https://orcid.org/0000-0002-7164-9320
Figures
Abstract
Background
Brucellosis is a widespread zoonotic disease with approximately 2.1 million new human cases annually. Traditional serological methods, which rely on lipopolysaccharide (LPS) as a diagnostic antigen, suffer from cross-reactivity with other Gram-negative bacteria. To address these issues, we aimed to develop a multiepitope fusion protein for improved brucellosis diagnosis.
Methods
We identified linear B-cell epitopes from Brucella proteins using the Immune Epitope Database (IEDB) and constructed a multiepitope fusion protein. The fusion protein was expressed through prokaryotic expression, purified, and was evaluated using an indirect enzyme-linked immunosorbent assay (iELISA) with serum samples from patients with confirmed brucellosis (n = 279) and negative controls from febrile patients (n = 126). To rigorously assess cross-reactivity, it was also tested against a separate panel of sera from 283 non-brucellosis febrile patients with laboratory-confirmed infections by other bacterial pathogens.
Results
The engineered fusion protein, consisting of 11 optimized linear B-cell epitopes derived from the consolidation of 23 initial epitopes obtained from the IEDB, was assessed using iELISA. This evaluation yielded an area under the curve (AUC) of 0.9912 relative to LPS, demonstrating a sensitivity of 95.34% and a specificity of 93.65% in comparison to the negative control group (n = 126). Critically, when tested against the distinct cross-reactivity panel, the fusion protein exhibited cross-reactivity with 9 serum samples from 283 patients infected with other bacterial pathogens, whereas LPS showed cross-reactivity in 41 out of 283 samples.
Conclusions
When evaluated against a clinically relevant control cohort of non-brucellosis febrile patients, the fusion protein demonstrated a substantial advantage by exhibiting significantly lower cross-reactivity compared to LPS, which frequently cross-reacted with other bacterial infections. The multiepitope fusion protein developed in this study demonstrates significant potential as a diagnostic tool for brucellosis. However, the study’s limitations, including a small sample size and lack of information on Brucella species, suggest the need for further research with larger and more diverse sample sets to fully validate its clinical application.
Author summary
This study developed a multiepitope fusion protein for improved diagnosis of human brucellosis. Linear B-cell epitopes for Brucella proteins were identified from the Immune Epitope Database (IEDB) and constructed a fusion protein comprising 11 optimized epitopes (derived from 23 initial epitopes). The fusion protein was expressed, purified, and evaluated using indirect enzyme-linked immunosorbent assay (iELISA) with serum samples from confirmed brucellosis and non-brucellosis patients. Results showed the iELISA using the fusion protein had an area under the curve (AUC) of 0.9912, with 95.34% sensitivity and 93.65% specificity, outperforming lipopolysaccharide (LPS) in diagnostic performance and cross-reactivity. The fusion protein exhibited low cross-reactivity with bacteremia infections (9 out of 283 sera), while LPS showed substantially higher cross-reactivity (41 out of 283 samples). Although the study demonstrated the fusion protein’s potential as a diagnostic tool for brucellosis, its limitations, including a small sample size and lack of Brucella spp. information, suggest the need for further research to fully validate its clinical application.
Citation: Zhan T, Li Y, Xie Y, Li S, Lian L, Han W, et al. (2025) Multiepitope fusion protein-based ELISA for enhanced brucellosis serodiagnosis. PLoS Negl Trop Dis 19(12): e0013804. https://doi.org/10.1371/journal.pntd.0013804
Editor: Richard A. Bowen, Colorado State University, UNITED STATES OF AMERICA
Received: August 25, 2025; Accepted: November 28, 2025; Published: December 3, 2025
Copyright: © 2025 Zhan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data generated or analysed during this study are included in this published article and its supplementary information files.
Funding: This work was supported by QingLan Project of Jiangsu Province Grant to DY (2024) and Postgraduate Research & Practice Innovation Program of Jiangsu Province Grant to YX (Grant number KYCX25-3209). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Brucellosis is a zoonotic infection of global significance, with an estimated 2.1 million new cases of human brucellosis reported annually [1]. The lack of specific clinical symptoms associated with brucellosis complicates early diagnosis, often resulting in misdiagnosis and the potential for chronic infection, which can have severe detrimental consequences [2,3]. Therefore, the accurate diagnosis of Brucella infection is critically important [3]. Currently, the diagnosis of brucellosis predominantly relies on serological methods, including tube agglutination tests, the tiger-red plate test, and enzyme-linked immunosorbent assays (ELISA) [4,5]. Traditional serological methods utilize lipopolysaccharide (LPS) as a diagnostic antigen; however, LPS is also found in other Gram-negative bacteria, leading to cross-reactivity with other bacteria such as Yersinia enterocolitica O9 and Escherichia coli O157:H7 during the serological diagnosis of brucellosis [6,7]. Consequently, current research in the serological diagnosis of brucellosis concentrates on the identification of appropriate alternative antigens.
Numerous studies in recent research have focused on the development of multiepitope fusion proteins through the application of bioinformatics techniques [8,9]. This approach entails predicting epitopes from Brucella antigenic proteins and assembling them into multiepitope fusion proteins for use as diagnostic antigens following prokaryotic expression [10,11]. In this methodology, the selection of Brucella antigenic proteins is therefore of paramount importance. A significant body of research has concentrated on Brucella outer membrane proteins, among other candidates; nonetheless, no antigenic protein has been identified that rivals LPS in efficacy [12]. Although bioinformatics technology has generated valuable insights for the identification of alternative antigens, it is important to acknowledge its limitations. Specifically, the predicted epitopes primarily consist of linear epitopes that can be expressed in prokaryotic systems, while the conformational epitopes, which represent the majority of epitopes, remain unexpressed [13,14]. Furthermore, the predictions made by bioinformatics methods are often associated with a significant degree of uncertainty. For instance, the prediction accuracy rates of Bepipred and SVMTriP are reported to be 65.93% and 55.2%, respectively [15,16]. In contrast, traditional methods for epitope characterization and identification involve the validation of overlapping synthetic peptides on an individual basis, which incurs substantially higher costs compared to bioinformatics predictions [17,18].
