Development and standardization of a western blotting test for detection of antibodies against B. abortus

Brucellosis is an infectious disease caused by bacteria of the genus Brucella spp. with diagnosis based on use of serological techniques. The present study aimed to develop and standardize a western blotting (WB) test for detection of antibodies against B. abortus. Samples from two groups of cattle were analyzed: group I: 60 serum samples from true positive and true negative vaccinated animals (30 positive samples from infected animals according to rose bengal test (RBT), 2-mercaptoethanol, serum agglutination test (SAT) and complement fixation test (CFT) and 30 RBT negatives samples); group II: 383 field samples (90 positive and 293 CFT negative sera). The most reactive band in the western blotting, which properly identified and separated infected from non – infected had a molecular weight of ≤ 20kDa. The sensitivity, specificity and accuracy of the WB compared to RBT was 93%, 99%, 98%, respectively and k= 0.938. When compared to CFT, the sensitivity, specificity and accuracy of the WB was 97%, 98% and 97%, respectively and k= 0.929. The WB developed and standardized in the present study is a serological test with potential use as a confirmatory test for the diagnosis of bovine brucellosis.

Isolation of Brucella abortus is considered the most reliable diagnostic method of brucellosis.The samples of choice for isolation include secretions, purulent content, abomasal content of aborted fetuses, cerebrospinal fluid, pleural fluid, synovial fluid and ascitic fluid (Mantur et al., 2006;Bovine…, 2012).However, bacterial isolation is a complex and expensive technique making it difficult to be routinely used in a laboratory.Serological methods work better in the laboratory environment and several serological tests are used for the diagnosis of bovine brucellosis worldwide (Nielsen et al., 2000).
In Brazil, the most common tests used are the rose bengal test (RBT), the fluorescence polarization assay (FPA), the serum agglutination test (SAT), the 2-mercaptoethanol (2-ME) and the complement fixation test (CFT) as golden standards for the diagnosis of brucellosis (Brasil, 2006;Brasil, 2017).RBT and FPA are highly sensitive; thus, in Brazil, they are used for screening individual animals and herds, respectively.Confirmatory tests such as SAT, 2-ME and CFT are highly specific; therefore, they are used to confirm the diagnosis of animals that were positive in screening tests.CFT is the method of choice for purposes of international trade (Nielsen et al., 2005;Brasil, 2006;Bovine…, 2012).
Western blotting (WB) is an immunoproteomic assay widely used for the detection of small amounts of antibodies (Lee, 2007).This method identifies antigen-antibody interaction and, consequently, detects immunospecific proteins.WB has become a decision-making and interpretation tool due to its high sensitivity and specificity (Kim et al., 2014).Manat et al. (2016) identified the outer membrane protein 28 (OMP28) as a major immunodominant antigen and a potential antigen for developing serodiagnosis of bovine brucellosis.
The present study aimed to develop and standardize a western blotting test for detection of antibodies against B. abortus using the B. abortus 1119-3 strain and to determine the method sensitivity and specificity compared to serological techniques routinely used for the diagnosis of bovine brucellosis in Brazil.

