Comparison of Cattle Serum Antibody Responses to Five Different Mycobacterial Antigens in an ELISA System

1. Introduction

Worldwide, tuberculosis (TB), mainly caused by Mycobacterium tuberculosis, is one of the top 10 causes of death. According to the World Health Organization (WHO) in 2017, millions of people fall ill with TB every year. Globally, the best estimate is that 10.0 million people (range, 9.0–11.1 million) developed TB disease in 2017: 5.8 million men, 3.2 million women, and 1.0 million children.

TB is considered a significant health and economic threat, particularly in developing countries ( 1 ). In addition to mycobacterial pathogens, other species such as environmental, atypical, and nontuberculous mycobacteria (NTM) may cause disease both in humans and livestock ( 2 ).

Mycobacterium avium subspecies paratuberculosis (MAP) is the etiologic agent of severe gastroenteritis in ruminants known as Johne’s disease ( 3 ). Moreover, MAP has a zoonotic role in Crohn's disease among children and adolescents. This species is prevalent in domesticated animals worldwide and significantly affects the global economy ( 3 ).

Currently, there are several challenges in the diagnosis of disease using traditional tests ( 4 ). This might be due to the structural complexity of the bacterium and the presence of common antigens between pathogenic and non-pathogenic Mycobacterium species and Mycobacterium bovis bacillus Calmette-Guérin (BCG) ( 5 ).

Mycobacterium species can be found using appropriate antigens for detection. In one study, mycobacterial antigens were classified into four groups by performing a gel double diffusion of bacterial antigens. The first group contained the antigens found in all mycobacteria. The second, third, and fourth groups comprised of antigens in slow- and fast-growers and species-specific antigens, respectively ( 6 ). 

As a result, the identification and use of specific and immunogenic antigens play a critical role in the diagnosis of disease by the constitution of serologic methods ( 7 ). For instance, ELISA is a rapid, easy-to-perform, and user-friendly test for early detection of infection ( 5 ).

The purpose of this study was to investigate different Mycobacterium species, including pathogenic Mycobacteria, NTM, and M. bovis BCG with whole proteins. Therefore, the ELISA method tested the protein antigens of different mycobacterial strains and compared with the purified protein derivative (PPD) skin test.

2. Materials and Methods

2.1. Bacterial Strains

In this study, five strains of Mycobacteria including Mycobacterium tuberculosis (ATCC:35808), phlei  (ATCC:11758), bovis (ATCC:1173-P2 BCG), bovis AN5 (ATCC:35726), and MAP (ATCC:316F) were provided by the Department of tuberculosis of Razi Vaccine and Serum Research Institute to investigate. The presence of all these strains has already been confirmed by the polymerase chain reaction test. 

The bacteria were cultured on a Lowenstein-Jensen medium and incubated at the temperature of 37°C for two months. The samples were transported to a biphasic potato-Dorset-Henley liquid medium for 60 days at the temperature of 37°C for bacterial adaption. For expression of bacterial antigens, the samples were transferred to Henley-Dorset liquid medium for 60 days at the temperature of 37°C. To inactivate bacteria, a Fernbach culture flask containing media and bacteria was placed at the temperature of 68°C for 60 minutes. The proteins in the culture were filtered using 0.22 µm Eks filters and precipitated by adding trichloroacetic acid (TCA).

2.2. TCA Precipitation

Precipitation of tuberculin was performed by adding TCA 40% with the ratio of 9:1 for 24 hours. This solution was allowed to stand overnight at room temperature without stirring. The next day, most of the supernatant was removed by aspiration and precipitated residue was used to resuspend and transfer to centrifuge bottles. The bottles were centrifuged at 2500 rounds per minute for 10 min to remove residual supernatant. The precipitate was washed once with TCA 1% then with NaCl 10%. 

At last, the extracted tuberculin was diluted with 3.5 mM Na₂HPO₄.7H₂O, 10% Glycerol, 5 mM Phenol, 8.5 mM NaCl, and D.W up to one liter. The pH was adjusted to 6.9 ± 0.1 with 10 N NaOH. This method of precipitation was used for five strains of M. tuberculosis, M. phlei, M. bovis AN5,  M. bovis BCG, M. avium subspecies paratuberculosis. 