The Immune Epitope Database (IEDB, www.iedb.org) is a publicly accessible resource that is funded by the National Institute of Allergy and Infectious Diseases (NIAID). This database compiles experimental data on antibody and T cell epitopes that have been studied in humans and various animal species, with a particular focus on infectious diseases, allergies, autoimmunity, and transplantation [19,20]. Our search of the IEDB revealed a significant number of experimentally validated linear B-cell epitopes associated with Brucella. As a result, we aggregated the documented Brucella linear B-cell epitopes from the IEDB, constructed a multi-epitope fusion protein, and evaluated its potential utility for the diagnosis of brucellosis.
2. Methods
2.1. Ethics statement
All methods were carried out in accordance with the Declaration of Helsinki. The study was reviewed and approved by the Ethics Committee of Xuzhou Medical University (approved number: xzhmu-2024Z052) and verbal informed consent was obtained from all subjects.
2.2. Serum samples
A total of 100 positive serum samples and 96 negative serum samples were obtained from the Xuzhou Center for Disease Control and Prevention, while 179 positive serum samples and 30 negative serum samples were sourced from the Shandong Center for Disease Control and Prevention. Each positive sample was validated as positive through the Standard Agglutination Test (SAT), while all negative serum samples were obtained from febrile patients suspected of brucellosis but subsequently confirmed to be negative by SAT. All negative and positive samples were collected from the same respective regions within Xuzhou and Shandong. Moreover, to rigorously evaluate cross-reactivity, serum samples from 283 non-brucellosis febrile patients with laboratory-confirmed infections of other pathogens were obtained from the First Affiliated Hospital of Jilin University. These included patients with confirmed bacterial infections (e.g., Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Staphylococcus aureus) and other common pathogens (as detailed in S2 File). The diagnoses were confirmed by microbial culture, PCR, or specific serological tests as part of the hospital’s routine diagnostic procedures. Importantly, all these patients were confirmed to be negative for brucellosis by the SAT to exclude any co-infection or misdiagnosis.
2.3. Selection of Brucella Linear B-cell Epitopes
The identification of linear B-cell epitopes for Brucella was performed utilizing the IEDB (https://www.iedb.org/) with the following fundamental parameters [21]: Epitope type: Linear peptide; Assay type: B Cell; Outcome: Positive; Epitope Source: Brucella. All other parameters were maintained at their default settings. The retrieved B-cell epitopes were subsequently optimized for integration. In instances where amino acid sequences overlapped, the epitope was regenerated to ensure that the integrated epitope encompassed the overlapping amino acid sequence. After that, BLAST analysis was performed to verify the specificity of the selected Brucella linear B-cell epitopes.
2.4. Construction and evaluation of Brucella multiepitope fusion protein
The optimized integrated Brucella linear B-cell epitopes were tandemly linked, with adjacent epitopes connected by the linker sequence ‘GGGS’ to construct their corresponding amino acid sequences. Subsequently, the aliphatic index, instability index, isoelectric point (pI) and molecular weight (MW) of the protein were predicted for the amino acid sequence of the tandemly linked fusion protein using ProtParam (http://web.expasy.org/protparam/), which is available on the ExPASy website. The physicochemical properties of the fusion proteins were thoroughly evaluated. Additionally, the three-dimensional molecular model of the fusion protein was predicted using I-TASSER (http://zhanglab.ccmb.med.umich.edu/I-TASSER/) to assess the spatial conformation of the fusion protein. The antigenicity of the fusion proteins was predicted and evaluated using VaxiJen software (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html, threshold: 0.4, default).
2.5. Preparation and purification of fusion proteins
In the preparation of fusion proteins, established references were employed to guide the prokaryotic expression and purification processes [10,11]. The specific methodology is outlined as follows:
2.5.1. Plasmid construction.
The amino acid sequence of the designed multiepitope fusion protein was reverse-translated and codon-optimized for expression in Escherichia coli (E. coli). A 6 × His tag was incorporated at the C-terminus to facilitate purification. The optimized gene sequence was synthesized de novo and cloned into the pET-30a(+) expression vector between the Nde I and Xho I restriction sites.
2.5.2. Transformation.
The constructed expression plasmid was transformed into E. coli BL21 (DE3) competent cells using a standard heat-shock method. Briefly, the plasmid was mixed with competent cells on ice, heat-shocked at 42°C for 90 seconds, and then placed back on ice. After the addition of LB medium and a recovery incubation, the cells were plated on LB agar plates containing the appropriate antibiotic and incubated overnight at 37°C.
2.5.3. Protein expression.
For initial small-scale expression, a single transformed colony was inoculated into 1.5 mL of LB medium and cultured at 37°C with shaking at 200 rpm until the OD600 reached 0.6-0.8. Protein expression was induced by adding isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final concentration of 0.5 mM, followed by incubation at 37°C for 2 hours. The bacterial cells were then harvested by centrifugation, resuspended in 10 mM Tris-HCl (pH 8.0), and lysed by boiling in 2 × SDS loading buffer. Expression of the target protein was analyzed by 12% SDS-PAGE.
