed from each animal and used for microbiological culturing or histological examination. The cecum was weighed prior to further examination. Aerobic and anaerobic culture techniques. Portions (0.5 g) of the specimens described above were homogenized, diluted, cultured onto appropriate media, and incubated as described previously (15). Histological methods. Portions of gastrointestinal specimens were frozen with contents intact in 2% methylcellulose in saline. The tissues were sectioned at 4 ym on a microtome cryostat and fixed onto slides in absolute methanol. Sections were stained by a tissue Gram stain and examined by light and phase-contrast microscopy.

RESULTS Characteristics of indigenous microbes. The characteristics of the strains of indigenous bacteria isolated from COBS mice are given in Table 1. The cellular morphology of the Lactobacillus, Bacteroides, Clostridium, and anaerobic coccus in culture are shown in Fig. 1 (a through d, respectively). The morphology of the Lactobacillus strain growing in vitro (Fig. la) differs markedly from that of cells of the microbe growing in vivo (Fig. 2a). Cecal size and appearance. Cecal size following challenge with S. typhimurium varied according to the type of treatment the mice received before the challenge (Table 2). The appearance of the cecum also varied with treatment. The ceca of germfree or nonimmunized, associated mice challenged with S. typhimurium were small, edematous, and white in appearance. The ceca of mice that had been immunized with heat-killed Salmonella, but not associated with indigenous microbes, were less edematous but were of a similar size to that of germfree mice challenged with S.

174

TANNOCK AND SAVAGE

INFECT. IMMUN.

TABLE 1. Characteristics of strains of indigenous bacteria isolated from COBS, CD-I mice Isolate

Gram 4) (ref. agarmorpholmedium Spores ogy onstain

mouse gastroingermfree Major (ref. acid 4,prodtract 15) Colonization oftestinal uct(s)

Lactobacillus (100-1)

Gram-positive rod

Absent

Lactic

Bacteroides (116-2) Clostridium (109-2)

Gram-negative rod Gram-negative rod

Absent

Acetic, propionic Acetic

E. coli (104-1) Anaerobic coccus (999-2)

Gram-negative rod Gram-negative

Absent

Present

Absent

coccus

Acetic, succinic Butyric, lactic

Epithelium-associated layers in stomach. Present in lumen throughout intestinal tract. Present in cecum and colon. Only colonizes if Bacteroides present. Epithelium-associated layers in cecum and colon. Present throughout gastrointestinal tract. Present in lumen of cecum and colon.

mice challenged with S. typhimurium compared to unchallenged germfree mice. This decrease in size was accompanied by edema and abnormal coloration of the cecal tissue. Germfree mice injected with heat-killed S. typhimurium before they were challenged with the Salmonella had somewhat larger ceca than the untreated germfree controls infected with the pathogen. Nevertheless, only gnotobiotes associated with the indigenous microbes and injected with heat-killed Salmonella, before they were challenged with the pathogens, had ceca as large as those of the uninfected germfree controls. With the exception of conventionalized and COBS mice, high population levels of Salmonella were present in the gastrointestinal tract. Therefore, even when high population levels of Salmonella were present in the lumen, the cecum did not reduce in size if the mice had been both injected with heat-killed Salmonella and associated with indigenous bacteria. The mechanism of this phenomenon is not known. We believe, however, it may involve synergism between the interference exerted by the indigenous microbes and the host's resistance mechanisms. Salmonella bacteria penetrate into the tissues of conventional mice soon after intragastric inoculation even if small inocula are used (16). In COBS and conventionalized mice, the host immunological defenses are stimulated by the large populations of many types of microbes in the gastrointestinal tract (3). Thus, small numbers of Salmonella cells entering the host tissue are quickly destroyed. The germfree mouse, however, has a smaller stock of immunocompetent cells than do conventional mice (2). Even small numbers of Salmonella entering the tissues of ex-germfree animals are likely to proliferate and cause tissue destruction. It is DISCUSSION possible that the Salmonella do invade cecal tisAs has been observed by others (2), we noted sue in germfree mice, and that fewer of them a marked decrease in cecal size in germfree invade the tissue when the indigenous microbes

