INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Within the genus Francisella there are two known species, Francisella tularensis and Francisella philomiragia, which have a 16S rRNA gene sequence similarity of more than 98% (6). F. tularensis, a gram-negative, small (0.2 to 0.7 by 0.2 µm) facultative intracellular bacterium, is distributed on all continents of the northern hemisphere. At least four subspecies of F. tularensis are known. Francisella tularensis subsp. tularensis (or type A; predominantly found in North America), Francisella tularensis subsp. holarctica, and Francisella tularensis subsp. mediaasiatica (both referred to as type B) cause the zoonotic disease tularemia in humans and animals (5, 11, 17). Only a few cases of human tularemia-like disease caused by the fourth subspecies, Francisella tularensis subsp. novicida, have been described (10). Depending on the site of entry and the predominant infectious or pathogenic process, tularemia can clinically manifest as an ulceroglandular, glandular, pulmonary, typhoidal, or ocular form. High fever and enlarged lymph nodes are predominant symptoms.

More than 250 different animal species can be infected by F. tularensis. Small rodents are the main natural hosts (reservoir), and blood-sucking ectoparasites are the most important vectors. In addition, the bacteria are quite stable in the environment under humidified and cold conditions. Humans can acquire the infection through the bites of infected arthropods or after contact with infected animals or contaminated water, food, dust, and aerosols (15, 17).

The detection of specific antibodies in serum is the most widely used approach for routine laboratory diagnosis of tularemia. However, antibodies to F. tularensis are not detected until 2 weeks or more after infection (2, 12). An early identification of the pathogen is important, as the course and the outcome of the disease is mainly dependent on early and adequate antibiotic treatment. The classical methods for identification of F. tularensis are the isolation of the pathogen by cultivation with subsequent identification by agglutination or immunofluorescence assay. These methods are time-consuming and require intensive handling of the infectious agent. Antigen detection in urine and RNA hybridization of wound specimens have also been used as diagnostic methods (7, 16). More recently, PCR has been applied for the detection of F. tularensis (6, 13, 18).

In this study we describe a highly sensitive capture enzyme-linked immunosorbent assay (cELISA) based on monoclonal antibodies (MAbs) and compare it with an immunochromatographic membrane-based handheld assay (HHA) and a PCR for detection of F. tularensis in naturally infected hare tissues.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Bacteria. The F. tularensis live vaccine strain (LVS; ATCC 29684) was cultivated on cysteine-heart agar (Difco, Augsburg, Germany) supplemented with 10% sheep blood. The bacterial isolates from European brown hares (Lepus europaeus) were obtained by direct cultivation on cysteine-heart agar as described previously (9). The bacteria were identified as F. tularensis by specific antisera. The classification of these isolates as F. tularensis subsp. holarctica was done on the basis of resistance to erythromycin and acid formation from glucose and maltose, as well as by the absence of utilization of glycerol and saccharose (4, 14). The sources of other bacteria used in this study are listed in Table 1. All bacterial isolates used were inactivated by 0.5 to 1.0% formalin at room temperature for 20 min, washed in phosphate-buffered saline (PBS; pH 7.4) and adjusted to an optical density at 560 nm (OD₅₆₀) of 1.0. In the case of F. tularensis, this OD corresponded to approximately 10⁹ CFU/ml (data not shown).

Tissue samples. Spleen, lung, and kidney specimens from European brown hares found dead or ill during an F. tularensis epizootic period of tularemia in northeast Austria during the spring of 1997 were investigated. The primary identification of the bacteria was performed by direct cultivation and indirect immunofluorescence assays, as well as by a slide agglutination assay with a specific polyclonal antiserum as described elsewhere (9).

Extraction of LPS of F. tularensis. The lipopolysaccharide (LPS) antigen was solubilized by incubating isolated bacteria (Table 1) or 0.5 ml of cut tissue samples in extraction buffer (Abbott, Solna, Sweden) at a ratio of 1:5 and boiling them for 15 min. Insoluble material was removed either by passing the suspension through 0.2-µm-pore-size filters or by centrifugation at 1,000 × g for 10 min. To create a standard curve, LPS was extracted from 10⁹ bacteria (LVS) and serial dilutions were prepared corresponding to an LPS content of 10⁷ to 10³ bacteria/ml.