For large-scale protein production, the above culture was scaled up to 250 mL of LB medium. When the OD600 reached 0.6-0.8, IPTG was added to a final concentration of 0.5 mM, and induction was continued at 37°C for 4 hours. The bacterial cells were harvested by centrifugation at 8000 rpm for 6 minutes. The cell pellet was resuspended in 20–30 mL of 10 mM Tris-HCl (pH 8.0) and lysed by ultrasonication (500 W, 180 cycles of 5 s sonication with 5 s intervals). The soluble fraction (supernatant) and insoluble fraction (precipitate) were separated by centrifugation at 12,000 rpm for 10 minutes and analyzed by 12% SDS-PAGE.
2.5.4. Protein purification.
The nickel affinity column (Ni Sepharose 6 Fast Flow, GE Healthcare) was initially washed with deionized water to achieve a pH of 7.0. Subsequently, the nickel was adjusted to a pH range of 2–3. The column was then washed again with deionized water to restore a pH of 7.0. Following this, the nickel column was equilibrated with approximately 100 mL of 10 mM Tris-HCl buffer at pH 8.0, followed by an additional equilibration with approximately 50 mL of the same buffer supplemented with 0.5 M NaCl. The samples were diluted appropriately for analysis, ensuring that the final concentration of sodium chloride in the samples was 0.5 M. At the conclusion of the sample run, the column was washed with 10 mM Tris-HCl buffer (pH 8.0) containing 0.5 M sodium chloride. Protein elution was performed using 10 mM Tris-HCl buffer (pH 8.0) containing varying concentrations of imidazole: 15 mM, 60 mM, and 300 mM, each supplemented with 0.5 M NaCl. The effectiveness of the protein purification process was assessed using 12% SDS-PAGE electrophoresis, and protein quantification was conducted utilizing the BCA Protein Quantification Kit (Beyotime, Shanghai, China). The purity of the purified fusion protein was determined by analyzing the Coomassie Brilliant Blue-stained SDS-PAGE gel using Image Lab software (Bio-Rad, USA). The software calculates the percentage purity based on the integrated intensity of the target protein band relative to the total integrated intensity of all protein bands in the lane. The purified protein was subjected to endotoxin removal and subsequent quantification utilizing the Protein Endotoxin Removal Kit (C0268S, Beyotime, China) and the Endotoxin Detection Kit employing the Limulus Reagent Dynamic Turbidimetric Method (C0271S, Beyotime, China).
2.6. Development of the IELISA method
The purified fusion protein was utilized, and a standardized protocol was employed to develop the ELISA method [22].
The fusion protein was diluted to a concentration of 5 μg/mL using a carbonate buffer solution (CBS, pH 9.6) and subsequently added to a 96-well microplate (Corning, USA) at a volume of 100 μL per well. The microplate was incubated overnight at 4°C, followed by three washes with PBST, utilizing 300 μL per well for each wash. A blocking solution composed of 5% skimmed milk powder was then added at a volume of 300 μL per well, and the plate was incubated at 37°C for 2 hours. After an additional round of washing with PBST, serum samples were introduced at a dilution of 1:200 in PBS, with 100 μL added to each well, and incubated at 37°C for 1 hour. Following further washes with PBST, 100 µL of HRP-conjugated rabbit anti-human IgG (diluted 1:10,000, Thermo Fisher, USA) was added to each well and incubated at 37°C for 1 hour. The plate was washed three times with PBST, after which TMB substrate solution was added, and the plate was incubated in the dark for 10 minutes to allow for color development. The reaction was terminated by the addition of 2 M H2SO4, and the OD450 was measured using a microplate reader (Versa Max microplate reader, MD, USA). Laboratory-stored LPS (provided by the China Animal Health and Epidemiology Center, 3 mg/mL, extracted and purified from B. abortus A19) was utilized as a control antigen [23]. Serum samples were tested in triplicate following the same procedure. The sensitivity, specificity, area under the curve (AUC), and cutoff values were determined through receiver operating characteristic (ROC) curve analysis.
2.7. Evaluation of cross-reactivity via Indirect ELISA Method
The cross-reactivity of the fusion protein and LPS was assessed by testing them against the panel of sera from SAT-negative, non-brucellosis febrile patients (n = 283) described in section 2.1. A sample was considered cross-reactive if its OD450 value exceeded the cut-off value established by the ROC curve.
3. Results
3.1. Selection of B-cell epitopes
A comprehensive search conducted within the IEDB yielded a total of 23 epitopes associated with the following proteins: Cu-Zn superoxide dismutase (SOD), 31 kDa outer-membrane immunogenic protein precursor (Omp31), chaperone protein DnaK, Omp2b, 26 kDa periplasmic immunogenic protein precursor (BP26), and elongation factor Tu (EF-Tu) (S1 File). Through a process of merging overlapping sequences, these 23 initial epitopes were consolidated into 11 non-redundant, optimized epitopes (Table 1). Following the BLAST analysis, the selected epitopes, which exhibit a high degree of conservation in classical Brucella species, were found to be well conserved among the primary pathogenic species of B. melitensis and B. abortus. The percent identity observed is 100% (Table 1 and S4 File).
3.2. Construction and evaluation of the fusion protein
The fusion protein comprises a total of 342 amino acids. Using the ProtParam tool available on the ExPASy server, we calculated the aliphatic index and isoelectric point of the fusion protein. The aliphatic index is 67.02, and the instability index is computed to be 33.46, this classifies the fusion protein as stable (< 40 indicates stability). The grand average of hydropathicity (GRAVY) being -0.460 indicates that the protein may exhibit relatively good hydrophilic properties. The molecular weight of the constructed fusion protein was predicted to be approximately 35.1 kDa, and its theoretical isoelectric point (pI) was determined to be 4.57. The overall prediction for the protective antigen was calculated to be 0.9559, indicating that it is a probable antigen. The 3D structure of the protein is illustrated in Fig 1B.