typhimurium. By contrast, the ceca of mice that had been both immunized and associated prior to challenge were not edematous and were similar in size and appearance to those of germfree animals which had not been exposed to S. typhimurium. Aerobic and anaerobic culture. The population levels of indigenous microbes and S. typhimurium in the gastrointestinal tract of mice subjected to various treatments are given in Tables 3 to 6. Only low population levels of Salmonella were present in the gastrointestinal tract of COBS mice, compared with germfree animals challenged with S. typhimurium. Conventionalized mice also harbored low population levels of Salmonella. Vaccination and/or association of germfree mice with indigenous microbes did not produce changes in Salmonella population levels in the gastrointestinal tract. It is noteworthy that the Lactobacillus populations were larger in immunized, associated mice than in nonimmunized, associated animals challenged with S. typhimurium. The lower population levels of Lactobacillus in nonimmunized mice may have been a consequence of tissue damage resulting from Salmonella invasion. Histology. Epithelium-associated layers of bacteria (gram-positive rods and fusiforms) were observed in histological sections of gastrointestinal specimens from associated (Fig. 2a, b), conventionalized, and COBS mice. The layers were judged to be thicker in gnotobiotic animals that had been injected previously with heat-killed S. typhimurium than in animals not so injected. The anaerobic cocci appeared to be confined to the lumen of the large intestine (Fig. 2c, d).

Cr

4% *t

.0

-

.,

.