Antibodies. The MAbs directed against F. tularensis had been developed as described previously (8). Selected MAbs were produced in serum-free medium at concentrations of approximately 500 to 1,000 µg/ml with a Miniperm bioreactor (Heraeus, Hanau, Germany). The MAb Ft-11 of immunoglobulin G1 (IgG1) isotype was purified by protein A affinity chromatography by a standard protocol and biotinylated with the Biotin Labeling kit (Boehringer, Mannheim, Germany). The MAb Ft-27 of IgM isotype was concentrated to 2 mg/ml by stirring it in a concentrator chamber containing a 100-kDa cutoff filter membrane (Amicon, Gloucestershire, United Kingdom) and was extensively dialyzed against PBS.

Western blotting. Inactivated LVS samples were adjusted to an OD₅₆₀ of 2.5. The samples were mixed at a 1:2 ratio with electrophoresis sample buffer (Novex, Frankfurt, Germany) and 5% mercaptoethanol. After 15 min of boiling, the solution was centrifuged for 20 min at 400 × g and submitted to sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis with a 4 to 20% separating Tris-glycine gel (Novex). The gel was electrotransferred onto a nitrocellulose membrane (0.45 µm) at 30 V for 1 h (Novex). The membrane was blocked with 5% skim milk in Tris-buffered saline (pH 8.1) overnight at 4°C. MAbs were incubated with membrane strips at 5 µg/ml in PBS-10% goat serum at room temperature for 2 h. For specificity control, the MAbs were preincubated with PBS or extracted LPS from various numbers of F. tularensis LVS overnight at room temperature. After three washes, the membranes were incubated with goat anti-mouse immunoglobulin-peroxidase (Dako, Glostrup, Denmark) or goat anti-human immunoglobulin-peroxidase (Sigma, St. Louis, Mo.) at room temperature for 1 h followed by substrate reaction with precipitating tetramethylbenzidine (Seramun, Dolgenbrodt, Germany).

cELISA. For cELISA, a 96-well microplate (Maxisorb; Nunc, Roskilde, Denmark) was coated with 100 µl of the MAb Ft-27 at 10 µg/ml in carbonate buffer (pH 9.0) overnight. The plate was washed and blocked at 37°C for 30 min with PBS containing 0.05% Tween 20 and 4% skim milk powder. Antigen was diluted in dilution buffer (PBS with 0.05% Tween 20 and 2% skim milk) and incubated at 37°C for 60 min. After being washed, the bound antigen was incubated with 100 µl of biotinylated Ft-11 per well at 2 µg/ml at 37°C for 60 min. After being washed, 100 µl of streptavidin-peroxidase (Dako) diluted in dilution buffer at 1:4,000 was added and incubated at 37°C for another 60 min. The plate was washed, and 100 µl of tetramethylbenzine substrate (Seramun) was added and incubated at 37°C for 10 min. The OD was read with a photometer (Digiscan; AsysHitech, Eugendorf, Austria) at wavelengths of 450 and 620 nm as references.

Immunochromatographic HHA. The HHAs and the antibodies used to manufacture them were produced at the Naval Medical Research Institute. This assay employs an affinity-purified polyclonal antibody and a MAb with an isotype of IgG₃. The specificity of the MAb is for LPS of F. tularensis LVS.

