Role of Cell Wall Structure of Salmonella in the Interaction with Phagocytes

Transcription

1 INFECTION AND IMMUNITY, Sept. 1970, p Copyright 1970 American Society for Microbiology Vol. 2, No. 3 Printed in U.S.A. Role of Cell Wall Structure of Salmonella in the Interaction with Phagocytes D. FRIEDBERG AND M. SHILO Department of Microbiological Chemistry, Hebrew University-Hadassah Medical School, Jerusaleml, Israel Received for publication 25 May 1970 Wild-type and cell wall mutants of Salmonella were examined for sensitivity to ingestion and intracellular killing in vitro by mouse peritoneal macrophages and guinea pig polymorphonuclear leukocytes. A complete polysaccharide core of the cell wall is important for resisting ingestion and intracellular killing, and the presence of the 0-specific side chains contributes further resistance. Uridine diphosphate-gal-4-epimeraseless mutants grown on galactose-supplemented medium, rendering them smooth phenotypes, showed resistance to ingestion and intracellular killing similar to the wild type. The virulence of bacteria in many phagocytic systems has been stressed as an important factor affecting ingestion and their subsequent intracellular fate. Most workers found that avirulent organisms are ingested more efficiently by phagocytic cells than are virulent ones. This was shown for gram-negative bacteria such as Salmonella (11, 12, 24) and Brucella (2), and for gram-positive bacteria such as Staphylococcus (3). Moreover, a correlation was found between the virulence of Salmonella and of Escherichia coli and their survival within macrophages (10-12, 18, 23-25). In contrast, Furness (9) showed that virulence influences the intracellular multiplication within macrophages, and not the rate of ingestion. After being opsonized with specific antiserum, virulent strains of Salmonella and of E. coli become as sensitive as avirulent strains to ingestion and intracellular killing by macrophages (11, 12) and to ingestion by Myxamoeba (14). In the interaction of the bacteria and phagocyte, the bacterial cell wall structure must be of decisive influence. Recently, a number of investigators attempted to determine which of the cell wall components are responsible for this effect. By using wild-type Salmonella typhimurium and its mutants with deficiencies of sugars in the cell wall polysaccharide, Nakano and Saito (20, 21) showed that the clearance rate, virulence, and immunogenicity of the different strains in mice varied. The cell wall polysaccharide composition of E. coli was found to affect ingestion by phagocytic cells and its virulence for mice (15). Similar results were obtained with wild-type and cell wall mutants of S. minnesota ingested by Myxamoeba (14). 279 Even at the subcellular level, a relationship between the cell wall structure of S. typhimurium and E. coli strains and their resistance to the bacteriocidal lysosomal fraction of polymorphonuclear (PMN) leukocytes (1) have been demonstrated (6, 7). To obtain more detailed information on the influence of bacterial cell wall composition in ingestion and survival within the phagocyte, we studied two species of Salmonella and their cell wall mutants in their interaction with mouse peritoneal macrophages and guinea pig peritoneal PMN leukocytes in vitro. The use of uridine diphosphate (UDP)-gal-4-epimeraseless mutants grown under various conditions allowed us to assess in yet another way the importance of cell wall structure in the same strain. MATERIALS AND METHODS Bacteria and culture conditions. Wild-type S. typhimurium and its cell wall mutants with defined deficiencies of sugars in the lipopolysaccharide structure were obtained through the courtesy of W. Braun (Rutgers Univ., Institute of Microbiology, New Brunswick, N.J.). Wild-type S. enteritidis and its UDP-gal4-epimeraseless mutant (11-1-M) were obtained through the courtesy of Dr. Nikaido (Department of Microbiology, Univ. of California at Berkeley). The cell wall composition of the strains is described in Table 1. Detailed information of Salmonella cell wall structure is reviewed by Luderitz et al. (13) and by Osborn (22). Cells of all the strains were grown and harvested in the logarithmic phase as previously described (6). Growth of the UDP-gal4-epimeraseless mutants of S. typhimurium and S. enteritidis 11-1-M in galactose supplemented medium which results in cells with

2 280 FRIEDBERG AND SHILO INFEC. IMUN. TABLE 1. Description of strains employed Strain Genotype Cell wall composition Salmonella typhimurium LT2 S (wild type) Contains basal core and side chains S. typhimurium TV-119 Ra Lacks side chains, but has complete basal core S. typhimurium TV-161 Rb Lacks side chains and N-acetyl glucosamine (GlcNAc) S. typhimurium LT2-M1 Rc Lacks side chains, GlcNAc, and galactose S. typhimurium SL-1032 Rd Lacks side chains, GlcNAc, galactose, and glucose S. typhimurium G-30/C-21 Re Lacks side chains, GlcNAc, galactose, glucose, and heptose-phosphate S. enteritidis 11 S (wild type) Contains basal core and side chains S. enteritidis 11-1-M Rc Lacks side chains, GlcNAc, and galactose complete wall structure (8) was previously described (6) Ċollection and treatment of phagocytes. PMN leukocytes were obtained from guinea