(A) Amino acid sequence of the fusion protein; (B) 3D structural models of fusion proteins predicted by I-TASSER.
3.3. Preparation of fusion protein
The multiepitope fusion protein was successfully expressed in E. coli BL21(DE3). Solubility analysis revealed that the majority of the recombinant protein was present in the soluble fraction of the cell lysate (Fig 2), which facilitated subsequent purification. The protein was efficiently purified using Ni-NTA affinity chromatography under native conditions. As shown in Fig 2C, SDS-PAGE analysis of the purified product displayed a single predominant band at approximately 45 kDa. Analysis of the purified protein using Image Lab confirmed that the final purity exceeded 90%. After endotoxin removal, the endotoxin level of fusion protein was 0.084 EU/mL.
(A) Protein expression results in small amounts. M, Marker; Lane 1, uninduced control (BL21); Lane 2-3, IPTG induced (BL21). (B) Protein bulk expression results. M, Marker; Lane 1, whole bacteria after sonication; Lane 2, supernatant after sonication; Lane3, precipitate after sonication. (C) Protein purification results. M: Marker; Lane 1, purified protein.
3.4. Results of iELISA
Following the analysis of the ROC curve, the AUC for the fusion protein in serum was determined as the Youden index. The reactivity profiles of positive and negative serum samples with both antigens are presented as a dot plot in Fig 3A, showing clear separation between the groups. The fusion protein demonstrated excellent diagnostic accuracy against the negative control samples (n = 126), with the AUC of 0.9912 (95% CI: 0.9857 - 0.9968). At the optimal cut-off value (OD450 > 0.4861), the assay achieved a sensitivity of 95.34% and a specificity of 93.65%. For comparison, the LPS-based iELISA tested against the same negative control samples showed an AUC of 0.9958, with a sensitivity of 96.77% and a specificity of 98.41% at its cut-off (OD450 > 0.3185). The comprehensive evaluation metrics, including accuracy, positive predictive value (PPV), and negative predictive value (NPV), are summarized in Table 2, and raw data is presented in S2 File.
(A) Dot plot of human serum samples. Fusion Protein Positive: OD450 values of positive sera tested with the fusion protein. Fusion Protein Negative: OD450 values of negative sera from the negative control samples (n=126) tested with the fusion protein. LPS Positive: OD450 values of positive sera tested with LPS. LPS Negative: OD450 values of negative sera from the negative control samples (n=126) tested with LPS. (B) ROC analysis of human sera.
However, when we integrate the control samples (409 sera include 126 negative serum and 283 cross-reactive samples), For, the AUC of Fusion Protein is 0.9930, with a sensitivity of 99.64% and a specificity of 92.18%. For LPS, the AUC is 0.9807, with a sensitivity of 93.91% and a specificity of 94.38% (Table A and Fig D in S3 File).
3.5. Results of cross-reactivity
The cross-reactivity of the fusion protein and LPS was assessed by testing them against the panel of sera from SAT-negative patients (n = 283) with confirmed active infections by other pathogens (as defined in Methods 2.2). Utilizing iELISA and employing cut-off values established through ROC curve analysis, cross-reactivity with fusion protein was identified in 9 out of 283 serum samples (5 cases of Escherichia coli infection, 2 case of Klebsiella pneumoniae,1 case of Pseudomonas aeruginosa, and 1 case of Staphylococcus aureus). Conversely, cross-reactivity with LPS was identified in 41 out of 283 serum samples. The pathogens exhibiting cross-reactivity with LPS included 18 cases of Escherichia coli infection, 7 cases of Klebsiella pneumoniae, 4 cases of Pseudomonas aeruginosa, 3 cases of Enterococcus faecium, 2 cases each for Acinetobacter baumannii, Klebsiella oxytoca, Staphylococcus aureus and Streptococcus agalactiae, and one case of Serratia marcescens. The detailed results are presented in S2 File.
4. Discussion
Among the currently recognized species of Brucella, the six classical species are the primarily cause of human infections, which are predominantly associated with smooth Brucella species, such as Brucella abortus (B. abortus) and Brucella melitensis (B. melitensis) [24]. The key characteristic of smooth strains is the presence of a complete lipopolysaccharide (LPS) in their outer membrane, featuring an O-polysaccharide chain (O-antigen). This O-antigen is the primary target of the host’s antibody response and is thus the basis for most conventional serological tests. In contrast, infections caused by rough Brucella species, such as Brucella canis (B. canis), which lack the O-antigen due to genetic mutations, are relatively infrequent and are notoriously difficult to diagnose with standard O-antigen-dependent assays [25]. Therefore, serological diagnostic methods that utilize LPS as an antigen are inherently limited; they may fail to detect rough Brucella infections and are prone to cross-reactivity with other Gram-negative bacteria (e.g., Yersinia enterocolitica O9) that share similar O-antigen structures [6,7]. Hence, the identification of proteins that are present in both smooth and rough Brucella species may offer a viable solution to this diagnostic challenge [26]. Additionally, the ubiquitous presence of LPS in Gram-negative bacteria heightens the risk of misdiagnosis.
