Characterization and Diagnostic Application of the Francisella Tularensis Capsule

NON-TECHNICAL SUMMARY: Tularemia is primarily a disease of wildlife, but can also be transmitted to food animals when bitted by an infected arthrood vector. Furthermore, tularemia is a potential bioterrorism and agroterrorism agent. At this time, there is no rapid diagnostic test that can be used in the field to diagnose F. tularensis in specimens. The purpose of this project is to develop rapid, sensitive and specific latex agglutination and biosensor diagnostic tests for tularemia that can be used in field conditions.

<P>

APPROACH: <BR> Objective 1: We will remove the cells by centrifugation, resuspend the bacteria in 1/10th volume 10 percent NaCl, and stir this mixture for 1 h at room temperature. The cells will be removed by centrifugation, and hexadecyl-trimethyl ammonium bromide at 0.05 M final concentration will be added. The precipitate will then be extracted with 0.4 M NaCl, and high molecular weight material precipitated with 3 volumes of -20 C ethanol. The precipitate will be resuspended in 0.1 M phosphate buffer, pH 7.0, and incubated with DNase, RNase, and proteinase K. This mixture will be extracted with an equal volume of 90 percent phenol at room temperature for 30 min. After the extraction the aqueous phase is separated by centrifugation, removed and extracted with fresh phenol two more times. The aquesous phase will be ultracentrifuged to remove LPS and lyopilized. <BR><BR>Objective 2: One or more rabbits will be immunized with 50 ug of purified capsule in Freunds Complete Adjuvant in several subcutaneous sites. A second immunization will be given about 3 weeks later in Freunds Incomplete Adjuvant. About 2 weeks later the rabbits will be immunized with 50 to 100 ug of capsule intravenously at weekly intervals until a high titer antiserum is obtained. Once monospecific antiserum to the capsule is made the IgG fraction will be isolated by Protein A/G affinity chromatography. This IgG fraction will be used to identify the location of the capsule on the cell with Protein A gold particles by immuno-electron microscopy. The bacteria will be grown, incubated with IgG and then Protein A-20 nm gold and electron microscopy carried out at the College of Veterinary Medicine's electron microscopy core laboratory. <BR><BR>Objective 3: The isotype of the antibodies made to the purified capsule will be determined by ELISA using commercial isotype-specific secondary antibody conjugates. monospecific IgG to the capsule will be isolated by Protein A/G affinity chromatography, and the IgG covalently coupled to carboxylate latex particles. The latex particles are activated by the addition of 2 percent 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide-HCl. After end-over-end mixing for 4 h at room temperature the particles are washed in borate buffer, pH 8.0, and various concentrations of IgG is added. Nonspecific binding sites are blocked after overnight mixing by the addition of 0.25 M ethanoloamine for 30 min. The particles are then washed and suspended at 1% w/vol in storage buffer. Variables that will be examined to optimize the sensitivity and specificity of the assay will include the size of the particles, the concentration of IgG, and the use of streptavidin-conjugated latex particles and biotin-hydrazide IgG, which may amplify the sensitivity of the assay. Once optimized, this test can detect as little as 250 pg of capsule, based on results with A. pleuropneumoniae. The antibodies will also be attached to photonic biosensors for an alternate diagnostic test.
<BR><BR>
PROGRESS: 2002/07 TO 2006/06<BR>
We have determined that the capsule of Francisella tularensis is atypical in that, unlike most bacterial capsules, it cannot be isolated from culture supernatant by precipitation with 75% ethanol or cetavlon. Scanning electron microscopy revealed that cells of F. tularensis type A and the LVS (live vaccine strain), both of which are reported to be encapsulated, appeared to have a small electron-dense surface material around them. For convenience, most work was therefore done with LVS, which can be worked with in a BSL-2 laboratory. The bacteria were passed in chemically defined broth medium for 17 passages, followed by culture on the same agar medium and incubated at 32 degrees C, rather than 37 degrees C. The electron dense material was greatly enlarged. However, it remained difficult to differentiate and separate the capsule from lipopolysaccharide (LPS). In other experiments, we generated a mutant with a base substitution in the gene wbtI, which is involved in LPS O-antigen biosynthesis. This mutant made a complete core carbohydrate, but no O-antigen. Complementation of this mutant with a normal copy of wbtI generated a recombinant strain with a normal LPS. The mutant was also passed in defined medium, and also made a large amount of electron-dense material. These passed mutant cells were gently extracted with 0.5% phenol, the cells removed, and the high molecular weight material precipitated by addition of excess cold ethanol. The precipitate was suspended in water, and subjected to ultracentrifugation at 41,000 rpm overnight. The supernatant was precipitated with cold ethanol, and the precipitate dialyzed, and fractionated through a Sephacryl-300 column. The carbohydrate-positive void volume was lyophilized. The pellet from the ultracentrifugation step was extracted with warm (50 degrees C) 1% sodium deoxycholate, the ultracentrifugation repeated, and the supernatant treated as above. The purified material was submitted to the Complex Carbohydrate Research Center at the University of Georgia for chemical analysis, and it was determined that the putative capsule contained glucose, galactose, mannose, and octadecanoic acid. The material was coupled to an amino-affinity column support by cyanogen bromide activation, and antibodies to the putative capsule from the immunoglobulin fraction of rabbit antiserum to LVS whole cells were affinity purified. The antibodies reacted with the antigen by ELISA, and Protein A-gold-immuno-transmission electron microscopy demonstrated that the antibodies bound to material on and just off the cell surface of LVS. This antigen had a ladder-like appearance (similar to LPS) on Western blot analysis when reacted with antiserum to whole cells, but not with antibodies to LPS.
<BR><BR>
IMPACT: 2002/07 TO 2006/06<BR>
At this time a vaccine is not licensed for the prevention of tularemia, and a rapid diagnostic test is not available. We expect that the capsule of F. tularensis is required for virulence, and that conjugation of the capsule to key antigens that induce the proper cellular immune response may be used as a protective vaccine. In addition, antibodies to the capsule may be used in various types of diagnostic tests for rapid diagnosis or for detection of the agent. We have isolated a novel glycolipid from F. tularensis that may prove to be an important factor in the virulence of this bacterium, and which can be used to design improved diagnostic tests and vaccines against tularemia.