Francisella tularensis--a model for studies of the immune response to intracellular bacteria in man.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to G Sandström.
Explore the source record for details and available documents.
In this study 17 strains of Porphyromonas gingivalis, both reference and clinical isolates, were investigated for their in vitro interaction with human polymorphonuclear leukocytes, hydrophobicity, density, and virulence in a mouse model. The results of the phagocytosis, hydrophobicity, and density experiments showed that P. gingivalis strains could be divided into two distinct groups. One group of strains were readily attached and phagocytosed when exposed to the leukocytes. These bacteria were hydrophobic and had a higher buoyant density than the other group, which were poorly phagocytosed, had a low buoyant density, and were hydrophilic. This latter group also exhibited an extracellular meshwork resembling a glycocalyx when examined by electron microscopy. There were also significant differences between strains in the mouse pathogenicity model. Two strains caused an invasive, spreading infection compared with the other 15 strains which produced small, localized abscesses. There was no clear correlation between the results of the phagocytosis assay and the virulence of the bacteria when injected subcutaneously in mice. Resistance to phagocytosis may be important for survival of these bacteria, but it does not in itself imply the ability to cause damage to the host.
A T-cell-stimulating, membrane-located 17-kDa protein of the live vaccine strain Francisella tularensis LVS has previously been cloned and sequenced. In the present study, it is shown to be a lipoprotein. When F. tularensis was grown in the presence of [3H]palmitate, several proteins of the organism, including a 17-kDa protein, were radiolabeled. The labeled 17-kDa protein was found by Western blot (immunoblot) analysis to be identical to the cloned protein. It was located in the detergent phase after partitioning with the nonionic detergent Triton X-114, thereby behaving like a hydrophobic integral membrane protein. The protein was predominantly hydrophilic and contained no putative transmembrane domain. The presence of fatty acids is therefore the probable explanation of the membrane location of the 17-kDa protein. The amino acid sequence of the 17-kDa protein contains the tetrapeptide Leu-Ala-Ser-Cys, which is a recognition sequence of the lipoprotein signal peptidase. Globomycin, a specific inhibitor of the peptidase, inhibited maturation of the 17-kDa lipoprotein. The protein incorporated [3H]palmitate also when expressed by Escherichia coli. The 17-kDa lipoprotein was recognized not only by T cells but also by serum antibodies of F. tularensis-primed individuals.
A 17-kDa major membrane protein of the facultative intracellular bacterium Francisella tularensis is recognized by T cells from individuals previously undergoing tularemia or tularemia vaccination. Here the nucleotide sequence of the encoding gene is presented. The A + T rich sequence deduced a protein of 149 amino acids with a predicted m.w. of 15,772. Evidence is presented suggesting that the 17-kDa protein is a lipoprotein. Also, homology was found to a peptidoglycan-associated lipoprotein of Escherichia coli. Thirteen overlapping synthetic peptides encompassing the entire deduced protein were tested on lymphocytes from F. tularensis-primed individuals. Three peptides induced a T cell response, as disclosed by lymphocyte proliferation and production of IL-2 and IFN-gamma. Out of five responding individuals, only one recognized more than one of the peptides. A murine mAb recognized an epitope separate from the T cell-reactive peptides. The identified T cell epitopes are presumably relevant to the host defense against F. tularensis, especially because IFN-gamma is known to enhance the antimicrobial activity of the host against intracellular bacteria.
Tularemia is a zoonotic disease, occurring throughout the Northern Hemisphere. The causative agent, the bacterium Francisella tularensis, is represented by two main types. Type A is found in North America, whereas type B is mainly found in Asia and Europe and to a minor extent in North America. No routine technique for rapid diagnosis of tularemia has been generally applied. We have partially sequenced 16S rRNAs of two F. tularensis strains, as well as the closely related Francisella novicida. Of 550 nucleotides analyzed, only one difference in 16S rRNA primary sequence was found. This 16S rRNA analysis enabled the construction of oligonucleotides to be used as genus- and type-specific probes. Such probes were utilized for the establishment of a method for rapid and selective detection of the organism. This method allowed identification of Francisella spp. at the level of genus and also discrimination of type A and type B strains of F. tularensis. The analysis also permitted the detection of F. tularensis in spleen tissue from mice infected with the bacterium. The results presented will enable studies on the epizootiology and epidemiology of Francisella spp.