A considerable number of antigenic proteins have been utilized in serological diagnostic studies of brucellosis, including outer membrane proteins; however, the diagnostic efficacy of individual proteins is frequently suboptimal [27]. The fusion of multiple proteins appears to enhance the performance of the assay; nevertheless, the resultant molecular weight is frequently too large for effective recombinant expression. Therefore, early studies have employed bioinformatics tools to predict B-cell epitopes within outer membrane proteins for the tandem construction of multiepitope fusion proteins [10,28]. Although some preliminary results have been obtained, the construction of these fusion proteins has largely been serendipitous due to the inherent uncertainty associated with the predictions generated by bioinformatics tools, indicating significant potential for improvement [11,29–31]. In the present study, we selected B-cell epitopes cataloged in the IEDB for the tandem construction of multi-epitope fusion proteins, which were subsequently experimentally validated, thereby addressing the limitations associated with inaccurate bioinformatics predictions. The present study advances prior foundational research by introducing several novel elements that differentiate it from earlier work. First, experimentally validated B-cell epitopes obtained from the IEDB were employed to construct the fusion protein, thereby avoiding the uncertainties inherent in bioinformatics-based predictions and enhancing the reliability of epitope selection. Second, the fusion protein developed herein demonstrated superior diagnostic performance, achieving an AUC of 0.9912, sensitivity of 0.9534, and specificity of 0.9365, surpassing the metrics reported in previous studies. However, the LPS-based iELISA demonstrated marginally superior metrics. Furthermore, the fusion protein exhibited substantially reduced cross-reactivity, with only 9 of 283 non-Brucella-infected samples showing cross-reactivity, in contrast to 41 samples observed with LPS. This reduction is critical for minimizing false-positive results and improving diagnostic specificity. Additionally, the validation process was expanded to include a larger and more diverse cohort, encompassing 283 non-Brucella-infected samples, thereby providing more robust evidence for the diagnostic potential of the fusion protein and highlighting its advantages over conventional methods.
The most significant impact of our multiepitope fusion protein on the serological detection of brucellosis lies in its substantial reduction of false-positive results when discriminating brucellosis from other confirmed bacterial infections, a context where LPS-based tests are prone to cross-reactivity, particularly with Gram-negative bacteria. For clinicians and patients, a reduction in false positives means fewer individuals undergoing unnecessary and potentially lengthy antibiotic treatments, reducing associated side effects, healthcare costs, and patient anxiety. Furthermore, it prevents the misallocation of public health resources and allows for a more accurate investigation of true brucellosis outbreaks. However, we should note a limitation of our cross-reactivity assessment. All the cross-reactive sera were confirmed bacterial infection samples. In reality, patients with fever who require differentiation from brucellosis do not have exclusively bacterial infections; other causes include viral and fungal infections. While we tested the fusion protein against a panel of sera from patients with other common infections, the definitive serological status of these samples regarding their respective pathogens was not reconfirmed in our study. Future studies would benefit from using well-characterized serum panels with confirmed, high-titre antibodies against specific cross-reactive agents (e.g., Yersinia enterocolitica O9), providing an even more stringent evaluation of diagnostic specificity.
The 23 epitopes derived from six antigenic proteins within the selected IEDB collection were validated using monoclonal antibodies, these proteins include Cu-Zn SOD, Omp31, DnaK, Omp2b, BP26 and EF Tu [32–44], and the epitopes selected from the IEDB were chosen based on their reported immunogenicity and potential applicability in diagnostic contexts. On the other hand, it is important to note that the IEDB does not provide information regarding the cross-reactivity of these epitopes with other pathogens, a factor that is essential for ensuring the specificity of the diagnostic tool. While these epitopes are highly conserved within Brucella species, as indicated by BLAST analysis, certain epitopes are also found to be prevalent in other pathogens. For example, the epitope ‘QTREHIL’ exhibits 100% similarity with epitopes identified in Pseudomonas. Moreover, additional epitopes demonstrate a relatively high degree of homology with those present in Ochrobactrum. Following our integration and optimization efforts, we developed a fusion protein that incorporates 23 linear B-cell epitopes, as cataloged by the IEDB. The results of this study confirmed that the protein demonstrated effective performance in the diagnosis of brucellosis, with detection results surpassing those obtained using LPS. In contrast, our serum was classified solely based on serological method (SAT), the absence of culture or PCR confirmation is a limitation of our study, as these methods could provide additional validation of the true status of the serum samples. Future studies should consider incorporating these techniques to enhance the accuracy and reliability of the diagnostic results. While the fusion protein demonstrated high sensitivity (95.34%) in recognizing positive samples, its principal advantage over LPS lies in the significant reduction of cross-reactivity, as detailed in the results.
However, it is likely that the protein’s lack of purity contributed to its reduced effectiveness in identifying negative samples. The apparent MW of the purified fusion protein on SDS-PAGE was approximately 45 kDa, which is higher than the theoretical molecular weight of 35.1 kDa predicted from its amino acid sequence. This discrepancy is a common observation with recombinant proteins and can be attributed to several factors: Amino acid composition, the theoretical calculation assumes a uniform peptide-mass ratio, but unusual amino acid composition can affect how the protein binds SDS, altering its electrophoretic mobility. Physicochemical properties: the relatively acidic theoretical pI (4.57) of our fusion protein may influence its SDS-binding capacity and thus its migration. Protein structure: Although SDS-PAGE is run under denaturing conditions, some residual protein structure or the presence of the flexible ‘GGGS’ linkers might prevent the protein from being fully linearized, leading to anomalous migration. Post-translational modifications: While less common in prokaryotic systems, minor modifications or the presence of the N-terminal methionine can contribute to small shifts in apparent molecular weight. Regarding implementation and cost, the recombinant multiepitope protein offers practical advantages. The iELISA protocol we developed is identical to standard ELISA procedures, requiring no specialized equipment or training, which facilitates its seamless adoption in clinical and public health laboratories. While a precise cost-benefit analysis is beyond the scope of this study, the production of recombinant protein in E. coli is generally scalable, reproducible, and cost-effective. In contrast, the extraction and purification of native LPS from pathogenic Brucella strains require high-level biosafety containment (BSL-3), which is complex, hazardous, and expensive. Therefore, our recombinant antigen not only eliminates the biosafety risks but is also likely to be more economical and sustainable for large-scale production.