MATERIAL AND METHODS
A saline suspension (0.85% NaCl) of B. abortus 1119-3 strain, acquired from experimentally infected animal by B. abortus provided by the National Agricultural Laboratory of Pernambuco (LANAGRO -PE, Brazil), was used for the extraction of whole-cell proteins.The strain was resuspended in phosphate buffered saline (PBS) to an initial concentration of 4.5 x 10 8 cells/mL estimated at 600nm on a spectrophotometer.The cell suspension was concentrated to 9.8 x 10 9 cells/mL in PBS and used for antigen preparation.Approximately 10mL of the cell suspension was centrifuged at 4000 x g for 10min at 4°C and the supernatant was discarded.Soluble components were removed by three consecutive centrifugation and washing steps with 9mL PBS.After the third washing, 3mL of formaldehyde (Sigma Aldrich, EUA) was added for bacteria inactivation and degradation as well as 6mL of PBS.Then, the cell suspension was incubated on an orbital shaker overnight at room temperature and the washing and degradation processes were repeated two more times.The cell suspension was washed once again with PBS and 3mL of PBS was added to it.Finally, Laemmli Sample Buffer (Sigma Aldrich, EUA) was added to the cell suspension in a 1:2 ratio resulting in the antigen suspension, which was stored at -20°C for later use.
The antigen suspension and a 250 -10kDa molecular weight marker were loaded on a 3%-15% Bis-Tris gel (Invitrogen, EUA) and separated at a constant voltage of 200V for 1 hour using a vertical electrophoresis unit (Invitrogen, EUA) and 20X nupage mops SDS running buffer (Invitrogen, EUA).The gel was stained with SilverXpress kit (Invitrogen, EUA) (Figure 1) and bands with a molecular weight equal or less than 20kDa, which are considered specific markers of B. abortus (Pajuaba et al., 2012;Kim et al., 2014), were used to define positive results.
The antigen was transferred from the gel to a 0.45µm nitrocellulose membrane (Invitrogen, EUA) by blotting at constant 30V for 1 hour using a vertical electrophoresis unit (Invitrogen, EUA) and 20X nupage transfer buffer (Invitrogen, EUA).After transfer, the lane containing the molecular weight marker was cut from the membrane, which was blocked overnight in Blocking Solution (Candor, Germany).After three washing steps with Washing Solution (Candor, Germany) (1:20 dilution) for 20min, the membrane was cut into strips of approximately 3mm.Some strips were used for membrane quality test and the remaining strips were stored in Falcon tubes at -20°C until use (Elschner et al., 2011).Sera from groups I and II were also tested by CFT following the methodology described in chapter 2.4.3 of the manual of diagnostic tests and vaccines for terrestrial animals from OIE (Principles…, 2013).CFT was performed using a whole cell antigen (Tecpar, Brazil) in 1:500 dilution and incubation at 37°C for one hour (Brasil, 2006).The western blotting standardization followed the methodology described by Elschner et al. (2011).
The sera from all 443 animals were diluted in Low Cross buffer (Candor, Germany) at a 1:50 ratio and incubated with the membrane strips for 1.5h on an orbital shaker, and washed three times with washing solution.
The Kappa (k) coefficient was used to study the agreement between the tests used, considering the conventional interpretation of k-values as follows: 0.00 -0.20= poor agreement; 0.21 -0.40= regular; 0.41 -0.60= moderate; 0.61 -0.80= good; 0.81-1.00=very good, negative values are interpreted as zero.The sensitivity (Se), specificity (Sp), positive predictive value (PPV), negative predictive value (NPV) and accuracy (ACC) (Pereira, 2008) of the WB developed in the present study were calculated by comparing the WB results with the RBT and CFT results using 2x2 tables.