2.3. Determination of Protein Amount by Lowry Method

The extracted protein was measured by Lowry assay, which is more sensitive, rapid, and accurate in comparison to other methods ( 8 ). This stage was performed on bacteria mass, which was inactivated by water steam (68°C) for one hour. In this process, bovine serum albumin (BSA) was used as a protein concentration standard. The samples’ absorption was read at 750 nm by a spectrophotometer.

2.4. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE)

After gel (12%) staining and visualization with Coomassie Blue dye, the Laemmli system was used as an SDS-PAGE method for separating proteins under denaturing conditions ( 9 ).

2.5. Study Population

Serum samples were collected from 54 livestock and obtained at different phases of the study from healthy, naturally infected, or experimentally sensitized cattle with different antigens of Mycobacterium tuberculosis, phlei, bovis BCG,  bovis AN5, and MAP crude antigens. In this study, a total of 12 Holstein bulls aged 1 to 6 years with the lowest antibody titers against mycobacterial antigens were included. They had negative mycobacterial culture test results for two months before and after sensitization. The samples were sensitized by crude antigens of mycobacteria based on World Organization for Animal Health guidelines. 

The negative control group included five samples with negative culture test results and no clinical symptoms. Moreover, the positive control group contained five samples whose fecal and ELISA test results confirmed the presence of infection. Sensitization was performed with crude antigens in the Deputy of tuberculin and mallein production department, Razi Vaccine and Serum Research Institute, University of Tehran, Tehran, Iran.

2.6. Tuberculin Skin Test

All the study population was evaluated by the tuberculin skin test provided by the Deputy of tuberculin and mallein production department, Razi Vaccine and Serum Research Institute. Before injection, skin thickness was measured using a caliper. The skin test reaction was read after 72 hours, and the induration was measured and recorded. The cattle were classified as PPD-positive if the skin-fold thickness at the injection site was more than 4 mm, negative if it was less than 2 mm, and suspicious if it was between 2 and 4 mm.

2.7. Checkerboard Titration Experiment to Optimize ELISA Parameters

To get the best results from your ELISA assay, the dilution factors of the sample and the detection antibodies must be optimized. If your sample or antibodies are too concentrated, you risk saturating the assay. If they are not concentrated enough, your signal will be weak and difficult to detect. For strong, quantifiable signals, use a checkerboard titration to test for the optimal concentration of sample and detection antibodies.

Dilution levels were evaluated by separate checkerboards with an indirect ELISA method to optimize test parameters, antibodies, and antigens (Kashyap, Rajan, et al. 2007). Serial dilution was performed by adding different mycobacteria antigens to discrete plates with the concentrations of 1-20 μg/mL to 0.1 M bicarbonate coating buffer, pH 9.6. Thereafter, 100μl of the solution was poured into the wells of ELISA microtiter Nunc-Immuno™ 96 well MicroWell™ PolySorp® plates from column A to G. The last column was considered as a negative control containing coating buffer without antigen.

The microtiter plate was incubated for 16 to 24 hours at a temperature of 4°C. Then, the wells were washed three times with 300 μl of 10 mM phosphate-buffered saline (PBS) solution. At this point, each well was filled with 150μl of PBS Tween-20 containing 2.5% casein (Maravel, Ashford, Kent, United Kingdom) as a blocker and placed for one hour at room temperature.

The obtained dilutions of 1/50 V/V and 1/100 V/V negative and positive control sera provided by 10 mM PBS Tween-20 containing 1% BSA were added to each well with the volume of 100 μl and incubated for 30 minutes at room temperature (20 to 25°C). After five times of washing with 300 μL of 1X Phosphate-Buffered Saline, 0.1% Tween® 20 Detergent, and 10 mM PBS containing 0.1% Tween-20, the optimized dilution of 1/5000 was prepared in 10mM PBS, pH 7.2, from anti-bovine anti-IgG Horseradish Peroxidase, conjugated antibody (AbD Serotec-UK).