The currently used live vaccine strain Francisella tularensis LVS was derived several decades ago from a wild strain of the species. In the present report, several membrane polypeptides of LVS are shown to be recognized by T cells from individuals immunized by natural infection with F. tularensis. Bacterial membranes of a capsule-deficient mutant of LVS were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Thereafter, gels were divided into seven fractions, each fraction containing a different number of peptide bands. From other gels, four bands were excised, each containing one major polypeptide. Eluates of each fraction and of each polypeptide band induced a proliferative response and an interleukin-2 response in lymphocytes from most of the individuals. When the lymphocytes were separated after induction, most of the proliferative response was found to occur in CD4+ T cells. Lymphocytes from nonimmune individuals responded poorly to all membrane polypeptides. To study the possible heterogeneity of antigen determinants among the polypeptides, T-cell clones were raised towards F. tularensis and tested for proliferative response to the four major membrane polypeptides. Five clones, all CD4+ CD8-, responded to one or more of the polypeptides, each clone with a unique pattern of response. In conclusion, F. tularensis possesses a high number of T-cell-reactive membrane polypeptides. There seems to be a heterogeneity of T-cell determinants among these polypeptides. Determinants involved in immunization by natural infection are well conserved in LVS.
Tularemia meningitis occurred in a previously healthy 64-year-old man from Northern Sweden. The patient was immunocompetent and recovered without sequelae. The bacterial isolate conformed by most criteria with the Euro-Asian biovar Francisella tularensis var. palaearctica. Tularemia meningitis may mimic meningitis caused by Mycobacterium tuberculosis.
Francisella tularensis septicemia has previously been reported in the USA, where Francisella tularensis biovar tularensis (type A) occurs endemically. The present report describes two cases of septicemia caused by the Euro-Asian Francisella tularensis biovar palaearctica (type B). The patients lived in an endemic region of Sweden and presented with high fever and pneumonia. Both patients recovered. The ability of Francisella tularensis biovar palaearctica to cause bacteremia has consequences for laboratory safety.
The isolation and expression in Escherichia coli of a gene encoding a T-cell stimulating 17 kiloDalton (kDa) membrane protein of Francisella tularensis is described. A genomic library of DNA from the live vaccine strain LVS of F. tularensis was constructed in the E. coli expression vector phage lambda gt11. The library was probed with antibodies directed against the 17 kDa protein. One recombinant phage was isolated, containing a 2.8 kilobase (kb) DNA insert. The insert was cleaved and a resulting 1.2 kb fragment was found to express the 17 kDa protein. The 1.2 kb fragment was inserted in the high copy number plasmid pUC18 and expressed in E. coli. Membrane preparations of these bacteria induced a response in T cells from F. tularensis-primed individuals but not in T cells from non-primed individuals. The cloned gene may become useful in studies on host interaction with F. tularensis and enable a precise identification of bacterial structures involved in the T-cell response.
Eight monoclonal antibodies were produced by immunizing mice with a high molecular-weight antigen from Francisella tularensis. The antigen consists of carbohydrates and proteins and was found to contain several different epitopes. By treating the antigen with periodate or proteinase K, it was found that the majority of antibody-reactive epitopes was mainly dependent on the carbohydrate part of the antigen. On intact F. tularensis bacteria the epitopes seem to be exposed at the bacterial surface. Both main biovars of F. tularensis, tularensis (type A) and palaearctica (type B), exhibited the epitopes.
