Assessment of serological response of chickens to Salmonella Gallinarum and Salmonella Pullorum by Elisa

This study was done to assess an enzyme-linked immunosorbent assay (ELISA) to detect chicken serological response against Salmonella enterica serotypes Gallinarum and Pullorum. The assays have used soluble bacterial proteins of Salmonella Gallinarum strain 9 (AgSG) as detecting antigen and peroxidase and alkaline phosphatase conjugates. According to the results, the antigen, sera and conjugate concentrations were optimized. In addition, the assay using alkaline phosphatase and peroxidase conjugates was helpful to distinguish positive serological reaction to serotypes Gallinarum and Pullorum from Enteritidis.

Because high concentrations of circulating IgG persist for several months following infection with Salmonella (Hassan et al., 1990;Nicholas & Cullen, 1991), the ELISA can overcome the problem of infection detection due to the intermittent faecal excretion of Salmonella. IgM and IgA may also be detected, but such classes of antibody do not persist to the same extent as IgG (Hassan et al., 1990(Hassan et al., ,1991. Because of the persistence of IgG, ELISAs based on this class of antibody have presently a greatest practical potential to indicate the infection status of the poultry flock. The ELISA based on IgG detection has also some advantages over established serological tests. It has been shown to be more sensitive than both the rapid slide agglutination test, which relies on the use of undiluted serum, and standard tube agglutination; besides, it is at least as sensitive as the microantiglobulin test (Cooper et al., 1989;Nicholas & Cullen, 1991). Unlike the slide agglutination test, developed originally to be used for infections caused by serotypes Pullorum and Gallinarum, ELISAs can also be used to examine eggs, an important aspect to the poultry breeding stock, table egg and food industries (Barrow & Lovell, 1991;Nicholas et al., 1990). The ability of the ELISA to differentiate infections caused by Salmonella serotypes has been demonstrated by using several detecting antigens prepared from different serotypes, such as LPS (Cooper et al., 1989;Dadrast et al., 1990;Nicholas & Cullen, 1991) or flagella (Hassan et al., 1990). The ELISA can also be used with serum eluted from blood dried onto absorbent paper, facilitating field sampling (Hassan et al., 1990;Minga & Wray, 1990).
Recent studies have shown the value of ELISA in studying the serological responses of chickens to experimental and natural infection with invasive Salmonella including Typhimurium (Hassan et al., 1990), Enteritidis (Nicholas & Cullen, 1991), and Gallinarum-Pullorum serotypes (Barrow et al., 1992;Iba et al., 1991) and has also indicated their value in monitoring field infections such as those with S. Enteritidis (Barrow, 1994b) and S. Gallinarum and S. Pullorum (Berchieri Jr et al., 1995). Most of these assays have used soluble proteins, LPS or flagella as detecting antigens. Barrow et al. (1992) showed that either soluble protein or LPS could be used to study experimental infections. The soluble protein was also used successfully in a small field study in which high optical densities were obtained in ELISA evaluation of sera from showing signs of fowl typhoid, but not from uninfected flocks. Afterwards, a larger survey was carried out showing the reliability of the test (Berchieri Jr et al., 1995).
The most effective means of controlling pullorum disease and fowl typhoid is a combination of stringent management procedures and eradication (Snoeyenbos, 1991). Eradication is normally done by identifying infected flocks and eliminating individual reactor birds detected by a serological test. Serology is used because S. Gallinarum and S. Pullorum are not excreted extensively in the feces, unlike many other Salmonella serotypes that are more frequently associated with human food poisoning. For many years, the choice has been the slide agglutination test, which was originally developed by Runnells et al. (1927) to test serum and was later adapted by Schaffer et al. (1931) to test whole blood using a stained antigen. The test is easy to use and has largely been used as a flock test and helped to eliminate these infections in many countries. However, the test can yield erratic results, which can be dependent on antigen quality. Quantification is also difficult and the test is not applicable to be used with eggs; such a test would be useful for screening breeding flocks (Barrow et al., 1992).
The Brazilian poultry industry is based on the importation of grandparent flocks and the official program of the Government establishes that the flocks must be free of S. Gallinarum and S. Pullorum and controlled for S. Enteritidis and S. Typhimurium. In many countries, S. Enteritidis was disseminated on poultry farms through vertical transmission (Baumler et al., 2000). In Brazil, dissemination was related to Salmonella isolation from imported ostrich flocks (Ribeiro et al., 2003). Environmental conditions facilitated the spreading of these serotypes. In view of the difficulties in controlling dissemination of S. Enteritidis, many commercial flocks of laying hens are vaccinated against it and there is a strong tendency to extend the vaccination program to breeding flocks (Scharr, 2003). In this case, the traditional pullorum test using whole blood will lose be of no value. Studies from Iba et al. (1991) and Berchieri Jr. et al. (1995) showed that the ELISA using plates coated with soluble protein as antigen and alkaline phosphatase conjugate was able to detect serological response to S. Pullorum and S. Gallinarum. In addition, the test was supposed to separate positive results of S. Gallinarum and S. Pullorum from S. Enteritidis. The present work was undertaken to assess the ELISA as a tool to search serological response to S. Gallinarum and S. Pullorum, and also to separate positive reactions of infections caused by S. Gallinarum and S. Pullorum from S. Enteritidis. The ELISA was established using soluble proteins and either alkaline phosphatase or peroxidase conjugates.