Thereafter, 100 μl of the solution was added to each well for 30 minutes at room temperature. Consequently, after washing the wells, according to the previous stage, 100 μl of tetra-methyl benzidine substrate was poured into the wells, and the plate was maintained for 15 minutes in darkness at room temperature. Finally, 100 μl of a stop solution, H₂SO₄ 0.1M was added and the plate was read at a wavelength of 450 nm by an ELISA reader (Bio Rad-Model 620). 

2.8. Indirect ELISA Protocol

According to the checkerboard, the best serum and antigen dilutions were identified, and indirect ELISA was performed with 54 samples. The appropriate concentrations of different mycobacterial antigens in the separate plate were mixed to 0.1 M bicarbonate coating buffer, pH 9.6. The other procedures were executed under the checkerboard and appropriate serum dilution of 1/50.

2.9. Data Analysis

Data analysis was performed using SPSS software version 25. The cutoff value for the ELISA assay was calculated using the receiver operating characteristic (ROC) curve based on 54 sera with 38 and 16 negative and positive PPD skin tests, respectively, as a standard test, and the sensitivity and specificity were determined. In all the measurements, P-value less than 0.05 was considered statistically significant.

3. Results

3.1. TCA Precipitation and Determining Protein Concentration with Lowry Protein Assay

The TCA precipitation of proteins is often used to concentrate protein samples or remove contaminants, involving salts and detergents, before downstream applications such as SDS-PAGE or 2D-gels. According to the absorbance readings by Lowry protein assay, the standard curve in Excel software was plotted and the protein concentration was calculated using the obtained formula. BSA or standard solutions were used to draw the standard curve. The concentration of M. bovis BCG, M. phlei, MAP, M. tuberculosis, and M. bovis AN5 proteins precipitation TCA was estimated by Lowry protein assay as 1.5 mg/ml, 5.65 mg/ml, 1.8 mg/ml, 1.5 mg/ml, and 1.4 mg/ml, respectively.

3.2. SDS-PAGE of Different Mycobacterial Crude Antigens

The separation was carried out in SDS-PAGE, based on the molecular weight of the compounds. The crude antigens were loaded on a 12% gel with a 10 to 180 kDa molecular weight marker. After deposition of the samples and staining with the Coomassie Blue method, the presence of proteins with different molecular weights was determined. As a result, different proteins were detected with molecular weights of 20 to 180 kDa, the highest protein accumulation in molecular weights was above 50 kDa. The number of bands observed per Mycobacterium was about 10 bands. The SDS-PAGE proteins of these Mycobacteria are demonstrated in figure 1.

Figure 1. SDS-PAGE* gel staining with Coomassie Blue in different mycobacterial crude antigensSDS-PAGE of TCA** precipitated proteins, right to left, 10 to 180 kDa protein ladder, M. tuberculosis, M. bovis BCG, M. phlei, M. avium subsp. Paratuberculosis, respectively. The number of bands observed in each Mycobacterium preparation was about 10 bands. The highest protein accumulation in molecular weights was above 50 kDa.* sulfate-polyacrylamide gel electrophoresis** trichloroacetic acid

3.3. Tuberculin Skin Test Results in Livestock

All 54 samples were examined by the PPD skin test, in which 38 livestock had negative skin test results, four of them had positive skin test results due to illness, and 12 samples had positive skin test results due to sensitization by Mycobacterium tuberculosis, phlei, bovis BCG, bovis AN5, and MAP antigens. Therefore, approximately 30% of PPD skin tested were positive and 70% were negative (Figure 2; Table 1).

Figure 2. Tuberculin skin test results in livestock54 livestock were examined by the PPD* skin test. There were 38 livestock with negative skin test results, four samples with positive skin test results due to illness, and 12 livestock positive skin test results due to sensitization by a specific antigen.Negative results are shown in the diagram with blue color, sick livestock are shown with red color and positive due to sensitization in green color.* Purified protein derivative

3.4. Checkerboard Results in the ELISA Method

The best concentration of antigen, according to the results of the checkerboard was about M. bovis BCG (7.5 μg/ml), M. phlei (15 μg/ml), MAP (0.9 μg/ml), M. tuberculosis (7.5 μg/ml), and M. bovis AN5 (0.7 μg/ml). The adequate dilution used for 54 sera and positive and negative control was 1/50. The amount of optimum density of five antigens is shown in table 2.