pigs as previously described (5). Macrophages were harvested from the peritoneal cavity of normal female albino mice in saline containing 5,ug of sodium heparin (Nutritional Biochemical Corp., Cleveland, Ohio) per ml as described previously (5). The cell suspensions were centrifuged (200 X g for 5 min in a horizontal International centrifuge), and the cellular pellet was washed once and resuspended in medium 199 (19) to the desired cell concentration as measured in a hemocytometer. Interaction of bacteria with phagocytes (reaction mixture). From 3 X 107 to 4 X 107 bacteria were added to each milliliter of the reaction mixture containing an equal concentration of phagocytes suspended in medium 199 (19) containing 10% newborn calf serum (Microbiological Associates, Inc., Bethesda, Md.). To suppress the extracellular multiplication of the bacteria, chloramphenicol (Abic Ltd., Israel) was added at a bacteriostatic concentration (6 or 12,g/ml). From 2 to 3 ml of the reaction mixture was incubated in 25-ml siliconized Erlenmeyer flasks at 37 C in a rotatory shaker bath under an atmosphere of air mixed with CO2 (5%) saturated with water. Estimation of ingestion and intracellular killing of the bacteria within the phagocytes. Samples (0.2 ml) were taken from the reaction mixture into 3.8 ml of medium 199 after different times of incubation. The phagocytes were sedimented by centrifugation (200 X g for 5 min) and washed once, and viable counts were carried out on the pooled upper parts of the supernatants (containing uningested bacteria). The number of ingested bacteria (I) was calculated by subtracting the number of viable extracellular bacteria (E) at a given time from the total number of viable bacteria (T) at zero time (T - E = I). The percentage of ingestion is expressed by the equation, I/T X 100. Viable counts of the surviving ingested bacteria were carried out after the washed sedimented phagocytes were disrupted by addition of 2.5% (w/v) saponin (The British Drug Houses, Ltd.). The saponin-treated suspensions were immediately diluted to avoid saponin damage to the bacteria. The percentage of intracellular survival was calculated from the number of viable bacteria (S) released from the washed phagocytes divided by the number of ingested bacteria (I) (S/I X 100). RESULTS Ingestion and intracellular killing of Salmonella typhimurium LT2 and its cell wail mutants by mouse peritoneal macrophages. The cell wall mutants of S. typhimurium with an incomplete core (Rb, Rc, Re types) were rapidly ingested by peritoneal macrophages (Fig. IA). At least 85% of the population of these rough strains was ingested during the first 15 min of incubation. Ingestion of the wild-type S. typhimurium, on the other hand, showed a lag, and only 40% of the bacterial population was ingested during the entire incubation period (2 hr). The rough Ra type strain, possessing a complete core but lacking the 0-specific side chains, showed a rapid initial rate of ingestion but reached a lower level after the first 15 min of incubation (60% of the Ra type population as opposed to 85 to 100% of the population of other rough strains). Like the wildtype bacteria, the galactose-grown UDP-gal-4- epimeraseless mutant (LT2-M1-gal, S phenotype) was ingested only after a lag. However, its rate of ingestion later accelerated, and 80% of the bacteria were ingested by the end of 2 hr. The survival of the ingested bacteria within the macrophages is shown in Fig. 1B. The rough Rb, Rc, Re type strains with an incomplete core were killed rapidly. During the first 15 min of incubation, only 2 to 12% of these bacterial populations survived in comparison with 60% of the wild-type population. The rate of killing of these mutants

3 VOL. 2, 1970 SALMONELLA CELL WALL STRUCTURE 281 became slower thereafter; after 2 hr of incubation, only 0.2 to 2% of these bacteria survived, as compared to more than 10% of the wild-type population. The high sensitivity of the Rb type mutant is noteworthy, but its cause is not clear. The Ra type mutant was killed at an initial rate similar to that of the other mutants, but the percentage of final survivors (at 2 hr) was greater. The galactose- grown UDP-gal4-epimeraseless mutant (LT2- Ml-gal, S phenotype) was killed during the first 30 min at a slower rate than cells of the same strain (Rc type) grown on unsupplemented medium; however, the rate of killing of the galactosegrown mutant remained steady up to 2 hours, but that of the Rc type decreased. Thus, the number of survivors of both phenotypes of this strain did c 0 ub "I A Time (minutes) Time (minutes) FIG. 1. Ingestion (A) and killing (B) of wild-type Salmonella typhimurium and its cell wall mutants by mouse peritoneal macrophages. Bacterial strains are as designated in Table 1; LT2-MI (gal) is the LT2-MJ (Rc) strain after growth on galactose. Numbers of survivors were determined by viable count of bacteria at given times of incubation of the reaction mixture. A 100 _ so c ~~~~~e 30~~~~~~~~~~U 70- ~~~~.- U. tn 10 5 B B zu Time (mirnutes) Time (minutes) FIG. 2. Ingestion (A) and killing (B) of two phenotypes ofsalmonella enteritidis JJ-I-M. Symbols: 0, 11-1-M (Rc); X, 11-1-M (gal)-s phenotype, the Rc strain after growth on galactose.