.-

~~~~~~~~~'Po

I a

a 0

-4

C 'S

.0 ., e*S,

1

..

*

tY

*~

S *¾4t-

* t_ *;

C.X, '40PS.\

FIG. 1. Cell morphology of indigenous microbes cultured in vitro. Gram stain, x4,500. (a) Lactobacillus 100-1, 10A agar medium; (b) Bacteroides 116-2, brain heart infusion agar; (c) Clostridium 109-2, sweet E agar; (d) anaerobic coccus 999-2, brain heart infusion - Tween 80 broth. 175

t-

INFECT. IMMUN.

TANNOCK AND SAVAGE

176

1,

.%,

:..

A,

a1,

It., .'

'..

f16-

t"

C.

177


VOL. 13, 1976

TABLE 2. Cecal size in germfree and gnotobiotic CD-I mice challenged with S. typhimurium Group

A B C D E a b c

Immunized with heat-killed S. typhimurium

No Yes Yes No No

Associated with

Challenged

Clatrodium

with S. typhimuriuma

Lactobacillus, No Yes No No Yes

Cecal size (% of body wt)

pb

5.6 (4.3-8.6)c 4.6 (3.9-6.1) 2.5 (0.9-3.4) 1.7 (1.2-2.4) 1.7 (1.3-2.0)

>0.1 0.01 >0.1

No Yes Yes Yes Yes

Intragastrically. Mann-Whitney U-test. Mean and range; five mice per group.

TABLE 3. Population levels of indigenous microbes and S. typhimurium in the stomach of CD-I mice Group

Treatment

Lactobacilli

Coliforms

Salmonella

A B C D E F G H I

None Immunized, associated, challenged Immunized, challenged Challenged Associated, challenged Immunized, associated, challenged

Na 9 (8-10)b

N N N N N 8 (7-9) N

5 (Ec-8) 5 (3-8) 3 (E-5) 4 (E-5) 7 (3-8)

N N 7 (5-7) 8 (8-9) 8 (8-9) 8 (7-10) 8 (8-9)

Immunized, associated, challenged Conventionalized, challenged COBS, challenged

4 (N-4) 5 (4-7)

N

6 (4-7) N (N-E) E (E-3)

a N, None isolated. b Median and range; log10 viable bacteria per gram; five mice per group. c Isolated from enrichment culture.

TABLE 4. Population levels of indigenous microbes and S. typhimurium in the ileum of CD-I mice Group

Treatment

A B C D E F G H I

None Immunized, associated, challenged Immunized, challenged Challenged Associated, challenged Immunized, associated, challenged Immunized, associated, challenged Conventionalized, challenged COBS, challenged

Lactobacilli

Coliforms

Salmonella

Na

N N N N N 8 (7-8) N 4 (4-5) 5 (N-8)

N 6 (3-7) 5 (4-6) 5 (5-7) 5 (3-5) 6 (5-7) 5 (5-6) N (N-Ec) E (N-E)

8

(6-8)b

N N 5 (5-7) 8 (8-9) 8 (7-10) 8 (7-9) 8

N, None isolated. Median and range; log1, viable bacteria per gram; five mice per group. c Isolated from enrichment culture.

a

b

FIG. 2. Epithelium-associated layers of indigenous bacteria in the gastrointestinal tract of gnotobiotic mice. (a) Gram-positive rods (Lactobacillus 100-1) on keratinized squamous epithelium of stomach. Gram stain, x1,800. (b) Fusiform-shaped rods (Clostridium 109-2) associated with colonic epithelium. Gram stain. Phase-contrast microscopy, x4,500. (c) Gram-negative fusiforms (Clostridium 109-2) associated with colonic tissue (bottom, left). Gram-positive cocci (anaerobic coccus 999-2) restricted to lumen. Gram stain, x1,800. (d) Same field as (c), but viewed with phase-contrast microscopy. Fusiform-shaped rods are more apparent. x1,800.

178

TANNOCK AND SAVAGE

INFECT. IMMUN.

TABLE 5. Population levels of indigenous microbes and S. typhimurium in the cecum of CD-1 mice

Group

Treatment

Lactobacilli

Coliforms

TotalobeS anaer-

Salmonella

A B C D E F G H I


Nb 8 (8-9)' N N 6 (5-7) 8 8 (8-9) 8 (8-9) 8 (8-9)

N N N N N 9 (8-9) N 5 (5-8) 6 (4-8)

N 11 (10-11) N N 9 (9-10) 11 (10-11) 11 (10-11) 10 (10-11) 11 (10-11)

N 9 (8-9) 10 (9-10) 9 (9-10) 8 (8-9) 8 8 (8-9) N (N-Ed) E (E-3)

a

Sweet E agar. N, None isolated. C Median and range; log,0 viable bacteria per gram; five mice per group. d Isolated from enrichment culture. a

b

TABLE 6. Population levels of indigenous microbes and S. typhimurium in the colon of CD-1 mice Group

Treatment

Lactobacilli

Coliforms

Totalob,esa anaer-

Salmonella

A B C D E F G H I


Nb 8 (7-8)' N N 6 (5-7) 8 8 (7-8) 8 8 (7-9)

N N N N N 8 (8-9) N 5 (4-5) 5 (4-7)

N 9 (8-10) N N 9 (9-10) 10 (10-11) 10 (9-11) 9 (8-10) 11 (10-11)

N 8 9 (9-10) 8 (8-9) 8 (7-8) 8 8

N (N-3) Ed (E-3)

Sweet E agar. N, None isolated. ' Median and range; log,0 viable bacteria per gram; five mice per group. d Isolated from enrichment culture. a

b