The specific clinical implications of our findings are twofold. Primarily, this fusion protein serves as a superior candidate antigen for developing next-generation commercial ELISA kits. However, whether it can serve as a safer and more specific alternative to LPS requires further validation through the collection of real-world samples. This is particularly crucial in regions where cross-reactive pathogens like Yersinia enterocolitica O9 are co-endemic. Secondly, its high specificity makes it an excellent confirmatory test following a sensitive screening test, or as a primary diagnostic tool in low-prevalence areas and reference laboratories, where the cost of a false positive is particularly high.
In addition, the results of the cross-reactivity assessment indicated that the fusion protein exhibited cross-reactivity with 9 out of 283 serum samples from bacterial infections. In contrast, LPS showed cross-reactivity with 41 samples, predominantly associated with Escherichia coli infections, which may be attributed to the widespread presence of LPS in Gram-negative bacteria [6,7]. It is crucial to acknowledge that the lack of validation concerning the specificity and coverage of the selected epitopes represents a limitation of our study. While the epitopes were sourced from the IEDB, which contains experimentally validated data, we did not perform individual verification of each epitope’s reactivity and coverage across various Brucella species. Furthermore, the Brucella species infecting the serum donors was not determined, which precluded an assessment of the fusion protein’s ability to distinguish between species such as B. melitensis and B. abortus. Validating each epitope would require additional experiments, including the synthesis of peptides and their conjugation to carriers; processes that are both complex and financially burdensome, exceeding our current funding capabilities. Future research involving larger sample sizes and a more diverse representation of Brucella species is essential to thoroughly assess the diagnostic potential of the developed fusion protein.
The multiepitope sequence developed here provides a robust foundation that can be further refined in future iterations. Firstly, the current construct is limited to linear B-cell epitopes expressible in a prokaryotic system. Incorporating conformational epitopes or T-cell epitopes, potentially using eukaryotic expression systems, could enhance its immunogenicity and diagnostic breadth. Secondly, as our epitopes are highly conserved in classical Brucella species, this construct is a powerful broad-spectrum tool. However, for species-specific diagnosis (e.g., to distinguish between B. melitensis and B. abortus), future designs could integrate epitopes unique to specific species. Finally, computational tools for predicting and minimizing homology with human proteins and other common pathogens could be employed to further reduce the risk of autoimmunity or residual cross-reactivity.
Our experiments confirmed the successful construction of a multicellular epitope fusion protein, which exhibits considerable potential for practical applications. A fundamental difference between this study and our previous research resides in the epitope selection methodology. Whereas earlier diagnostic fusion proteins were designed based on epitopes predicted through bioinformatics tools, the present construct is exclusively composed of experimentally validated linear B-cell epitopes obtained from the IEDB. This strategy reduces the uncertainty associated with computational predictions and is grounded in established empirical evidence. Nevertheless, several limitations were identified in our study. The clinical validation of our fusion protein, while promising, was conducted on a cohort where the infecting Brucella species was not identified. Future studies involving larger, well-characterized sample sets, including infections caused by rough Brucella species such as B. canis, will be essential to fully delineate the diagnostic scope and versatility of this assay. Also, it remains unclear whether the patients included in our study were experiencing acute or chronic infections, making it impossible to ascertain whether the protein facilitates the early diagnosis of brucellosis. Future research should prioritize increasing the sample size and utilizing random sampling methods to validate the application potential of the fusion protein.
5. Conclusions
In this study, we successfully developed a novel multiepitope fusion protein for the serodiagnosis of brucellosis by integrating experimentally validated linear B-cell epitopes from the IEDB. The fusion protein demonstrated excellent diagnostic performance in iELISA, rivaling that of conventional LPS-based assays. Its most significant advancement lies in a substantial reduction in cross-reactivity with bacteremia infections, effectively addressing a critical flaw of current standard methods. By providing a safer, more specific, and scalable antigen alternative to LPS, this fusion protein represents a promising candidate for next-generation brucellosis diagnostics. Future validation with larger, well-characterized cohorts will be essential to fully translate this robust diagnostic potential into clinical practice.
Supporting information
S1 File. Epitope table export results from IEDB.
https://doi.org/10.1371/journal.pntd.0013804.s001
(XLSX)
S3 File. Original images for Fig 2 and I-ELISA results of human serum samples (Integrate 126 negative serum and 283 cross-reactive samples).
https://doi.org/10.1371/journal.pntd.0013804.s003
(DOCX)
Acknowledgments
We thank the China Animal Health and Epidemiology Center for the gift of LPS and Brucella strains, and the Xuzhou Center for Disease Control and Prevention for the gift of human brucellosis sera.