4 282 FRIEDBERG AND SHILO INFEC. IMMUN. B -- en Time (minutes) T i me ( minutes) FIG. 3. Ingestion (A) and killing (B) of wild-type Salmonella typhimurium and its cell wall mutants by guinea pig PMN leukocytes. Bacterial strains are as designated in Table 1; LT2-MI (gal) as in Fig. 1. not differ by the end of the incubation period (2 hours). Phagocytosis of Salmonella enteritidis cell wall mutant (11-1-M) by mouse peritoneal macrophages. Ingestion (Fig. 2A) and intracellular killing (Fig. 2B) of cells of S. enteritidis mutant grown under standard conditions (Rc type) and grown on galactose-supplemented medium (S phenotype) were compared. Ingestion of the Rc type cells proceeded rapidly, and 99.9% of the population was ingested during the incubation period (2 hr), whereas the galactose-grown cells (S phenotype) were ingested after a lag and showed a lower per cent of ingested cells (Fig. 2A). The galactose-grown cells also showed increased resistance to intracellular killing, and 30% of the population survived after 2 hr incubation, whereas only 1% of the Rc type cells survived within the macrophages (Fig. 2B). In general, the S phenotypes of both S. typhimurium LT2-M1 and S. enteritidis 11-1-M were more resistant to ingestion and killing than the corresponding Rc cells. However, the differences in resistance between the two types of cells (S phenotype and Rc type) were greater with S. enteritidis than in the respective strains of S. typhimurium(fig. 1 and 2). Phagocytosis of wild-type salmonellas and their cell wall mutants by guinea pig PMN leukocytes. All the rough mutants of S. typhimurium were much more rapidly ingested by PMN leukocytes than the wild-type cells or the galactose-grown UDP-gal-4-epimeraseless mutant (S phenotype; Fig. 3A). The rates of intracellular killing of the rough U)~~~~~n o > 10 Rb 5 Re $ S 9 RcLT2-M Time (hours) FIG. 4. Intracellular fate of wild-type Salmonella typhimurium and its cell wall mutants in guinea pig peritoneal PMN leukocytes. After 30 min of incubation of leukocytes with the bacteria without chloramphenicol, the leukocytes were washed three times and resuspended in fresh medium lacking the antibiotic. The per cent of survivors at zero time (after resuspension in fresh media) equals 100. Bacterial strains are as designated in Table 1; LT2-MJ (gal) as in Fig. 1.

