Development and Evaluation of a Latex Agglutination Test for the Identification of Francisella tularensis Subspecies Pathogenic for Human

Tularemia is a zoonotic, highly infectious disease caused by an intracellular Gram-negative bacterium, Francisella tularensis. The disease affects a wide range of hosts including invertebrates, mammals and birds. Humans can become infected by direct contact with an infected animal (through broken skin, scratch or tissue injury), through a bite of haematophagous arthropods (e.g. fleas, lice, midges, bedbugs, mosquitoes, ticks), by drinking contaminated water, eating contaminated food, or through inhalation of contaminated dust (Formińska et al., 2015). The clinical presentation in humans depends on the route of infection and varies in symptoms and severity (Eliasson et al., 2006). There are four subspecies of F. tularensis: tularensis, novicida, mediasiatica and holarctica. Of these, subsp. tularensis and subsp. holarctica cause disease in humans, whereas subsp. mediasiatica is believed to be of relatively low virulence in humans, and only rare cases of human disease caused by this subspecies are known. F. tularensis subsp. novicida is non-pathogenic for humans (Celli and Zahrt, 2013).

According to WHO case definition, a confirmation of tularemia case requires recovery of an isolate and identification of the culture as F. tularensis by antigen or DNA detection. Commercial biochemical identification systems available in clinical diagnostic laboratories are not suitable for accurate identification of F. tularensis. Alternatively, paired serum specimens with a fourfold difference in titre (tube or microagglutination assay) or significantly (ELISA), with at least one serum positive, are also considered confirmatory (WHO, 2007). However, antibody against F. tularensis are detectable in patients’ serum 10–20 days post-infection (Koskela and Salminen, 1985). Thus, usefulness of antibody detection tests is limited in severe cases and when a rapid preventive action must be undertaken. On the other hand, a F. tularensis antiserum (Becton Dickinson Diagnostic Systems) recommended by WHO for the slide agglutination test for F. tularensis culture identification has been withdrawn from the manufacturer’s offer and is not available on the market any more. Below we described the latex agglutination test (LAT) for the rapid identification of F. tularensis isolates that could be an alternative for the classical slide agglutination test.

For the preparation of sera for coating of latex beads we used pooled serum samples obtained from 25 patients with high level of IgM antibodies to F. tularensis, specified as OD₄₅₀ higher than 1,80 by ELISA (Rastawicki and Wolaniuk, 2013; Rastawicki et al., 2015). The gamma globulin fractions of sera were isolated by 40% ammonium sulfate fractionation in cold-bath, and the precipitate was resuspended in phosphate-buffered saline (PBS), pH 7.4. The solution was subsequently dialyzed against PBS for 48 hours until ammonium sulfate had been removed. Then, 5.0 ml of washed, 1% suspension of 0,81 μm latex particles in glycine-saline buffer (pH 8.2) was added to equal volume of twice-diluted in glycine-saline buffer purified gamma globulins. The mixture was vortexed for 1 min and then allowed to incubate with gentle stirring at 37°C for 6 h. After incubation, sensitized latex particles were recovered by centrifugation, washed twice with glycine-buffered saline and finally diluted to 1% with glycine-buffered saline (pH 8.2) containing 0.1% sodium azide and 0.3% of BSA. For control, the latex reagent particles were sensitized with bovine albumin (Sigma Chemical Co., USA).

The investigated strains of F. tularensis (Table I) were cultured on enriched chocolate agar plates. After 48 h of incubation, a small loopful of bacteria from the strain investigated was suspended in 100 μl of PBS. The latex particles sensitized with gamma globulins and the particles sensitized with albumin bovine-control latex reagent, were parallel mixed with the same volume (25 μl) of bacterial suspensions on a black glass plate. The results were read after 1, 3 and 5 minutes of rocking the plate. Agglutination (clumping of cells) was scored as: -negative; +/++/+++ weak/strong/very strong positive. A positive latex agglutination test and a negative latex control test were confirmatory for F. tularensis. To assess potential cross-reactivity of the developed latex test we used the 16 different control bacterial strains as presented in Table I. The procedures of LAT with control strains were the same as for F. tularensis.

