A total of 156 bovine sera (brucellosis-positive sera = 43, brucellosis-negative sera = 113) and 126 sheep sera (brucellosis-positive sera = 34, brucellosis-negative sera = 92) were provided by the China Animal Health and Epidemiology Center (Qingdao, China). All brucellosis-positive sera were confirmed by the Rose Bengal plate agglutination test (RBPT) and tube agglutination test (SAT). These seropositive sera were obtained from animals confirmed by bacteriological identification for Brucella. Negative sera were originated from a brucellosis-free area in China. The bovine and sheep negative and positive control sera for brucellosis were purchased from the China Veterinary Drug Control Institute for antigenicity analysis and indirect enzyme-linked immunosorbent assay (iELISA).

The reference DNA sequence of Brucella melitensis M28 (NCBI reference numbers are NC_017244.1 and NC_017245.1 for chromosome 1 and chromosome 2, respectively) was used for designing primers, comparing amplified sequences, and so on. In our study, three Tat substrate proteins, namely, ErfK (A0577, Accession: WP_002963714.1), YxeI (A1479, Accession: WP_002964574.1), and EntF (B0249, Accession: WP_002968302.1), were selected, which are related to Brucella virulence. The DNA fragments of ErfK, YxeI, and EntF, synthesized by Sangon Technology Co., Ltd. (Shanghai, China), were inserted into the expression vector pET30a. Expression and purification of the recombinant proteins were performed in our laboratory as described elsewhere (4).

Western blot analysis of antigens was performed as previously described (7). Following separation by SDS-PAGE, the target proteins were transferred to a nitrocellulose membrane. The membrane was blocked with 5% skim milk in phosphate-buffered solution (PBS) for 1 h at room temperature, washed three times with PBST (PBS containing 0.5% Tween-20), and incubated with bovine and sheep brucellosis-positive sera (1,1,000) overnight at 4°C. The membrane was washed with PBST three times and incubated with horseradish peroxidase (HRP)-labeled protein G (Thermo, USA) used at a 1:8,000 dilution. Finally, the membrane was washed with PBST three times, and development of the HRP signal was performed using an electrochemiluminescence (ECL) reagent (Yeasen, China).

The expressed proteins rErfK, rYxeI, and rEntF were analyzed for their efficiency in detecting bovine and sheep brucellosis. The different antigens and combinations are shown in Table 1.

Brucella antibodies in bovine and sheep sera were detected by indirect ELISA. For the procedure, the Tat substrate-coated antigen was produced by a combination of equal concentrations of rErfK, rYxeI, and rEntF and used as one coating antigen. 100 μL of the combined protein group and purified single recombinant Tat substrate proteins at a final concentration of 0.78 μg/mL in 0.05 M bicarbonate buffer were added to each well in a 96-well plate (NUNC, Denmark), and the plate was incubated for 1 h at 37°C. After washing three times with PBST, the plate was blocked in 5% skim milk in PBS for 90 min at 37°C. Next, the plate was washed three times with PBST, 100 μL of serum was added to each well, and the plate was incubated for 2 h at 37°C. Subsequently, the plate was washed three times with PBST, HRP-labeled protein G was added (1,8,000 dilution, PBS) (Thermo, USA), and the plate was incubated at 37°C for 1 h. After washing three times with PBST, 100 μL tetramethylbenzidine (TMB) substrate solution (Sigma, USA) was added to each well for 5 min at room temperature in complete darkness. Then, the reaction was stopped by adding 50 μL of ELISA Stop Solution (Solarbio, China). Finally, the OD₄₅₀ was measured using an ELISA plate reader (TECAN, Switzerland).

Dot plots, receiver operating characteristic (ROC) curves, and the area under the ROC curve (AUC) were obtained to determine the effectiveness of single Tat substrate antigens and combined antigens (22). A cutoff value was calculated by the Youden index (specificity + sensitivity - 1), and true positives (TP), true negatives (TN), false positives (FP), false negatives (FN), accuracy, positive predictive value (PPV), and negative predictive value (NPV) were also obtained. Accuracy, PPV, and NPV were calculated according to the following equations: Accuracy = (TP + TN/TP + FN + TN + FP) × 100; PPV = (TP/TP + FP) × 100; and NPV = (TN/TN + FN) × 100 (11).

To verify the specificity of the three proteins as diagnostic antigens, we analyzed rabbit serum that had been immunized with pathogenic bacteria that were prone to cross-reaction with Brucella, including Yersinia enterocolitica O9, Escherichia coli O157:H7, Mycobacterium tuberculosis, Vibrio cholerae, Legionella, and Salmonella, which was purchased from Tianjin Biochip Corporation (Tianjin, China). The serum was selected for the cross-reaction assay using HRP-labeled sheep anti-rabbit IgG secondary antibody (1:5,000) (Thermo, USA). To estimate the inter- and intra-assay precision, the coefficients of variation (CVs) of repeated tests within batches and between batches, respectively, were all calculated five times by using the positive and negative brucellosis serum samples (12). The detection procedures were the same as those described in section serum testing.

Dot plot and receiver operating characteristic (ROC) analyses were performed using GraphPad Prism version 10.0.0. A p-value of <0.05 was considered to be statistically significant.

Tat substrate proteins ErfK, YxeI, and EntF were expressed and purified using a previously described hybrid procedure of denaturing–renaturing procedure (13, 14). The SDS-PAGE analysis showed that the majority of the expressed three Tat substrate proteins were found in the inclusion body (Figure 1), and the Ni-NTA Spin Kit purified the inclusion body. After expression and purification, SDS-PAGE confirmed that the three Tat substrate proteins were successfully purified, with the molecular weights of rErfK, rYxeI, and rEntF being 25 kDa, 38 kDa, and 21 kDa, respectively (Figure 2).