Preparation of DNA and PCR. In order to isolate DNA, two different protocols for treatment of tissue and spiked serum samples prior to PCR were assessed. For method A, a 400-µl aliquot of ground tissue slurry or of spiked serum was mixed with 50 µl of TEN buffer (100 mM Tris, 10 mM EDTA, 1 M NaCl, pH 8.0), 20 µl of SDS (20% [wt/vol]), and 50 µl of proteinase K (1 mg/ml) and incubated at 56°C for 3 to 5 h. After two phenol extractions followed by two extractions with chloroform-isoamyl alcohol (24:1), the DNA was ethanol precipitated. The pellet was washed twice with 70% ethanol and resuspended in 20 µl of water. Method B was applied for spiked serum samples only. Lysis buffer (50 mM Tris, 1 mM EDTA, 0.5% Tween 20 [pH 8.0]; 190 µl) and 10 µl of proteinase K (1 mg/ml) were added to 200 µl of spiked serum. Incubation at 56°C for 3 h was followed by 10 min at 95°C to inactivate the proteinase K. The PCR was performed as described by Sjöstedt (18) with 1 µl of DNA template prepared by either method A or B and 40 cycles of amplification with primers specific for sequences of the gene encoding a 17-kDa lipoprotein, TUL 4 (13). After amplification, 5 µl of each reaction mixture was analyzed on a 2% agarose gel.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Development and evaluation of the cELISA. Two specific murine MAbs to F. tularensis (8), Ft-11 (IgG1) and Ft-27 (IgM), were utilized for the cELISA. Both MAbs reacted strongly with the LPS of F. tularensis as shown by Western blot analysis (Fig. 1). The reactivity of the MAbs could be blocked by free LPS of F. tularensis in a concentration-dependent fashion, indicating the specificity of the binding. A key step in the configuration of the assay was the extraction of the LPS from the bacteria. Compared to the crude bacterial preparation, a 10-fold increase in sensitivity was achieved with cELISA (Fig. 2). Furthermore, the handling of infectious material was reduced by this first step of the assay, as the LPS extraction was accompanied by inactivation of the bacteria. This procedure was the method of choice, as it also allowed the lysis of eukaryotic cells, which is important for the availability of intracellularly localized bacteria.

View larger version (56K): [in this window] [in a new window]   FIG. 1.   Western blot of LPS-specific MAbs to F. tularensis. Whole bacterial antigen of F. tularensis was separated by SDS-polyacrylamide gel electrophoresis with 4 to 20% Tris-glycine gel and blotted onto a nitrocellulose membrane. The MAbs Ft-27 and Ft-11 were used at 5 µg/ml in PBS (A). For specificity control, the MAbs were preincubated with extracted LPS from 104 (B), 106 (C), and 108 (D) bacteria/ml (lane E, molecular mass markers).

View larger version (21K): [in this window] [in a new window]   FIG. 2.   Comparison of detection of whole bacteria and extracted LPS of F. tularensis by cELISA. Serial dilutions of whole bacteria and extracted LPS antigen of F. tularensis LVS from spiked PBS and pooled human sera were tested in the LPS-specific cELISA. The results are given as OD450.

Detection limits of cELISA, HHA, and PCR. In order to compare the sensitivities of the different assays used in this study, serial dilutions of F. tularensis LVS were prepared in PBS and in human sera from healthy donors, representing biological fluids which could contain inhibiting factors for the assays. The detection limit of the cELISA was 10³ bacteria/ml in PBS and 10⁴ bacteria/ml in human serum (Fig. 3). This represents detecting 10² and 10³ bacteria/100 µl of sample. Further evaluation of the assay has revealed that the bacteria and the LPS antigen, respectively, could be detected in spiked human urine, sputum, and stool with nearly the same sensitivity (data not shown).

View larger version (26K): [in this window] [in a new window]   FIG. 3.   Sensitivity of cELISA, HHA, and PCR for the detection of F. tularensis. Serial dilutions of the solubilized LPS antigen of F. tularensis in PBS and in pooled human normal sera (NS-P) were tested in the specific cELISA (upper diagram). The results are given as OD450. The same dilutions of bacteria were tested in the immunochromatographic HHA and in the PCR (lower table). The results are expressed as relative intensity of obtained visible bands. +, light; ++, medium; +++, dark; neg, negative; n.t., not tested.

Detection of F. tularensis in tissue specimens. For further evaluation, the assays were applied to detect and quantify F. tularensis in tissue samples from dead or sick and killed European brown hares (Table 2). The bacteria were isolated by direct cultivation or were detected by immunofluorescence from 14 samples from hares (no. 3 to 16) and were classified as F. tularensis subsp. holarctica. In one case (no. 17) tularemia was diagnosed serologically by a slide agglutination assay. In two cases (no. 1 and 2) the bacteriological and serological studies remained negative. Eleven of the animals (no. 3 to 13) showed pathomorphologic signs of an acute septicemic tularemia with enlarged spleen. Pneumonia, pericarditis, or nephritis was observed in four animals (no. 14 to 17) and was considered to represent chronic forms of the disease. In 15 cases, spleen tissues were studied; in two cases, kidney or lung, respectively, was available.