Glanders in horses in some selected areas of Bangladesh and comparison between CFT and Immunoblot used for the screening of glanders

Glanders is an infectious disease of solipeds and humans caused by the gram negative bacterium Burkholderia mallei. It is the oldest known diseases of horses (Marr and Malloy, 1996). The organism was isolated first by Loeffler and Schultz in Germany in 1882 and then by the French microbiologists Bouchard, Charrin and Captain in the same year (Marr and Malloy, 1996). The disease is characterized by chronic suppurative lesions of the skin and mucous membranes, pneumonia and septicemia. In horses, the disease is mainly chronic in nature and causes severe economic losses in equine related trade. The natural hosts of B. mallei are the horse, donkeys and mules (Khan et al. 2013). The disease has public health importance and humans become infected through direct contact with organisms and longtime contact with infected animals. An extremely high rate of mortality can occur in untreated cases.
        
Agglutination test and complement fixation test (CFT) are still the mostly used serological tests for the diagnosis of B. mallei infection or proof of the presence of a specific delayed hypersensitivity reaction after application of mallein intra cutaneously is also used (Neubauer et al., 2005; OIE, 2008). The CFT is the only officially recognized serological test for glanders in international trade of equidae. The CFT has a sensitivity of at least 97% (Cravitz and Miller, 1950), but a notable number of nonspecific, false positive results occur (Wilson and Miles, 1946; Misra and Arora, 1988; Turnbull et al., 2002). False positive results due to cross reactions may be seen in horses suffering from strangles, equine influenza or petechial fever. The test can also not  be applied on sera having so called “anticomple-mentary activity”. In general, serological tests may be negative in emaciated and chronically debilitated animals suffering from glanders (Wilson and Miles, 1946). In glanders free areas or in areas with very low prevalence of glanders, highly specific tests are needed to minimize the number of false positive results (Sprague et al., 2009). In Western blot analysis or competition ELISA, lipopolysaccharide (LPS) preparations have already been used to detect anti- B. mallei antibodies in serum of horses (Katz et al., 1999; Sprague et al., 2009). But current protocols for the extraction and purification of LPS from B. mallei are time consuming and hazardous for the operator (Pitt et al., 1992; Anuntagool and Sirisinha, 2002). However, it is necessary to find out an appropriate test that could be performed easily.
        
Glanders is a widely spread zoonotic disease and according to the Office International des Epizooties (OIE), it is a notifiable disease in 178 countries. The disease is still endemic in the Middle East, Asia and South America. From 2000 to 2014, equine glanders was reported from 12 countries in Asia, two countries in Europe (Lithuania, Russia) and two each from South America (Brazil and Bolivia) and Africa (Eritrea and Ethiopia) (Bazargani et al., 1996; Al-Ani et al., 1998; Mota et al., 2000; Muhammad et al., 2001; Elschner et al., 2009; Elschner et al., 2011).
        
Bangladesh is an agro-based country where livestock is playing an important role to uplift the socio-economic status of the rural farmers. Among livestock, the horse is widely used for drought purposes such as pulling carts, transportation, land tillage and racing. Bangladesh is always in danger of this disease as the neighboring countries like Pakistan and India are affected by this disease. However, in spite of its importance, so far no study has been carried out on glanders in the horses of Bangladesh. Therefore, the present study was conducted to determine the seroprevalence of glanders in horses using CFT and the positive sera were further confirmed by immunoblot.

Study area and collection of samples
 
The study was carried out at the Zoonotic Disease Diagnostic Laboratory, Department of Medicine, Faculty of Veterinary Science, Bangladesh Agricultural University, Mymensingh and in the OIE and National Reference Laboratory for Glanders, Friedrich-Loeffler-Institut, Federal Research Institute for Animal Health, Institute of Bacterial Infections and Zoonoses, NaumburgerStrasse 96a, 07743 Jena, Germany.
        
Approximately 5-7 ml of blood was randomly collected from 301 horses from three districts: Mymensingh (115), Tangail (76) and Jamalpur (110) along with clinical signs using a sterile disposable syringe. Samples were placed on a tray undisturbed for at least 1 hr at room temperature in a slightly inclined position to facilitate clotting and separation of serum. The clotted blood samples were then stored overnight at 4°C. Sera were transferred to a fresh test tube and centrifuged at 2,500 rpm for 10 minutes. The clear supernatants were transferred to fresh tubes and stored at -20°C until further use.
 
Complement Fixation Test and Immunoblot
 
The CFT on 301 serum samples was performed according to the instructions of the OIE Manual of Diagnostic tests and Vaccines for Terrestrial Animals (OIE, 2008) using a certified, commercially available antigen (cc-pro GmbH, Oberdorla, Germany), a complement and ready-to use hemolytic system (InstitutVirion/Serion GmbH, Würzburg, Germany) (Khan et al. 2014). Samples were considered negative when 100% haemolysis occurred at 1:5 dilution, suspicious when 25-75% haemolysis was seen at dilution 1:5 and positive when no haemolysis was detected at dilution 1:5.
        