The live vaccine strain (LVS) of Francisella tularensis is killed by human polymorphonuclear leukocytes as a result of strictly oxygen-dependent mechanisms (S. Löfgren, A. Tärnvik, M. Thore, and J. Carlsson, Infect. Immun. 43:730-734, 1984). We now report that a capsule-deficient (Cap-) mutant of LVS survives in the leukocytes. In contrast to the encapsulated parent strain, the Cap- mutant was avirulent in mice and was susceptible to the bactericidal effect of nonimmune human serum. The mutant was killed by serum as a result of activation of the classical pathway of complement by naturally occurring immunoglobulin M. This killing by serum was mitigated by the presence of human polymorphonuclear leukocytes. After opsonization in complement component C5-deficient nonimmune serum, the Cap- mutant was ingested and survived in the leukocytes. Under these conditions, the parent strain was killed. The leukocytes responded to both the parent and the Cap- strain with a very low chemiluminescent response. Only the response to the parent strain was inhibited by superoxide dismutase. When the Cap- mutant was opsonized with immunoglobulin G, it induced a higher and superoxide dismutase-inhibitable chemiluminescent response and was killed by the leukocytes. In conclusion, the capsule of F. tularensis LVS seemed to protect this organism against the bactericidal effect of serum. When deprived of the capsule, the organism failed to induce an antimicrobial response in polymorphonuclear leukocytes and survived in the leukocytes. Survival in phagocytes is a key characteristic of intracellular parasites. The Cap- mutant of F. tularensis may become a useful tool in experiments to explain the differences between pathways of ingestion of intracellular parasites, evidenced by the death or survival of the parasite.
A comparative study of the antibody titer against Francisella tularensis biovar palaearctica in serum and lung extract was made using different immunoassays. Samples were taken from experimentally-infected goshawks (Accipiter gentilis) and European beavers (Castor fiber), and in a field survey of wild beavers. Good accordance between the antibody titer in serum and in lung extract was found in the experimental studies. However, the antibody titer was generally one- to three-fold lower in lung extract than was the titer of serum. Results in the field survey were less reliable. Estimating antibody titer in lung extract is a practical method for surveys for antibody levels, and an alternative to serological surveys, despite lower sensitivity as compared to serum.
The occurrence of tularemia was studied in 1,500 hares submitted to the National Veterinary Institute, Uppsala, Sweden for postmortem examination during 1973 through 1985. A total of 109 tularemia cases was recorded based on the fluorescent antibody (FA) test for Francisella tularensis and on the gross and microscopic pathology. Tularemia was diagnosed only in the varying hare (Lepus timidus) and not in the European brown hare (Lepus europaeus). The geographical distribution of the 109 cases indicates that tularemia has not spread in Sweden during the last 45 yr, with the exception of an endemic occurrence of the disease on the island of Stora Karlsö in the Baltic sea. The disease was most frequent in the autumn and only a few cases were recorded during winter. Cases were not seen in the spring. The annual prevalence varied, with several cases in 1974 and 1981, but there were no cases in 1976 and 1980. The postmortem findings in hares dying of tularemia in the autumn were characterized by focal coagulative necrosis in liver, spleen and bone marrow, with high numbers of bacteria FA-positive for F. tularensis. In hares dying during winter months, the most characteristic findings were hemorrhagic enteritis and typhlitis, although necrotic lesions could occur in liver, spleen and bone marrow. Diseased hares on the island of Stora Karlsö were demonstrated to be infected with ticks, while hares on the mainland of Sweden generally were fed upon by mosquitoes. Twenty-six of the 109 hares with tularemia were examined bacteriologically and F. tularensis biovar palaearctica was isolated from eight. The lung extract antibody test for F. tularensis was performed in 18 of the 109 hares. All were negative. In addition to the field study, an experimental study with F. tularensis biovar palaearctica was performed. Four varying hares and three European brown hares were inoculated. None of the hares died from tularemia, and generalized infection was not demonstrated.
Explore the source record for details and available documents.
