Animals younger than 6 months of age are infected by ingesting bacilli through feed and water, as well as, sucking on teats contaminated with infected feces (Castellanos, 2010).

Affected bovines suffer chronic enteritis, the signs of which are diarrhea, submandibular edema, weight loss, and low body condition that leads eventually to death (Cirone et al., 2007; Speer et al., 2006). Clinical signs are only observed in adults between 18 and 24 months of age. Appetite and general behavior of the animal remain normal during the early stages, but milk production and body condition worsen due to nutrient malabsorption. As the disease advances, there is lethargy, depression, bristly hair, hypoproteinemia and submandibular edema (Mortier et al., 2015).

Tests based on the humoral immune response are the most frequently used for the diagnosis of the infection. One of the more evaluated tests has been the ELISA which is considered a good option for diagnosis, since there are several brands of commercial kits; it has a sensitivity of 50 to 83% and a specificity of 70 to 89% (Martinez et al., 2012).

Another technique that allows the determination of the antigen-antibody interaction is the FPA. The principle of this test is based on the property that antibodies have a high affinity and specificity to join a particular antigen; this property is used to recognize a specific analyte of interest, and by different mechanisms, generate a light signal that may be measured. In the case of FPA, the signal which is being measured is polarized fluorescent light. The size of the molecule is the main factor that influences the rotation speed. For example, a small molecule rotates at a higher speed than a larger molecule. These molecules move and rotate freely in the solution, and the result that is obtained is going to depend on how much the molecule has been able to rotate during the time the excited fluorescence state lasts. The smaller the molecule, the faster it shall rotate, and its polarization of fluorescence shall be lower (Nielsen and Gall, 2001; Marcelo et al., 2017).

If a molecule is marked with a fluorochrome, the rotation time through a 68.5° angle can be established by measuring the intensity of polarized light in vertical or horizontal planes. A large molecule emits more light in a single plane (more polarized) than a small one that rotates faster and emits more depolarized light (Nielsen and Gall, 2001; Marcelo et al., 2017). Polarized fluorescence has been used for the diagnosis of bovine brucellosis and tuberculosis, demonstrating a sensitivity of 95.5% and specificity of 99.0% for brucellosis and 92.9 and 98.3%, respectively for tuberculosis (Surujballi et al., 2002; Jolley et al., 2007; Marcelo et al., 2017).

Paratuberculosis diagnostic methods must be based on their capacity to detect infected animals; the specificity and reliability of serodiagnosis tests is determined by the type of mycobacterial antigens that are used, hence the identification of the structural antigenic components must be well characterized and specific for each mycobacterial species to be able to provide the means to improve specificity and sensitivity of immunodiagnostic assays.The economic losses caused by paratuberculosis on the animal husbandry industry are a serious problem, and therefore it is necessary to have quick tests for diagnostic of the disease. The objective of this work was to compare FPA and ELISA to detect paratuberculosis infected bovines.

Samples

The sampling was carried out in dairy herds, with and without prior paratuberculosis diagnosis, which belonged to cooperating cattlemen. The dairy herds were located in the States of Mexico, Hidalgo, Queretaro and Aguascalientes (Table 1). To estimate the sample size, the cattle census of each State of the Mexican Republic included in the study were taken into consideration, as well as a paratuberculosis prevalence of 10% with a confidence level of 95% and 5% error. Six-hundred and three serum and feces samples were taken from bovines, older than two years of age, most of them from the Holstein breed. The blood was taken from the caudal vein in vacutainer tubes and centrifuged at 180 xg, for ten minutes to obtain the serum; the feces samples were directly taken from the rectum with palpation gloves and were kept at 4°C until their use.

Nested polymerase chain reaction (n-PCR)

DNA extraction from feces was carried out according to the protocol described by Jaimes et al. (2008).

The primers described by Erume et al. (2001), were used for the first PCR reaction: Paratb1 (5´ TGA TCT GGA CAA TGA CGG TTA CGG A 3´) and Paratb 4 (5´ CGC GGC ACG GCT CTT GTT 3´) and the product that was obtained had 563 bp; for the second reaction, Paratb 2 (5´ GCC GCG CTG CTG GAG TTA A 3´) and Paratb 3 (5´ AGC GTC TTT GGC GTC GGT CTT G 3´) were used, obtaining a final product of 210 bp.

For the first reaction, 2 µl (15 ng/ µl) of DNA, obtained from bovine feces were used, with the following reagent conditions: for 48 µl of premix solution, 5 µl were used of Reaction Buffer 10X (67 mM/µl), 4 µl MgCl ₂ 30 mM 20 X, 1 µl DNTP (200mM), 1 µl Paratb 1 (25 pMol), 1 µl Paratb 4 (25 pMol), 0.25 µl Polimerase 500 U and 35.75 µl double-distilled water. Amplification was carried out in an Axigene  thermocycler   using   the   following    program:  an   initial denaturing at 98°C/3 min, followed by 35 cycles of denaturing at 98°C/30 s, annealing at 65 °C/30 s, and extension at 72 °C/30 s, followed by a final extension at 72°C/3 min and then 4°C /three min. For the second reaction, 3 µl of the first reaction were taken as a DNA template and were transferred into PCR microtubes containing the same amount and concentration of reagents previously described with the exception that the Paratb1 and Paratb4 primers were substituted by primers Paratb2 (25 pMol) and Paratb3 (25 pMol). The same thermocycler program was used. For the visualization of the amplification products, electrophoresis was carried out on 2% agarose gels and stained with ethidium bromide.