All manipulations with viable F. tularensis strains were done under biosafety level 3 (BSL-3) conditions. Keeping in mind that most diagnostic laboratories work under BSL-2 conditions, to minimize the risk of infection we also evaluated the test using inactivated bacterial suspension. For inactivation the bacterial suspension in PBS was heated at 96°C for 15 min, then cooled and used for the LAT. To verify the effectiveness of inactivation, 50 μl of the suspension were inoculated onto enriched chocolate agar plates and incubated at 37°C for 10 days.

The agglutination reactions with F. tularensis subsp. holarctica and F. tularensis subsp. tularensis strains after 3 minutes were very strong without any differences between live and inactivated suspensions used. We did not observe positive reactions for F. tularensis with the control latex reagent. No positive reactions were observed also by the LAT with F. tularensis subsp. novicida as well as with the most of control strains. However, a very strong reaction of LAT with S. aureus ATCC 25923 was found after 3 minutes of rocking the plates. For this reason, we decided to investigate the additional five S. aureus strains isolated from hospital patients. A weak positive reaction after 3 minutes and a strong reaction after 5 minutes of rocking were observed in two cases.

Antibody coated latex particles are commonly used in diagnostic microbiology for detection, identification or serotyping of many different microbes (Miller et al., 2008; Porter et al., 2008; Sumithra et al., 2013). In the previous work we developed the LAT for detection of antibodies against F. tularensis in serum samples (Rastawicki et al., 2015). Here, we present the LAT for identification of F. tularensis that could be cultured from all kinds of samples like environmental, food, human and animal tissue samples, etc. In accordance with our expectation, the test recognized F. tularensis subsp. tularensis and F. tularensis subsp. holarctica but not F. tularensis subsp. novicida. It is because of the unique LPS composition of both subsp. tularensis and subsp. holarctica, which is different from that of F. tularensis subsp. novicida (McLendon et al., 2006). The LAT recognises clinically relevant subspecies of F. tularensis, opposite to PCR detection of tul4 gene, which gives positive results for all Francisella species and additional PCRs for other targets are necessary to differentiate Francisella species and subspecies (WHO, 2007). The lack of cross-reactivity of the LAT with other bacteria, except S. aureus, is in accordance with other researchers’ results on antibody-based F. tularensis identification methods such as cELISA and immunochromatographic assay (Grunow et al., 2000), and reveals that the test is highly specific. The cross-reactivity with some S. aureus strains is probably related to the presence of protein A in the cell wall of these S. aureus strains. It has been demonstrated that protein A expressed by some S. aureus strains has a high ability to bind immunoglobulins (King and Wilkinson, 1981; Romagnani et al., 1981). Our experiments with other latex tests of different commercial companies, for example dedicated to detection of Salmonella or E. coli strains, also showed cross-reactivity with some S. aureus (data not shown). However, it is quite easy to differentiate between F. tularensis and S. aureus based on commercially available latex agglutination test for S. aureus or ability to grow on various microbiological media, the colonial morphology, or Gram staining. F. tularensis grows on rich media (enriched chocolate agar – CA, buffered charcoal yeast extract – BCYE, cystine heart agar with 9% chocolatized blood – CHAB, thioglycollate-glucose blood agar – TGBA, GC Agar II with 1% haemoglobin and 1% IsoVitaleX, sheep blood agar – SA) but does not grow on ordinary media; whereas, S. aureus easily grows on ordinary media such as nutrient agar (NA) and brain heart infusion agar (BHI). Also, some selective media can be applied.

The LAT developed in our study is inexpensive, simple, rapid and does not need any specialized equipment to be performed. We recommend that the results should be read after 3 minutes of rocking the plate. Thus, the test can be performed much faster compared to PCR or real-time PCR which needs at least one hour, even when the fast polymerases are used (Zasada et al., 2013). Moreover, the test works well with inactivated samples which minimizes the risk of laboratory acquired infection and allows to perform the test under BSL-2 conditions. It is the important characteristic as majority of diagnostic laboratories work under BSL-2 conditions with no access to a BSL-3 laboratory. The LAT reagents shelf life is at least 2 years when stored at 4°C, as it was shown by manufacturers of commercially available latex tests for other microbes as well as our experience with in-house tests. The method is also highly reproducible between different operators (data not shown). The LAT could be an alternative for the slide agglutination test described in WHO guidelines for tularemia (WHO, 2007) when the Francisella tularensis antiserum is unavailable. Moreover, the use of latex particles coated with antibodies increases sensitivity of antigen detection significantly when compared to the slide agglutination test with antiserum (Drożdż, 2006).