Samples of serum Reference sera
Adult chickens with no signs of clinical salmonellosis were inoculated intra-muscularly with 0.5mL of inactivated cultures of Salmonella (10 8 cfu/mL) to obtain serological response to individual Salmonella serotypes. The Salmonella serotypes included Enteritidis, Typhimurium, Gallinarum, Pullorum, Infantis, Montevideo, Binza, Livingstone, Anatum, Stanley, Eimsbuettel, Ealing and Virchow. The 10mL broth cultures (Bacto-Brain Heart Infusion Agar -BHI) were inactivated by the addition of 0.2mL of 40% formaldehyde. Negative serum samples to Salmonella was also obtained from SPF chickens maintained by Institute for Animal Health, UK. All serum samples were stored at -20ºC.

Test sera
Samples of serum from flocks naturally and experimentally infected with 1mL broth culture of S. Gallinarum (10 8 cfu/mL) or S. Pullorum (10 8 cfu/mL) were tested. In addition, the serological response of layers vaccinated and vaccinated/challenged against S. Enteritidis was analyzed.

ELISA Procedure Peroxidase
The ELISA was conduced as described by Hassan et al. (1990) and Barrow et al. (1992) with some Oliveira GH de, Berchieri Júnior Â, Montassier HJ, Fernandes AC Assessment of Serological Response of Chickens to Salmonella Gallinarum and Salmonella Pullorum by Elisa modifications. Washed whole cells of a virulent S. Gallinarum, strain 9 Nal r (10 8 cfu/mL) (Barrow et al., 1992), were used to prepare the soluble protein detecting antigen using 8 cycles of sonication (Branson Sonifer 250, USA) at 85 Watts and 30-second intervals. Flexible polyvinyl ELISA microplates (Cliniplate Labsystems, Finland) were used. The assay included five steps, using 50mL of the appropriate reagent in four steps and 100mL in one. After each step, reagents were removed by aspiration, and wells were washed with PBS containing 0.1% Tween 20 (PBST). Firstly, the antigen was diluted in carbonate-bicarbonate buffer pH 9.6 (15mM Na 2 CO 3 , 35mM NaHCO 3 , 0.3mM NaN 3 ), added to the wells and incubated for 18h/4ºC. In the second step, 100mL carbonate-bicarbonate buffer pH 9.6 with 10% skimmed milk (Molico Nestlé, Brazil) was added to each well for 45min/37ºC. Test serum samples were diluted in PBST with 10% skimmed milk and added to the plate. Afterwards, it was added in each well rabbit anti-chicken IgG peroxidase conjugate (Sigma, A-9046) diluted 1:2,000 in PBST. Finally, the substrate orthophenylenediamine (10mg/mL) (OPD Sigma, P-8287) was diluted in 25mL of citrate phosphate buffer pH 4.9-5.2 (0.1M C 6 H 8 O 7 , 0.2M NaHPO) and 100mL of hydrogen peroxide. The substrate was then added to the wells and the microplate was incubated for 15 minutes at room temperature. The reaction was stopped by addition of 2N chloride acid (HCl). Absorbance values were read at 490nm using an automatic ELISA reader (Microplate Reader model 550, Bio Rad). Each serum sample was tested in duplicate and the mean was calculated and used in further analysis. All incubations were done in a moist chamber.

Alkaline Phosphatase
The steps of the alkaline phosphatase ELISA were the same as described above for peroxidase ELISA, except for two alterations. In this case, a rabbit antichicken IgG alkaline phosphatase conjugate (Sigma A-9171) was diluted 1:1,000 in PBST and the substrate p-nitrophenylphosphate (5mg/mL) (pNPP -Sigma N-9389) was diluted in 5mL of diethanolamine buffer pH 9.8 (100mM diethanolamine, 500mM MgCl 2 ). After the substrate was added to the wells, the microplate was incubated for 30 minutes at room temperature. The reaction was stopped by the addition of 3M sodium hydroxide and absorbance values were read at 405nm.