References
- 1. Laine CG, Johnson VE, Scott HM, Arenas-Gamboa AM. Global Estimate of Human Brucellosis Incidence. Emerg Infect Dis. 2023;29(9):1789–97. pmid:37610167
- 2. Bosilkovski M, Siskova D, Spasovska K, Vidinic I, Dimzova M. The influence of illness duration before diagnosis on clinical characteristics and outcome in human brucellosis. Trop Doct. 2019;49(3):177–81. pmid:31060447
- 3. Qureshi KA, Parvez A, Fahmy NA, Abdel Hady BH, Kumar S, Ganguly A, et al. Brucellosis: epidemiology, pathogenesis, diagnosis and treatment-a comprehensive review. Ann Med. 2023;55(2):2295398. pmid:38165919
- 4. Yagupsky P, Morata P, Colmenero JD. Laboratory Diagnosis of Human Brucellosis. Clin Microbiol Rev. 2019;33(1):e00073-19.
- 5. Freire ML, Machado de Assis TS, Silva SN, Cota G. Diagnosis of human brucellosis: Systematic review and meta-analysis. PLoS Negl Trop Dis. 2024;18(3):e0012030. pmid:38452046
- 6. Bonfini B, Chiarenza G, Paci V, Sacchini F, Salini R, Vesco G, et al. Cross-reactivity in serological tests for brucellosis: a comparison of immune response of Escherichia coli O157:H7 and Yersinia enterocolitica O:9 vs Brucella spp. Vet Ital. 2018;54(2):107–14. pmid:30019327
- 7. Chart H, Okubadejo OA, Rowe B. The serological relationship between Escherichia coli O157 and Yersinia enterocolitica O9 using sera from patients with brucellosis. Epidemiol Infect. 1992;108(1):77–85. pmid:1547843
- 8. Rahimnahal S, Yousefizadeh S, Mohammadi Y. Novel multi-epitope vaccine against bovine brucellosis: approach from immunoinformatics to expression. J Biomol Struct Dyn. 2023;41(24):15460–84. pmid:36927475
- 9. Saadi M, Karkhah A, Nouri HR. Development of a multi-epitope peptide vaccine inducing robust T cell responses against brucellosis using immunoinformatics based approaches. Infect Genet Evol. 2017;51:227–34. pmid:28411163
- 10. Yin D, Li L, Song X, Li H, Wang J, Ju W, et al. A novel multi-epitope recombined protein for diagnosis of human brucellosis. BMC Infect Dis. 2016;16:219. pmid:27206475
- 11. Yin D, Bai Q, Wu X, Li H, Shao J, Sun M, et al. Paper-based ELISA diagnosis technology for human brucellosis based on a multiepitope fusion protein. PLoS Negl Trop Dis. 2021;15(8):e0009695. pmid:34403421
- 12. Cardoso PG, Macedo GC, Azevedo V, Oliveira SC. Brucella spp noncanonical LPS: structure, biosynthesis, and interaction with host immune system. Microb Cell Fact. 2006;5:13. pmid:16556309
- 13. Ramana J, Mehla K. Immunoinformatics and epitope prediction. Methods Mol Biol. 2020;2131:155–71.
- 14. Raoufi E, Hemmati M, Eftekhari S, Khaksaran K, Mahmodi Z, Farajollahi MM, et al. Epitope Prediction by Novel Immunoinformatics Approach: A State-of-the-art Review. Int J Pept Res Ther. 2020;26(2):1155–63. pmid:32435171
- 15. Saha S, Raghava GPS. Prediction of continuous B-cell epitopes in an antigen using recurrent neural network. Proteins. 2006;65(1):40–8. pmid:16894596
- 16. Yao B, Zheng D, Liang S, Zhang C. SVMTriP: A Method to Predict B-Cell Linear Antigenic Epitopes. Methods Mol Biol. 2020;2131:299–307. pmid:32162263
- 17. Lupu L, Kleinekofort W, Morgner N. Epitope characterization of proteins and aptamers with mass spectrometry. Eur J Mass Spectrom (Chichester). 2023;29(5–6):359–69. pmid:37957929
- 18. McMaster M, Mohr K, Page A, Closmore A, Towne F, Brooks BD. Epitope characterization of anti-drug antibodies-a tool for discovery and health: an overview of the necessity of early epitope characterization to avoid anti-drug antibodies and promote patient health. Expert Opin Biol Ther. 2021;21(6):705–15. pmid:33317351
- 19. Chronister W, Sette A, Peters B. Epitope-Specific T Cell Receptor Data and Tools in the Immune Epitope Database. Methods Mol Biol. 2022;2574:267–80. pmid:36087207
- 20. Mahajan S, Vita R, Shackelford D, Lane J, Schulten V, Zarebski L, et al. Epitope Specific Antibodies and T Cell Receptors in the Immune Epitope Database. Front Immunol. 2018;9:2688. pmid:30515166
- 21. Vita R, Blazeska N, Marrama D, IEDB Curation Team Members, Duesing S, Bennett J, et al. The Immune Epitope Database (IEDB): 2024 update. Nucleic Acids Res. 2025;53(D1):D436–43. pmid:39558162
- 22. Wu Q, Yuan L, Guo X, Sun M, Yao M, Yin D. Study on antigenic protein Omp2b in combination with Omp31 and BP26 for serological detection of human brucellosis. J Microbiol Methods. 2023;205:106663. pmid:36592896
- 23. Xie Y, Lin S, Guo L, Qi X, Zhao S, Pei Q, et al. Development and evaluation of the recombinant BP26 protein-based C-ELISA for human brucellosis diagnosis. Front Microbiol. 2025;15:1516915. pmid:39831112
- 24. Deng Y, Liu X, Duan K, Peng Q. Research Progress on Brucellosis. Curr Med Chem. 2019;26(30):5598–608.
- 25. Akinyemi KO, Fakorede CO, Amisu KO, Wareth G. Human and animal brucellosis in Nigeria: a systemic review and meta-analysis in the last twenty-one years (2001-2021). Vet Sci. 2022;9(8):384.