3.5. Determining Cutoff for ELISA System with ROC Curve

Based on the results of optical absorption in indirect ELISA and PPD skin test, the ROC curve was plotted by SPSS software. Considering the best sensitivity and specificity, the optimum cutoff point in the graph was determined, which was 0.452, 0.436, 0.257, 0.111, and 0.923 for M. tuberculosis, MAP, M. phlie, M. bovis AN5, and M. bovis BCG, respectively (Figure 3).

Figure 3. ROC* diagram plotted for different Mycobacterium antigensThis figure shows the ROC diagram plotted for different mycobacterial antigens. A:M. bovis BCG, B:M. avium subsp. Paratuberculosis, C:M. tuberculosis, D:M. phlei, E:M. bovis AN5. The PPD skin test was considered the standard. The red dot on each graph represents the cutoff point desired based on sensitivity and specificity.* receiver operating characteristic

3.6. Determining the Sensitivity and Specificity of the ELISA Test and Agreement of the Tests Based on Kappa Coefficient Calculation

Statistical indices of sensitivity, specificity, and positive and negative predictive values were used for data analysis. The sensitivity of the test for M. tuberculosis, MAP, M. bovis AN5, M. phlie, M. bovis BCG was 83.33, 33.33, 87.5, 66.66, and 66.66, respectively. The Specificity of the test for these species was 72.91, 95.83, 76.08, 70.83, and 68.75, respectively. In addition, the efficiency was measured as 74.07, 88.88, 77.77, 70.37, and 68.51, respectively (Table 3; Figure 3).

Based on the comparison of PPD skin test and ELISA results, in all crude antigens Kappa agreement coefficient was significant (P<0.05).

4. Discussion

TB, the second deadliest infectious disease worldwide is caused by bacteria called Mycobacterium tuberculosis complex ( 10 ). NTM is an infectious disease like TB, but it does not cause TB. This complex can cause a wide range of infections, the most often is pulmonary infection (65–90 %) ( 11 ).

Mycobacterium tuberculosis BCG is the only vaccine widely used in tuberculosis and most countries, including Iran, have used it. Although the BCG vaccine can prevent the spread of tuberculosis, it may not provide complete immunity to people with tuberculosis ( 12 ).

PPD test which is used for immunologic diagnosis of Mycobacterium tuberculosis infection has many constraints, such as being confounded by bacillus Calmette-Guerin (BCG) vaccination or exposure to NTM ( 13 ). Therefore, a system design is needed to identify and differentiate mycobacteria.

In recent years, culture tests were mainly replaced with antigen or antibody detection tests such as enzyme immunoassays for the diagnosis of tuberculosis. These methods can be performed in intermediate-level laboratories with relatively simple equipment ( 14 ).

In this study, the crude antigens were prepared after precipitation with 4% TCA mycobacteria to have crude proteins, followed by Lowry protein assays to confirm the presence and concentration of proteins. At last, the ELISA plate was coated and adjacent to the target sera and the obtained data were analyzed with SPSS software version 25.

According to the results, the sensitivity, specificity, and efficiency of this method for MAP were 33.33%, 95.83%, and 88.88%, respectively. Therefore, this test was capable of designing a diagnostic ELISA system for MAP. A positive predictive value of 50% indicated suitable testability in the detection of MAP cases. On the other hand, a negative predictive value of 92% indicated a proper ability to detect healthy livestock.

In a study conducted in 2016 to develop and optimize an ELISA system for rapid detection of paratuberculosis in cattle, the secreted antigens of MAP was precipitated by the TCA method. The ELISA method was performed on the crude antigen. The Specificity and sensitivity of the system were determined as 99% and 86%, respectively. The results of the mentioned study showed that this system was more sensitive in comparison with the IDEXX ELISA kit ( 15 ).