RESULTS
Considering the WB and RBT results for cattle of group I, WB showed 100% Se, Sp, PPV, PNV, ACC and k= 1.Similar results were observed when WB was compared to CFT for the same group of animals.
Brucellosis diagnosis is a constant challenge being restricted to serological tests and bacterial isolation (Alton et al., 1988).However, serological tests neither differentiate vaccinated animals from naturally infected animals, nor identify Brucella species nor avoid crossreactions with other microorganisms because the LPS-O antigen present in the B19 vaccine is also present in Gram-negative bacteria (Nielsen et al., 2004;Olsen and Stoffregen, 2005;Al Dahouk et al., 2006;Ko et al., 2012).Thus, the development of tests that do not require the use of LPS O-antigens is very promising because it would help reduce costs associated with vaccination, diagnosis and slaughter of positive animals (Abalos et al., 2000;Pajuaba et al., 2012).
In contrast to other applications of the western blotting test (Al Dahouk et al., 2006;Connolly et al., 2006;Elschner et al., 2011;Lee, 2007;Ko et al., 2012;Kim et al., 2014;Wareth et al., 2014), our study sought to adapt and standardize a technique developed for proteomic studies to serological diagnosis (Towbin et al., 1979), noting that there are no studies on the use of the western blotting test for detection of antibodies against B. abortus.
Studies have described Brucella sp.proteins that could be employed in WB.Pajuaba et al. (2012) and Kim et al. (2014) reported that proteins with ≤ 20kDa molecular weight are specific markers for animals naturally infected with B. abortus, differentiating them from animals vaccinated with the B19 vaccine.For this reason, we chose the ≤ 20kDa molecular weight as a limit between positive from negative animals in the WB developed herein (Figure 2).
In its "Principles and methods of validation of diagnostic assays for infectious diseases", the OIE states that in order to be considered validated, an assay must go through four stages of evaluation, namely, analytical characterization, diagnostic characterization, reproducibility and implementation also known as the "assay validation pathway".The first two stages (analytical and diagnostic characterization) are required for the standardization of an assay, while the latter two (reproducibility and implementation) are mandatory for validation.In the present study, we completed the first two stages recommended by the OIE, analytical and diagnostic characterization, so the WB developed herein could be considered a standardized test (Principles…, 2013).The analytical characterization corresponded to the screening of true positive and true negative samples by WB and methods routinely used in Brazil (RBT and CFT) with comparison of their results and calculation of analytical sensitivity and specificity.The diagnostic characterization was comprised of the screening of field samples for calculation of diagnostic sensitivity and specificity.According to the OIE, an assay test could be considered validated if its sensitivity and specificity values are greater than 95% (Principles…, 2013).The WB standardized in our study showed excellent specificity (99%) but sensitivity (93%) below of limit set by OIE (Principles…, 2013) when compared to RBT; thus, WB should not be considered as an option to replace RBT for herd screening because it did not meet the OIE criteria of ≥ 95% sensitivity and specificity.When compared to CFT, WB had high sensitivity (97%) and specificity (98%); therefore, it could replace CFT as a confirmatory diagnostic tool (Stemshorn et al., 1985;Dohoo et al., 1986;Mac Millian et al., 1990;Nielsen et al., 1999;Pajuaba et al., 2012;).
The k coefficients for WB x CFT and WB x RBT were classified as very good when considering the conventional interpretation of k-values (Pereira, 2008;OIE, 2013).corroborating with results obtained in previous studies on standardization of western blotting for the diagnosis of brucellosis (Pajuaba et al., 2012;Wareth et al., 2014).The use of new tests that use immunospecific proteins for the diagnosis of brucellosis adds greater confidence in the interpretation of the results; however, the WB developed in the present study has to be validated before it can be routinely used.

CONCLUSIONS
There is still a need for the development of serological tests using antigens capable of detecting species-specific against B. abortus antibodies in order to reduce cross-contamination with other microorganisms and, especially, to differentiate vaccinal immunity from acquired immunity, eliminating the false positive results caused by vaccination.In addition, easier and safer diagnostic methods are critically needed in the laboratory routine.The immunoproteomic technique could be used as a confirmatory diagnostic tool for bovine brucellosis in Brazil.It is quick to perform, less dependent on reagents and more reliable and with potential to be capable to differentiate vaccinated from naturally infected animals.Vet. Immunol. Immunopathol., v.160, p.218-224, 2014.

Figure 1 .
Figure 1.3% -15% Bis-Tris Gel silver stained showing the antigen and a 250 -10kDa molecular weight marker.Four hundred and forty three bovine serum samples were tested using the western blotting technique developed in the present study.The samples belonged to two groups, group I: 60 serum samples from 30 true positives animals with clinical signs of brucellosis, B. abortus isolated and positive results in RBT, 2-ME, SAT and CFT and 30 true negative vaccinated animals with negative results in RBT, 2-ME, SAT and CFT; group II: 383 field samples from CFT positive (n= 90) and negative (n= 293) animals.All experimental procedures with animals were conducted in accordance with the ethical principles for animal testing adopted by the Ethic Committee for the use of Animal at the Federal Rural University of Pernambuco.Protocol Number: 102/2015.All the serum samples from groups I and II were screened by RBT following the methodology recommended by chapter 2.4.3 of the manual of diagnostic tests and vaccines for terrestrial