are present than when they are absent. Those Salmonella that do enter the tissue may be prevented from rapidly proliferating by the host resistance mechanisms that are stimulated when the animals are injected with heat-killed Salmonella. Thus, the indigenous microbes and the host's resistance mechanisms would be acting synergistically to retard or prevent S. typhimurium infection. Such a phenomenon has been described in resistance of mice to V. cholerae infections (14). Since only low population levels of Salmonella were present in the gastrointestinal tract of conventionalized or COBS mice challenged with S. typhimurium, other interference effects, besides the phenomenon observed in these experiments, must also be operating in thse animals. Our experiments show that E. coli and anaerobic cocci, as represented by our strains, are not involved in this interference mechanism. Thus, the interference we observed was exerted by anaerobic bacteria that associate with epithelial surfaces. We postulate that

host resistance to infection of the gastrointestinal tract by S. typhimurium consists of at least three parts: (i) interference in the lumen by as yet undescribed mechanisms; (ii) interference at the epithelial surface of host tissue mediated by epithelium-associated bacteria; and (iii) host resistance mechanisms operating in the tissue and possibly at epithelial surfaces. ACKNOWLEDGMENTS This investigation was supported by Public Health Service research grant AI-08254 from the National Institute of Allergy and Infectious Diseases and by Public Health Service International Research Fellowship F05 TWO 1855-02 from the Fogarty International Center. LITERATURE CITED 1. Abrams, G. D., and J. E. Bishop. 1967. Effect of the normal microbial flora on gastrointestinal motility. Proc. Soc. Exp. Biol. Med. 126:301-304. 2. Gordon, H. A., and L. Pesti. 1971. The gnotobiotic animal as a tool in the study of host microbial relationships. Bacteriol. Rev. 35:390-429. 3. Hess, M. W., H. Cottier, B. Sordat, D. D. Joel, and A.

VOL. 13, 1976

4. 5.

6. 7.

8. 9. 10.


D. Chanana. 1973. The intestinal barrier to bacterial invasion, p. 447-451. In W. Braun and J. Ungar (ed.), Non-specific factors influencing host resistance. Karger, Basel. Holdeman, L. V., and W. E. C. Moore (ed.). 1972. Anaerobe laboratory manual. Virginia Polytechnic Institute Anaerobe Laboratory, Blacksburg, Va. Maier, B. R., A. B. Onderdonk, R. C. Baskett, and D. J. Hentges. 1972. Shigella, indigenous flora interactions in mice. Am. J. Clin. Nutr. 25:1433-1440. Meynell, G. G. 1963. Antibacterial mechanisms of the mouse gut. II. The role of Eh and volatile fatty acids in the normal gut. Br. J. Exp. Pathol. 44:209-219. Miller, C. P., and M. Bohnhoff. 1963. Changes in the mouse's enteric microflora associated with enhanced susceptibility to Salmonella infection following streptomycin treatment. J. Infect. Dis. 113:59-66. Robinson, R. A., and W. A. Royal. 1971. Field epizootiology of Salmonella infection in sheep. N.Z.J. Agric. Res. 14:442-456. Salisbury, R. M. 1958. Salnonella infections in animals and birds in New Zealand. N.Z. Vet. J. 6:76-86. Savage, D. C. 1972. Survival on mucosal epithelia, epithelial penetration and growth in tissues of pathogenic bacteria, p. 25-57. In H. Smith and J. H.

11.

12.

13. 14. 15.

16.

17.

179

Pearce (ed.), 22nd Symp. Soc. Gen. Microbiol. Cambridge University Press, Cambridge, England. Savage, D. C., J. S. McAllister, and C. P. Davis. 1971. Anaerobic bacteria on the mucosal epithelium of the murine large bowel. Infect. Immun. 4:492-502. Schaedler, R. W., and R. Dubos. 1962. The faecal flora of various strains of mice. Its bearing on their susceptibility to endotoxin. J. Exp. Med. 115:11491160. Schaedler, R. W., R. Dubos, and R. Costello. 1965. The development of the bacterial flora in the gastrointestinal tract of mice. J. Exp. Med. 122:59-66. Shedlofsky, S., and R. Freter. 1974. Synergism between ecologic and immunologic control mechanisms of intestinal flora. J. Infect. Dis. 129:296-303. Tannock, G. W., and D. C. Savage. 1974. Influences of dietary and environmental stress on microbial populations in the murine gastrointestinal tract. Infect. Immun. 9:591-598. Tannock, G. W., and J. M. B. Smith. 1971. A Salmonella carrier state involving the upper respiratory tract of mice. J. Infect. Dis. 123:502-506. Tannock, G. W., and J. M. B. Smith. 1972. The effect of food and water deprivation (stress) on Salmonellacarrier mice. J. Med. Microbiol. 5:283-289.