Establishing test sensitivity and specificity

Receiver operational characteristics (ROC) analysis was carried out, which is a graphic representation of sensitivity and specificity of a binary system according to the discrimination threshold variation. The Kappa Test or Concordance Index was estimated to measure the association with the results obtained in FPA, ELISA and nested PCR. The statistical analysis was carried out with the Intercooled Stata 7.0 and Epidat 3.1 software packages.

The results obtained with the FPA (142/603) and ELISA (141/603) tests were similar; the FPA test identified only one more serum positive (Table 2). The sensitivity and specificity of the FPA technique were 88.50 and 91.42%, respectively, when compared with ELISA and n-PCR. The FPA assay shows high sensitivity and specificity (Table 3, Figure 1). Concordance between tests showed a  good   correlation  of  FPA  when  compared  to  ELISA (0.6742%) and n-PCR (0.7314%); the positive predictive value of FPA was 75.6% when compared with ELISA and 87.7% compared with polymerase chain reaction (PCR). Likewise, the negative predictive values were 90.8 % and 89.3% respectively, which indicates that the test has agreater specificity (Table 4).

Currently, research lines in mycobacteriosis are focusing on the assessment of low molecular weight protein type antigens. These antigens, since they are highly specific, are the most viable candidates to be used in diagnostic tests to detect animals infected with tuberculosis and/or paratuberculosis. The low molecular weight antigens have been previously used in serological diagnostic techniques such as ELISA and FPA, where the sensitivity and specificity has been demonstrated to be higher in both techniques (Beck et al., 2005; Chaubey et al., 2016). In this work, the FPA and ELISA techniques were evaluated using as a confirming test n-PCR. FPA had an 88.5% sensitivity and 91.42% specificity, while ELISA had  83.86%    and    89.87%    respectively;    when  the concordance analysis was carried out, an acceptable result of (K=0.6742) was obtained. Therefore, both tests are a good option to determine the presence of anti-Map antibodies in cattle.

With n-PCR, 162 positive samples were detected while with FPA and ELISA 142 and 141 positives were detected, respectively. It is possible to find samples that are positive in n-PCR but negative in the serology tests, because in the subclinical stage of paratuberculosis the humoral immune response decreases and the cellular immune response increases. Another reason for alteration of antibody production can be the loss of immune response to the infection. This suppressing activity is known as anergy and occurs in immunocompromised animals such as old animals or that are in the final stages of the disease. Nevertheless, it is considered that these animals may be eliminating the bacilli that are not detected by serological diagnosis but are detected by n-PCR (Martinez et al., 2012; Chaubey et al., 2016). Nielsen and Gall (2001), mentioned that the FPA test is more specific than sensitive; albeit Jolley et al. (2007), when using the FPA test with the 22 kDa protein MPB70 from an M. bovis strain for the diagnosis of bovine  tuberculosis,  obtained  a  99.9% specificity. Also,  Surujballi et al. (2009), used FPA for diagnosis of tuberculosis in cervids and obtained a sensitivity and specificity above 81.00%.

FPA has been used in the diagnosis of bovine brucellosis, and the sensitivity and specificity has been reported at 99.00 and 95.50%, respectively (Marcelo et al., 2017). Even though the FPA technique is simple and has several advantages over other serological techniques, up until now there had not been studies on its use and application to the diagnosis of paratuberculosis. The results obtained in this work show that the use of FPA might contribute to and be an alternative for the diagnosis of bovine paratuberculosis.

The Map 3065 strain protoplasmic antigen that was used for ELISA, has the characteristic of being a complex preparation of lipids, carbohydrates, proteins and nucleic acids that contain antigenic determinants in common with most of the Map strains, with which an acceptable sensitivity and specificity has been obtained (Martinez et al., 2012). In comparison, the antigen that is used in the FPA test is a protein fraction of 35 kDa, and as such has more specificity. Proteins below 50 kDa are involved in the humoral and cellular response to intracellular microorganism infections, and therefore they are candidates to be used as antigens in the FPA test since they allow for more specificity of the assay (Nielsen and Gall, 2001; Franco et al., 2005; Mon et al., 2012; Chaubey et al., 2016).

Bauman et al. (2019)  evaluated Bulk tank milk (BTM) to determine Map's presence at the herd level in production, using   fecal- culture, PCR,  mELISA paratuberculosis tests in 29 dairy goat herds  and 21 dairy sheep flocks; theirs results showed that  mELISA   was more  sensitive   for  identifying  farms  with  affected  animals  (those  with  detectable  circulating  antibodies  in  their  serum  or  milk).  Increasing sensitivities to 87.50% (serum ELISA as reference  test)  and  71.40%  (milk  ELISA  as  reference  test)  in  dairy  goats  and  72.70%  (serum  ELISA  as  reference  test) and 87.50% (milk ELISA as reference test) in dairy sheep; While the sensitivity was 50% and speficivity  83.00%;  when used PCR of the BTM and fecal culture. BTM has been evaluated in dairy cattle as a potential herd-level paratuberculosis testing strategy  so  inclusion of FPA for the detection of anti-map antibodies in milk, is an alternative that would allow to have a diagnostic test with greater sensitivity and specify, as well  as  determine  the health status of paratuberculosis of the herd in production.  

Control of paratuberculosis in herds is based on the opportune identification and elimination of infected animals. Nevertheless, the low sensibility that diagnostic tests have and the subclinical presentation of the infection, make the diagnosis of the disease difficult, especially in young animals (Chaubey et al., 2016).

The FPA test using the 35 kDa antigen may be considered as an alternative that allows better sensitivity and specificity to diagnose paratuberculosis.

The authors have not declared any conflict of interests.

The authors are grateful to INIFAP SIGI 12484419581 for the financial resources provided.