Regarding the results, the sensitivity, specificity, and efficiency of the test for M. phlei antigens were 66.66%, 70.83%, and 70.37%, respectively. In this study, there was high antigenic cross-reaction considering the detection of M. phlei. The positive and negative predictive values of 22% and 94.44% indicated inappropriate and proper ability in detecting M. phlei infected and healthy livestock, respectively. Therefore, it was not possible to design an ELISA system for the detection of M. phlei. In a study performed in 1984, the preabsorption of tested sera with M. phlei was performed to avoid false-positive reaction on detection of MAP with ELISA method in bovine serum ( 16 ).

In M. bovis BCG antigens, with the test sensitivity of 66.66%, a specificity of 68.75%, and efficiency of 68.51%, it had an average ability to detect healthy and infected livestock. It is impossible with this system to design an ELISA system for the detection of M. bovis BCG or vaccinated cattle. A positive predictive value of 21.05% indicated the inconvenient ability of the test due to high antigenic cross-reactions. The negative predictive value was 94.28%, which indicated an acceptable ability to detect healthy livestock. In one study carried out in 1994 ( 17 ) on the crude antigen of M. bovis BCG, SDS-PAGE analysis showed 10 major and 10 minor structural polypeptides, and all regions were found with SDS-PAGE had cross-reaction with various mycobacteria.

In M. tuberculosis antigens, the sensitivity of the ELISA test was 83.33%, the specificity of the test was 72.91%, and the test efficiency was 74.07%. The negative predictive value of this test was 97.22% indicating a satisfactory ability to detect healthy livestock. The positive predictive value of 27.77% demonstrated a low ability to detect cases of M. tuberculosis and the presence of the antigenic cross-reactions. This antigen was a suitable candidate for the design of the diagnostic ELISA system if the precipitation of protein or fractionation with chromatography was performed. In one study conducted in 1983 ( 18 ), by measuring the IgG antibody to PPD antigen using the ELISA technique, patients with active pulmonary TB could be distinguished from patients with other pulmonary diseases.

In M. bovis AN5 crude antigen, with a sensitivity of 87.5%, a specificity of 76.08%, and 77.77% test efficiency, the test was capable of detecting healthy and infected livestock. This antigen was a suitable candidate for the design of the ELISA system. The negative predictive value in this test was 97.22% indicating an appropriate ability to detect healthy animals, while the positive predictive value of 38.88% showed that the ability to detect infected samples was moderate. Regarding the results of a study performed in 2009 ( 19 ), crude protein fractions were used for the identification of new M. bovis antigens. A proteomic approach allows the identification of new antigens for the detection of bovine TB. Even if the sensitivity and specificity of the tests based on these antigens were not optimal, they may form part of a combination or cocktail for the diagnosis of bovine TB.

In this study, all antigens studied had more than 92% detection ability in healthy animals. The highest sensitivity in ELISA tests was related to the crude antigen of M. bovis AN5, which was more than 87.5%. The highest diagnostic specificity was for MAP crude antigen (95.83%). In addition, the highest efficiency of diagnostic tests with more than 88.88% was about this antigen. In conclusion, MAP and M. bovis AN5 antigens were suitable candidates for the design of diagnostic ELISA due to their sensitivity, specificity, and efficiency.

Authors' Contribution

All authors had an equal role in the design, work, statistical analysis, and manuscript writing.

Ethics

All animal injections and bleeding activities were fully compliant with general animal ethics. Moreover, animal housing was according to standard operating procedures of the Razi Institute, which were mainly based on normal farming methods.

Conflict of Interest

The authors declare that they have no conflict of interest.

Grant Support

Razi Vaccine and Serum Research Institute, Agricultural Research Education and Extension Organization (AREEO), Tehran, Iran. Part of the work was also supported by Lahijan Azad University as a Ph.D. project fund to A. Bahmanjeh.

Acknowledgement

We are thankful to all personnel of Razi Vaccine and Serum Research Institute, especially the deputy of the tuberculin and mallein production department. We acknowledge Vaccine Production Unit, Research &amp; Production Complex, and also Arna Gene Avisa Company for their support.