- 26. Bulashev A, Eskendirova S. Brucellosis detection and the role of Brucella spp. cell wall proteins. Vet World. 2023;16(7):1390–9. pmid:37621538
- 27. Bai Q, Li H, Wu X, Shao J, Sun M, Yin D. Comparative analysis of the main outer membrane proteins of Brucella in the diagnosis of brucellosis. Biochem Biophys Res Commun. 2021;560:126–31. pmid:33989903
- 28. Yao M, Liu M, Chen X, Li J, Li Y, Wei YR, et al. Comparison of BP26, Omp25 and Omp31 and a Multiepitope-Based Fusion Protein in the Serological Detection of Canine Brucellosis. Infect Drug Resist. 2022;15:5301–8. pmid:36101776
- 29. Xie Y, Guo L, Qi X, Zhao S, Pei Q, Chen Y, et al. Establishment of an I-ELISA method based on multi-epitope fusion protein for diagnosis of human brucellosis. PLoS Negl Trop Dis. 2025;19(4):e0012995. pmid:40193524
- 30. Wang G, Qi X, Zhao S, Pei Q, Chen Y, Yin D, et al. Preparation of a Brucella multiepitope fusion protein based on bioinformatics and its application in serological diagnosis of human brucellosis. Sci Rep. 2025;15(1):19106. pmid:40447804
- 31. Wu Q, Yuan Y, Guo L, Xie Y, Yao M, Yin D. Preparation and application of a Brucella multiepitope fusion protein based on bioinformatics and Tandem Mass Tag-based proteomics technology. Front Immunol. 2025;15:1509534. pmid:39867886
- 32. Tabatabai LB, Pugh GW Jr. Modulation of immune responses in Balb/c mice vaccinated with Brucella abortus Cu-Zn superoxide dismutase synthetic peptide vaccine. Vaccine. 1994;12(10):919–24. pmid:7526568
- 33. Cassataro J, Estein SM, Pasquevich KA, Velikovsky CA, de la Barrera S, Bowden R, et al. Vaccination with the recombinant Brucella outer membrane protein 31 or a derived 27-amino-acid synthetic peptide elicits a CD4+ T helper 1 response that protects against Brucella melitensis infection. Infect Immun. 2005;73(12):8079–88. pmid:16299302
- 34. Cassataro J, Pasquevich KA, Estein SM, Laplagne DA, Velikovsky CA, de la Barrera S, et al. A recombinant subunit vaccine based on the insertion of 27 amino acids from Omp31 to the N-terminus of BLS induced a similar degree of protection against B. ovis than Rev.1 vaccination. Vaccine. 2007;25(22):4437–46. pmid:17442465
- 35. Li J, Hu F, Chen S, Luo P, He Z, Wang W, et al. Characterization of novel Omp31 antigenic epitopes of Brucella melitensis by monoclonal antibodies. BMC Microbiol. 2017;17(1):115. pmid:28506316
- 36. Zhang F, Li Z, Jia B, Zhu Y, Pang P, Zhang C, et al. The Immunogenicity of OMP31 Peptides and Its Protection Against Brucella melitensis Infection in Mice. Sci Rep. 2019;9(1):3512. pmid:30837598
- 37. Vizcaíno N, Kittelberger R, Cloeckaert A, Marín CM, Fernández-Lago L. Minor nucleotide substitutions in the omp31 gene of Brucella ovis result in antigenic differences in the major outer membrane protein that it encodes compared to those of the other Brucella species. Infect Immun. 2001;69(11):7020–8. pmid:11598077
- 38. Vizcaíno N, Cloeckaert A, Zygmunt MS, Dubray G. Cloning, nucleotide sequence, and expression of the Brucella melitensis omp31 gene coding for an immunogenic major outer membrane protein. Infect Immun. 1996;64(9):3744–51. pmid:8751924
- 39. Vizcaíno N, Zygmunt MS, Verger JM, Grayon M, Cloeckaert A. Localization and characterization of a specific linear epitope of the Brucella DnaK protein. FEMS Microbiol Lett. 1997;154(1):117–22. pmid:9297829
- 40. Paquet JY, Diaz MA, Genevrois S, Grayon M, Verger JM, de Bolle X, et al. Molecular, antigenic, and functional analyses of Omp2b porin size variants of Brucella spp. J Bacteriol. 2001;183(16):4839–47. pmid:11466287
- 41. Seco-Mediavilla P, Verger J-M, Grayon M, Cloeckaert A, Marín CM, Zygmunt MS, et al. Epitope mapping of the Brucella melitensis BP26 immunogenic protein: usefulness for diagnosis of sheep brucellosis. Clin Diagn Lab Immunol. 2003;10(4):647–51. pmid:12853399
- 42. Qiu J, Wang W, Wu J, Zhang H, Wang Y, Qiao J, et al. Characterization of periplasmic protein BP26 epitopes of Brucella melitensis reacting with murine monoclonal and sheep antibodies. PLoS One. 2012;7(3):e34246. pmid:22457830
- 43. Guo X, Sun M, Guo Y, Wu Y, Yan X, Liu M, et al. Production and evaluation of anti-BP26 monoclonal antibodies for the serological detection of animal brucellosis. Front Vet Sci. 2024;11:1389728. pmid:38957801
- 44. Zhao N, Jiang Y, Ming S, Liu S, Zhao X, Wang F. Monoclonal Antibody Preparation and Epitope Identification for Brucella melitensis Elongation Factor Tu. Front Microbiol. 2019;10:1878. pmid:31474961