Membranes from a capsule-deficient mutant of the live-vaccine strain of Francisella tularensis (F. tularensis LVS) were treated with N-lauroyl sarcosinate (Sarkosyl; CIBA-GEIGY Corp., Summit, N.J.). When the Sarkosyl-insoluble fraction was heated in the presence of sodium dodecyl sulfate and mercaptoethanol and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, several polypeptides were distinguished. Four major polypeptides were eluted from the gel, each of which stimulated lymphocytes from tularemia-vaccinated individuals but not from nonvaccinated individuals. The stimulation occurred mainly in T lymphocytes. Radioactive labeling of surface proteins of the capsule-deficient bacteria indicated that at least two of the four polypeptides originated from outer membrane proteins. The results suggest that several membrane proteins of F. tularensis LVS induce a specific T-lymphocyte response.
The lymphocyte stimulation test was used to estimate specific cell-mediated immunity after vaccination with the live vaccine strain of Francisella tularensis. Nonvaccinated individuals and individuals vaccinated 1, 5 to 6, 7 to 8, and 9 years previously were tested. Lymphocytes from most vaccinees responded to an antigen preparation of the vaccine strain, and those vaccinated 9 years before testing responded to a similar extent as did vaccines in the other groups. A new technique was developed to study the presence of T lymphocytes among the stimulated cells. Stimulated cells were allowed to incorporate [14C]thymidine and were then fractionated into T and non-T lymphocytes. Most of the incorporation appeared in the fraction containing T lymphocytes. The data indicate that cell-mediated immunity endures for at least 9 years after vaccination with the live F. tularensis vaccine.
After tularemia vaccinations, most individuals respond with cell-mediated and humoral immunity as disclosed by the lymphocyte stimulation test and enzyme-linked immunosorbent assay (ELISA), respectively. There is, however, no correlation between the magnitudes of the two responses, and some individuals show one of the responses only. We now report that the two responses are directed towards different antigenic determinants of the bacterium. Ether-water extraction of the live vaccine strain of Francisella tularensis gave a high yield of material reacting in ELISA as well as in the lymphocyte stimulation test. However, the specificity of the extract was low insofar as it reacted not only with lymphocytes and antibodies of tularemia-vaccinated individuals but also to a fairly high extent with those of nonvaccinated individuals. By using the extract as a starting material, a preparatory procedure was developed, resulting in antigen of high specificity in the two tests. The antigen prepared was a high-molecular-weight, carbohydrate-protein complex. Proteinase K treatment of the antigen abolished the lymphocyte-stimulating activity but did not decrease ELISA activity at all. Periodate treatment, on the other hand, greatly reduced ELISA activity but did not decrease the lymphocyte-stimulating activity. Thus, determinants of F. tularensis responsible for immunospecific lymphocyte stimulation seem to reside in protein, whereas ELISA activity seems to be due mostly to carbohydrate determinants.
The antigenic composition of the live vaccine strain of Francisella tularensis was investigated. Ether-water extracts, water-soluble material from freeze-pressed bacteria and detergent-eluted material from bacterial envelope were allowed to react in immunodiffusion and immunoelectrophoresis with rabbit antiserum against disintegrated bacteria of the vaccine strain. Ten antigenic factors were distinguished in an ether extract. When the extract was precipitated with ammonium sulphate 15 antigenic factors were distinguished in the precipitate and 14 factors in an ethanol precipitate of the supernatant fluid. In the water extract of freeze-pressed material 17 antigenic factors were found. Comparative immunoelectrophoresis of all these fractions demonstrated a minimum of 20 antigenic factors. When envelope material of the vaccine strain was treated with Triton X-100, three more antigenic factors were found to be solubilized. Thus, a total of 23 antigenic factors were distinguished in the extracts. There was a wide variation in heat and trypsin sensitivity between the antigenic factors. A few of the factors had esterase or alkaline phosphatase activity, whereas acid phosphatase, beta-glucuronidase or peroxidase activities were not found in any of the factors.