All samples showing positive or suspicious results in CFT were tested by immunoblot as described previously (Elschner et al., 2011). Briefly, after transfer of LPS-containing B. mallei antigen on nitrocellulose membrane, the membrane was blocked overnight in the blocking solution (Candor Bioscience, Wangen, Germany). After washing and cutting, strips were incubated with equine serum, for 1.5 hour diluted 1:50 in low cross buffer (Cando Bioscience) at room temperature. After incubation, membrane was washed for three times for 20 minutes each. Then it was incubated for 1.5 hour at room temperature in low cross buffer containing alkaline phosphate-conjugated rabbit anti-horse-IgG (A6063, Merck, Darmstadt, Germany). The strips were washed for three times again. Finally, it was stained with NBT-BCIP solution (Merck). In case of positive B. mallei, the banding pattern of LPS ladder within the region of 20 to 60 kDa should be clearly visible, weak color represented in suspicious case and negative case should be colorless.
 
Statistical analysis
 
Descriptive statistics, 95% confidence interval of prevalence and Pearson chi-square test to determine the level of significance between glanders detection level among CFT and immunoblot positive horse serum were performed in SPSS version 20.0 (Statistical Package for Social Services).

The overall seroprevalence of glanders was 34.9% (n = 105) using CFT and 24.8% (n = 26) of positive CFT samples were also confirmed using immunoblot (Table 1). One sample showed anticomplementary activity. One CFT positive sample was suspicious in immunoblot (Fig 1). There were also 21 suspicious cases in CFT and 3 suspicious cases in immunoblot. Out of 115, 110 and 76 sera from Mymensingh, Jamalpur and Tangail districts, 43, 42 and 20 sera showed positive reaction to CFT, respectively. The CFT based prevalence of glanders was found to be 37.4% (95% Confidence Interval (CI): 28.69-46.94) in Mymensingh, 38.2% (95% CI: 29.23-47.97) in Jamalpur and 26.3% (95% CI: 17.18-37.87) in Tangail district (Table 1). The highest immunoblot based prevalence of glanders was recorded for Jamalpur district (30.9%, 95% CI: 18.1-47.2), while 25.6 % (95% CI: 14.0-41.5) for Mymensingh and 10.0% (95% CI: 1.8-33.1) for Tangail districts were found. The difference in detection level of glanders in immunoblot test from these three districts was not statistically significant (p = 0.194) (Table 1). With respect to clinical signs, among the 105 CFT positive horses, 29 horses were affected by parasites, 28 had anorexia and fever, 8 showed signs of skin diseases, 7 had lameness and arthritis, 9 were affected by diarrhea andcolic,1 had influenza, 6 had lacrimation, 2 had abortions, 1 had wounds, 1 had epistaxis and 13 animals were apparently healthy (Table 2).
 

 

 

        
Among the 26 immunoblot positive cases, 6 horses were affected by parasites, 6 had anorexia and fever, 2 showed signs of skin diseases, had lameness and arthritis,4 were affected by diarrhea and colic, 3 had lacrimation and 3 animals were apparently healthy (Table 3). These clinical signs are supportive to glanders infection.
 
  
 
There are many tests which have been used in the diagnosis of glanders such as fluorescent antibody test, indirect hemagglutination test, immune electrophoresis which have low sensitivity and specificity (Zhang and Lu, 1983; Ma, 1986; Misra and Arora, 1988; Katz et al., 1999; Naureen et al., 2007). Complement Fixation Test is the test used worldwide for serodiagnosis of glanders because of its high sensitivity. But, CFT can give false positive results thus confirmation is needed. The cELISA is considered to be a reliable test due to its sensitivity and specificity of almost 100%. Polysaccharide based microarray is a new test for the serodiagnosis of glanders and melioidosis. But, both tests are too expensive for routine diagnosis of glanders in developing countries. The Westernblot is a very time-consuming, expensive and difficult to standardize test. It’s use is limited to few reference laboratories till now. Considering the costs, the effective routine complement fixation test and the confirmatory immunoblot technique were used in this study. An unexpected high prevalence of glanders in horses of 34.9% was observed with the CFT and still 24.8% proved immunoblot positive/suspicious. It is of utmost importance now to start cultivation of samples of horses which have been tested positive by immunoblot to unravel the true prevalence of glanders as it can be supposed that various positives may have been caused by infection with B. pseudomallei which is endemic in soils of South East Asia. For this setting tests distinguishing both diseases are needed and independent control programs for both diseases should be implemented taking into account the very special economic situation of Bangladesh. 
 
It is concluded that CFT is still the most valuable screening test currently available, but there are many false positive results found in the Bangladeshi setting where immunoblot has showed that it can help to reduce the number of false positive results made by CFT.

Authors are thankful to DAAD (Deutscher Akademischer Austauschdienst-German Academic Exchange Service), Bonn, Germany for financial support to the first author and Friedrich-Loeffler-Institut, Federal Research Institute for Animal Health, Institute of Bacterial Infections and Zoonoses, Naumburger Strasse 96a, 07743 Jena, Germany to take the sera samples from Bangladesh to Germany and then technical supports to examine the sera.