References

1. Olaleye A, Beke A. Determinants of Survival of Patients with Tuberculosis in Developing Countries. Tuberculosis: IntechOpen; 2018.
2. Honda J, Virdi R, Chan ED. Global environmental nontuberculous mycobacteria and their contemporaneous man-made and natural niches. Front Microbiol. 2018; 9:2029.
3. Harris NB, Barletta RG. Mycobacterium avium subsp. paratuberculosisin veterinary medicine. Clin Microbiol Rev. 2001; 14(3):489-512.
4. Tientcheu LD, Koch A, Ndengane M, Andoseh G, Kampmann B, Wilkinson RJ. Immunological consequences of strain variation within the Mycobacterium tuberculosis complex. Eur J Immunol. 2017; 47(3):432-45.
5. Xu J-N, Chen J-P, Chen D-L. Serodiagnosis efficacy and immunogenicity of the fusion protein of Mycobacterium tuberculosis composed of the 10-kilodalton culture filtrate protein, ESAT-6, and the extracellular domain fragment of PPE68. Clin Vaccine Immunol. 2012; 19(4):536-44.
6. Harboe M, Mshana R, Closs O, Kronvall G, Axelsen N. Cross‐reactions between Mycobacteria: II. Crossed Immunoelectrophoretic Analysis of Soluble Antigens of BCG and Comparison with Other Mycobacteria. Scand J Immunol. 1979; 9(2):115-24.
7. Wang X, Chen S, Xu Y, Zheng H, Xiao T, Li Y, et al. Identification and evaluation of the novel immunodominant antigen Rv2351c from Mycobacterium tuberculosis. Emerg Microbes Infect. 2017; 6(1):1-8.
8. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951; 193:265-75.
9. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970; 227(5259):680.
10. Cheon SA, Cho HH, Kim J, Lee J, Kim H-J, Park TJ. Recent tuberculosis diagnosis toward the end TB strategy. J Microbiol Methods. 2016; 123:51-61.
11. Porvaznik I, Solovič I, Mokrý J. Non-tuberculous mycobacteria: classification, diagnostics, and therapy. Respiratory Treatment and Prevention: Springer; 2016.
12. Barreto ML, Pereira SM, Ferreira AA. BCG vaccine: efficacy and indications for vaccination and revaccination. J Pediatr. 2006; 82(3):s45-s54.
13. Mori T, Sakatani M, Yamagishi F, Takashima T, Kawabe Y, Nagao K, et al. Specific detection of tuberculosis infection: an interferon-γ–based assay using new antigens. Am J Respir Crit Care Med. 2004; 170(1):59-64.
14. Araujo Z, De Waard J, Fernández de Larrea C, López D, Fandiño C, Maldonado A, et al. Study of the antibody response against Mycobacterium tuberculosis antigens in Warao Amerindian children in Venezuela. Mem. Inst Oswaldo Cruz. 2004; 99(5):517-24.
15. Rouhollah K, Ali M, Nader M, Keyvan T, Shamseddin G, Farzaneh F, et al. Development and optimization a high sensitive and specific ELISA system for rapid detection of paratuberculosis in cattle. Int J Adv Biotechnol Res. 2016; 7(1):1-9.
16. MERKAL RS. A method for avoiding false-positive reactions in an enzyme-linked immunosorbent assay (ELISA) for the diagnosis of bovine paratuberculosis. Jpn J Vet Res. 1985; 47(1):111-9.
17. Gupta V, Ram G, Bansal M. Antigenic characterization of Mycobacteriumbovis BCG soluble antigens. Vet Microbiol. 1994; 38(3):227-40.
18. Radin RC, Zeiss R, Phair JP. Antibodies to purified protein derivative in different immunoglobulin classes in the diagnosis of tuberculosis in man. Int Arch Allergy Immunol. 1983; 70(1):25-9.
19. Meikle V, Alito A, Llera AS, Gioffré A, Peralta A, Buddle B, et al. Identification of novel Mycobacterium bovis antigens by dissection of crude protein fractions. Clin Vaccine Immunol. 2009; 16(9):1352-9.