The ability of these three Tat substrate proteins in diagnosing bovine and sheep brucellosis was evaluated using bovine and sheep positive and negative sera by the Western blot method. The results showed that the three Tat substrate proteins of B. melitensis M28 were recognized by the positive sera against B. melitensis and B. abortus. All tested proteins were not reacted with negative sera (Figure 3). Taken together, our results suggest that the Tat substrate proteins ErfK, YxeI, and EntF of B. melitensis M28 could be used in preparation of Tat substrates by ELISA.

The optimal conditions for antigen coating and serum dilution in iELISA were determined by chessboard titration. Individual Tat substrate proteins and their combination were tested. The combination tests showed the best abilities in discerning bovine and sheep brucellosis-positive sera from negative sera, with maximum P/N ratios of 2.29 and 2.59, respectively. For the EntF group, the maximum ratio of P/N was 2.03 and 2.50, respectively, with an optimal dilution of serum being 1:400. ErfK and YxeI showed similar abilities to discern sheep brucellosis-positive and negative sera, with maximum P/N ratios of 2.37 and 2.33, respectively (Table 2).

Using brucellosis-positive and brucellosis-negative sera under the aforementioned optimal conditions, the efficiencies of individual Tat substrate proteins and their combination in serologically detecting bovine and sheep brucellosis were evaluated and compared.

For the detection of bovine brucellosis, according to the iELISA results, the average OD₄₅₀ values of 43 positive sera were 0.781, 0.686, 0.736, 0.503, and 0.783 for the combination and ErfK, YxeI, EntF, and lipopolysaccharides (LPS), while those of 113 negative sera were 0.341, 0.352, 0.318, 0.262, and 0.525, respectively (Table 3). The dot plot demonstrated that the combination and individual Tat substrate proteins ErfK, YxeI, and EntF showed difference in distinguishing bovine brucellosis-positive and negative sera (Figure 4A). The ROC curves were also obtained for these four antigens (Figure 5A). For the detection of bovine brucellosis, the AUC values of the combination, ErfK, YxeI, EntF, and LPS were 0.9549 (95% CI, 0.9190 to 0.9908), 0.8927 (95% CI, 0.8301 to 0.9553), 0.9121 (95% CI, 0.8474 to 0.9768), 0.9164 (95% CI, 0.8591 to 0.9738), and 0.965 (95% CI, 0.9354 to 0.9946), respectively. The combination showed improved accuracy and was superior to individual Tat substrate proteins alone. Based on the Youden index, the optimal cutoff value for the combination was 0.462. At the optimal cutoff value, 41 out of 43 positive samples were accurately diagnosed, and 106 out of 113 negative samples were accurately diagnosed. The positive predictive value (PPV) was 85.42%, and the negative predictive value (NPV) was 98.15%, making the combination the best performing antigen in serological diagnosis of bovine brucellosis. Correspondingly, the combination showed the highest accuracy (94.23%) in the serological detection of bovine brucellosis, similar to LPS (94.87%). The accuracies of individual Tat substrate proteins ErfK, YxeI, and EntF were 91.02, 93.58, and 91.02%, respectively (Table 3).

For the detection of sheep brucellosis, the average OD₄₅₀ values of 34 positive sera were 0.518, 0.549, 0.545, 0.490, and 0.704 for the combination, ErfK, YxeI, EntF, and LPS, while those of 92 negative sera were 0.199, 0.227, 0.234, 0.195, and 0.391, respectively. The dot plot demonstrated that the combination and individual Tat substrate proteins ErfK, YxeI, and EntF showed diverse efficiency in distinguishing brucellosis-positive and brucellosis-negative sera (Figure 4B). The ROC curve analysis showed that the AUC values of the combination, ErfK, YxeI, EntF, and LPS were 0.8864 (95% CI, 0.7742 to 0.9585), 0.8008 (95% CI, 0.7088 to 0.8928), 0.8197 (95% CI, 0.7188 to 0.9205), 0.8378 (95% CI, 0.7588 to 0.9168), and 0.9019 (95% CI, 0.9081 to 0.9916), respectively. Based on the Youden index, the optimal cutoff value for the combination was 0.3131. The positive predictive value (PPV) was 74.36%, and the negative predictive value (NPV) was 94.25%, indicating the combination as the best performing antigen in serological diagnosis of sheep brucellosis. The accuracies of the combination, ErfK, YxeI, EntF, and LPS were 88.10, 81.75, 86.51, 84.92, and 92.86%, respectively (Table 3).

To verify whether Tat substrate proteins employed as diagnostic antigens possess cross-reactivity with other bacteria, we selected rabbit sera that were infected with other bacteria or immunized with zoonotic pathogens for a cross-reactivity test. The results showed that ErfK had cross-reactivity with Salmonella according to an S/N (OD₄₅₀, sample/negative)>2.1, and the combination, YxeI and EntF, showed no cross-reactivity to the rabbit sera immunized with Yersinia enterocolitica O9, Escherichia coli O157:H7, Mycobacterium tuberculosis, Vibrio cholerae, Legionella, and Salmonella, exhibiting relatively good specificity (Table 4). Regarding the intra- and inter-assay precision, the coefficients of variation (CVs) of repeated tests within batches and between batches were confirmed to be less than 10% (Table 5), suggesting the good stability of Tat-iELISA.