RESULTS
According to the results, the test should be performed using the antigen diluted at 1:25,000 (5.67mg/mL) for both conjugates and the serum sample diluted at 1:1,000. Results were considered positive when OD=0.800. In this condition, the test is suitable to detect serological response to S. Gallinarum and S. Pullorum. Additionally, the assay performed with alkaline phosphatase and peroxidase conjugate is able to separate positive reactions to S. Gallinarum and S. Pullorum from other serotypes, including positive reaction to S. Enteritidis (Tables 1 and 2).

DISCUSSION
Unlike avian parathyphic Salmonella serotypes, Salmonella enterica serotypes Gallinarum and Pullorum are not frequently excreted in chicken feces, but infected birds tend to produce humoral response (Barrow et al., 1992). Breeding flocks must be free of Salmonella serotypes Gallinarum and Pullorum, therefore, it is very important to detect them as soon as possible to prevent both disease and dissemination. Among other measures, the pullorum test has been the method of choice to detect birds infected with both serotypes. Besides being a very simple method, since it uses whole blood sample, this test has also been very useful to control and eradicate S. Pullorum (Barrow, 1994b). Currently, farmers are allowed to vaccinate breeding flocks against Salmonella enterica serotype Enteritidis in many areas in the world, including Brazil. Since they have similar somatic antigens, cross-reactions may occur. Therefore, another serological test would be interesting for flock assessment instead of using only bacteriological examination.
According to Cooper et al. (1989), the ELISA is effective to detect serological response in chickens against Salmonella Enteritidis and Hassan et al. (1990) demonstrated that the ELISA could be performed using soluble protein antigen. This study assessed the ability of an immunoenzimatic assay performed with either peroxidase or alkaline phosphatase conjugates to investigate serological response to a soluble protein antigen from Salmonella Gallinarum 9 strain. The antigen was capable to promote reliable serological reaction as suggested by Iba et al. (1991) and Barrow et al. (1992).
The antigen diluted at 1:25,000 was able to detect IgG in serum samples diluted 1:1,000. According to , IgG was produced in chickens infected with invasive serotypes like Gallinarum, Pullorum, Typhimurium and Enteritidis and persisted longer than the others immunoglobulins. Thus, the test might be useful for monitoring programs since S. Gallinarum and S. Pullorum are not often present in feces. In these conditions, ELISA can be performed with peroxidase or alkaline phosphatase conjugates. However, the test was more specific when using the alkaline phosphatase conjugate (Table 1). Therefore, the alkaline phosphatase ELISA can be used as a serological test for monitoring poultry flocks to search for Salmonella specific IgG and could be a useful to examine serum samples, egg yolk and dried blood from any kind of poultry flock (Barrow, 1994a).
In order to specifically detect pullorum disease and fowl typhoid, it would be essential to serologically differentiate these diseases from infections caused by other invasive Salmonella serotypes, including Typhimurium and Enteritidis, and some other serotypes in group D, which might induce cross-reacting circulating IgG following oral infection (Hassan et al., 1990;Barrow et al., 1992). The differentiation between serotype infections was assessed in this work. The test using either peroxidase conjugate or alkaline phosphatase conjugate was able to detect humoral response to S. Gallinarum and S. Pullorum. Some intermediary reactions were seen when the ELISA was performed with peroxidase conjugate. However, in such cases, optical density was lower than 0.500. In addition, when it was tested serum samples from commercial flocks infected with S. Enteritidis or vaccinated with S. Enteritidis bacterin, the peroxidase ELISA did not indicate positive reactions. Despite that serum samples which were positive to S. Enteritidis showed OD=0.496 in the ELISA performed with peroxidase conjugate, this test was able to separate positive reactions between S. Gallinarum and S. Pullorum from S. Enteritidis, such as the alkaline phosphatase ELISA. On the other hand, positive results in the peroxidase test exhibited a very strong color, which could be detected with no need of a microplate reader. For this reason, in view of the lack of agreement among results of the experimental serum samples and field serum samples (Table 2), the peroxidase ELISA could be used in monitoring programs in which no discrimination between S. Gallinarum, S. Pullorum, S. Enteritidis and S. Typhimurium is needed (Barrow, 1994a). Under experimental conditions, peroxidase and alkaline phosphatase ELISAs were able to identify reactor birds.