5 VOL. 2, 1970 SALMONELLA CELL WALL STRUCTURE 283 (A L PMN leukocytes from guinea pigs, show the importance of the cell wall structure of salmonellas in the interaction of the bacteria with phagocytes. The significance of intracellular events in the early 11 (S) stages was stressed by Hsu et al. (10) who showed that smooth and rough strains of S. typhimurium differ in the extent of intracellular killing rather than in subsequent multiplication. Rough strains of S. typhimurium lacking sugars in the cell wall polysaccharide were more sensitive than the smooth wild type to ingestion and early intracel (Rc) lular killing by mouse peritoneal macrophages and guinea pig PMN leukocytes in vitro. The rough strain, posessing a complete core by lacking 0-specific side chains (Ra type), showed intermediate sensitivity (especially pronounced in intracellular killing) between the rough strains with incomplete cores (Rb, Rc, Rd, Re types) -n 03 5 and the more resistant wild type (S type). Thus, it appears that the possession of a complete core is important for resisting ingestion and for intracellular survival and that the presence of the 0- specific side chains contributes further resistance. These results agree with the observation of Medearis et al. (15), showing that wild-type E. coli 0111 with 0-specific side chains resists ingestion by PMN leukocytes in vitro and by macrophages in vivo better than its call wall mutant L2 1 3 Time (hours) lacking 0-specific side chains, which in turn was more resistant than the mutant J-5 with an in- FIG. 5. Intracellular fate of wild-typ;e Salmonella complete polysaccharide core. In addition, S. eniteritidis 11 (S) and its cell wall miutant 11-1-M minnesota with 0-specific side chains resists (Rc) in peritoneal PMN leukocytes. Detuails of experi- phagocytosis by the amoeba Dictyostelium ment as in Fig. 4. discoideum, whereas all its cell wall mutants are susceptible (14). Nakano and Saito (20) suggested strains lacking sugars in the core of the cell wall that the tetrasaccharide sequences of abequosyltypes) were mannosyl-rhamnosyl-galactose on the bacterial polysaccharide (Rb, Rc, Rd, and Re greater than those of the wild-type and the ga- surfaces of S. typhimurium might be the deter- (S pheno- minant for preventing engulfment of the bacteria lactose-grown epimeraseless mutantt type; Fig. 3B). The Ra type mutant seeems to have by the reticuloendothelial system (RES). Bacterial intermediate sensitivity. multiplication in the mouse was shown by them Figure 4 shows similar differen4 ces between to be correlated with the presence of the 0-specific rough and smooth types of S. typhimrurium in the side chains and all the sugars in the core (21). rate of killing in an experimental system lacking The enhanced resistance to ingestion and chloramphenicol. This indicates thiat chloram- intracellular killing of the galactose-grown S. phenicol does not affect the differentt sensitivities typhimurium LT2 Ml and S. enteritidis 11-1-M of the various strains. Likewise, the wild-type S. mutant strains corroborates the results obtained enteritidis and its cell wall mutant (RLc type) were with bacterial cells of different genotypes. The killed at similar rates during the earlier period of phenotypic transformation achieved by growing incubation (up to 1 hr) in the abserice of chlor- the rough mutants on galactose-supplemented amphenicol in the reaction mixture (Fig- 5). medium permits complete cell wall synthesis (8). Later, however, the wild-type bacetri a multiplied, Medearis et al. (15) also found that, after such a whereas the number of surviving mutant bacteria continued to decrease. phenotypic modification, the susceptibility of an E. coli J5 cell wall mutant to phagocytosis is diminished. Since complete cell wall structure in DISCUSSION these strains depends on the availability of galacwith mouse tose, they might have attenuated virulence. In- The results of experiments made peritoneal macrophages, as well as wiith peritoneal deed, it has been shown that such an S phenotype

6 284 FRIEDBERG AND SHILO INFEC. IMMUN. of E. coli is less virulent for mice than the wild type (15). The prolonged survival of the S phenotypes within the phagocytes suggests that such strains might be used in immunization. A correlation has been found between the cell wall composition and the infectivity of S. typhimurium (21) and of E. coli (15) to mice. Nakano and Saito (21) found that the number of 0-specific side chains is important for the infectivity of S. typhimurium. Although there are some differences related to details of the bacterial cell wall composition in the various systems of phagocytosis and virulence examined, it is obvious that a major determinant is the polysaccharide composition of the cell wall. Results of in vitro studies show that peritoneal macrophages ingest and kill avirulent strains of Salmonella much more rapidly than virulent strains (10-12, 17, 18, 23-25). These differences may now be interpreted in terms of differences in the polysaccharide composition in the cell walls of the avirulent and virulent strains. Since salmonellas are considered facultative intracellular parasites of the RES and multiply within macrophages, most studies on their phagocytosis and virulence have been carried out with macrophages (12, 23, 24). Our results show that the cell wall composition of the Salmonella strains affects phagocytosis by PMN leukocytes as it does phagocytosis by macrophages. Moreover, the differences in susceptibility to intracellular killing among the bacteria with different cell wall compositions seem to be more pronounced when examined with PMN leukocytes. Since metabolic and bactericidal activities of phagocytes are known to be influenced by the process of ingestion (1, 4, 16), the question must be considered whether differences in intracellular killing are possibly a secondary result of changes in the phagocytes rather than due directly to the different susceptibilities of the bacterial strains. This possibility is unlikely in the case of PMN leukocytes at least, since it has been shown that wild-type S. typhimurium and its cell wall mutants differ in susceptibility to the bactericidal action of the lysosomal fraction of PMN leukocytes (6). However, the Ra type mutant (TV-1 19) showed resistance similar to the wild type in the subcellular bactericidal system (5), whereas at the cellular level it appears to be more susceptible than the wild type. In this particular case, it might be that the process of ingestion renders the bacteria somewhat more sensitive. ACKNOWLEDGMENTS We thank W. Braun and D. Sulitziano for critical reading of the manuscript, and B. Golek for her aid in editing it. We also thank D. Varon for able technical assistance. LITERATURE CITED 1. Allen, J. M Lysosomes in bacterial infection. In J. T. Dingle and H. B. Fell (ed.), Lysosomes in biology and pathology. North Holland Publishing Co., Amsterdam, London. 2. Braun, W., A. Pomalis-Lebr6n, and W. R. Stinebring Interactions between mononuclear phagocytes and Brucella abortus strains of different virulence. Proc. Soc. Exp. Biol. Med. 97: Cohn, Z. A., and S. I. Morse Interactions between rabbit polymorphonuclear leucocytes and staphylococci. J. Exp. Med. 110: Cohn, Z. A., and S. I. Morse Functional and mnetabolic properties of polymorphonuclear leucocytes. J. Exp. Med. 111: Friedberg, D., and M. Shilo Infection-promoting activity of mouse high molecular peritoneal weight infections microbial with polysaccharide Pasteurella pestis. on J. Infec. Dis. 115: nuclear leukocytes. I. Role of cell wall composition of Salmonella typhimurium. Infec. Imnmaun. 1: Friedberg, D., and M. Shilo The interaction negative bacteria with lysosomal fraction of polymorpho- of gram- 7. Friedberg, D., 1. Friedberg, and M. Shilo The interaction of gram-negative bacteria with lysosomal fraction of polymorphonuclear leukocytes. II. Changes in the cell envelope of Escherichia coli. Infec. Immun. 1: Fukasawa, T., and H. Nikaido Galactose-sensitive mutants of Salmonella. If. Bacteriolysis induced by galactose. Biochim. Biophys. Acta 48: Furness, G Interaction between Salmonzella typhimuriut and phagocytic cells in cell culture. J. Infec. Dis. 103: Hsu, H. S., and A. S. Radcliffe Interactions between macrophages of guinea pigs and salmonella. I. Fate of Salmonella typhimurium within macrophages of normal guinea pigs. J. Bacteriol. 96: Jenkin, C. R., and B. Benacerraf In vitro studies on the mouse peritoneal anid and Escherichia J. Med. interaction between strains of Salmonzella macrophages coli. Exp. 112: and D. Rowley. Jenkin, C. R., Basis for immunity to and the typhoid in mice question of "cellular immunity." Bacteriol. Rev. 27: Liuderitz, O., A. M. Staub, and 0. Westphal Immunochemistry of 0 and R antigens of Salmonzella and related Enterobacteriaceae. Bacteriol. Rev. 30: Malchow, D., 0. Luderitz, 0. Westphal, G. Gerich, and V. Reidel Polysaccharide in Vegetativen und Aggregationsreifen Amoben von Dictyostelium discoideum. Eur. J. Biochem. 2: Medearis, D. N., B. M. Camitta, and E. C. Heath Cell wall composition and virulence in E. coli. J. Exp. Med. 128: Melly, M. A., J. B. Tomison, and D. E. Rogers Fate of staphylococci within human leukocytes. J. Exp. Med. 112: MitsuLhashi, S., I. Sato, and T. Tanaka Experimental salmonellosis. Intracellular growth of Salmonella enteritidis ingested in mononuclear phagocytes of mice, and cellular basis of immunity. J. Bacteriol. 81: Morello, J. A., and E. E. Baker Interaction of Salmonella with phagocytes in vitro. J. Infec. Dis. 115: Morgan, J. F., H. J. Morton, and R. C. Parker Nutrition of animal cells in tissue culture. I. Initial studies on a synthetic medium. Proc. Soc. Exp. Biol. Med. 73: Nakano, M., and K. Saito The chemical compositions in the cell wall of Salmonella typhimurium affecting the clearance-rate in mouse. Jap. J. Microbiol. 12:

7 VOL. 2, 1970 SALMONELLA CELL WALL STRUCTURE Nakano, M., and K. Saito Chemnical components in the cell wall of Salmonella typhimwrium affecting its virulence and immunogenicity in mice. Nature 222: Osborn, M. J Structure and biosynthesis of the bacterial cell wall. Annu. Rev. Biochem. 38: Roantree, R. J Salmonella 0 antigens and virulence. Annu. Rev. Microbiol. 21: Rowley, D Phagocytosis. Advan. Immunol. 2: Rowley, D., and J. L. Whitby The bactericidal activity of mouse macrophages in vitro. Brit. J. Exp. Pathol. 40: Younger, J. S., and D. S. Feingold Interferon production in mice by cell wall mutants of Salmonella typhimurium. J. Virol. 1: