…

The burden of brucellosis remains one of the most underappreciated problems in global infectious disease control. With an estimated 2.1 million human cases each year, the disease persists wherever close contact between people and livestock occurs, and where unpasteurized dairy products remain part of daily diets. The causative bacteria, members of the genus Brucella, are facultative intracellular pathogens, meaning they survive and replicate inside host cells, sheltered from many components of the immune response. This intracellular lifestyle is precisely what has frustrated traditional vaccine development for human use. Live attenuated vaccines that work reasonably well in animals, such as strains used to immunize cattle, sheep, and goats, are generally considered too risky for humans because even weakened Brucella organisms can cause disease in people with compromised immune systems. Subunit vaccines, which use only purified pieces of the pathogen rather than living organisms, sidestep this safety concern but historically have struggled to provoke immunity strong enough to clear an intracellular bacterium. The study by Zhang and colleagues addresses this central tension by combining the safety of a subunit approach with computational strategies designed to maximize immunogenicity.

Reverse vaccinology, the overarching strategy employed here, represents a fundamental shift from classical vaccine discovery. In the traditional paradigm, researchers identified vaccine candidates by growing pathogens in the laboratory and testing individual proteins one at a time, a slow process that favored abundant surface antigens. Reverse vaccinology instead begins with the pathogen’s genome sequence. Bioinformatic algorithms scan the predicted proteins of an organism, evaluating each for surface localization, conservation across strains, and the presence of short peptide segments that immune cells can recognize. The approach became feasible once complete bacterial genome sequences were available, and it has since been applied to pathogens ranging from group B meningococcus to various intracellular bacteria. For Brucella, whose genome encodes thousands of proteins, computational triage allows researchers to focus experimental resources on a handful of promising antigens rather than screening exhaustively in the laboratory.

The two antigens selected in this work, Omp28 and GroEL, were chosen for complementary reasons. Omp28 is an outer membrane protein, a class of molecules positioned at the interface between the bacterium and its environment. Outer membrane proteins are attractive vaccine targets because antibodies and immune cells encounter them early during infection, and because they often play essential roles in nutrient uptake and structural integrity, making it difficult for the bacterium to alter them without fitness costs. GroEL, by contrast, is a molecular chaperone located in the cytoplasm, part of the heat shock protein family that helps other proteins fold correctly under stress. Its inclusion may seem counterintuitive for a surface-targeted vaccine, but GroEL is highly conserved among Brucella species and is known to be released or presented during infection, where it can stimulate both antibody production and cellular immune responses. Prior studies have identified GroEL fragments as immunodominant in infected animals, supporting its role here as a source of T cell epitopes. Combining an outer membrane antigen with a conserved intracellular chaperone broadens the repertoire of immune recognition the vaccine can engage.

The epitope counts reported in the study illustrate the depth of the computational screen. Sixteen cytotoxic T lymphocyte epitopes, ten helper T lymphocyte epitopes, and three B cell epitopes were derived from these two proteins. Each category corresponds to a different arm of adaptive immunity. Cytotoxic T lymphocytes recognize peptide fragments displayed on major histocompatibility complex class I molecules, a presentation pathway used by virtually all nucleated cells to signal internal contents, including the contents of cells harboring intracellular bacteria. When cytotoxic T cells detect foreign peptides in this context, they kill the infected cell, denying the pathogen its replication niche. Helper T lymphocytes, recognizing peptides on class II molecules expressed by professional antigen-presenting cells, orchestrate the response by secreting cytokines that activate macrophages, support B cells, and license cytotoxic T cells. Because Brucella survives within macrophages, the activation of these cells by helper T cell cytokines such as interferon-gamma is considered the decisive mechanism for bacterial clearance in natural infection. B cell epitopes, meanwhile, are the surface features of proteins that antibodies bind directly, contributing to opsonization and complement-mediated killing of bacteria circulating outside cells. A vaccine that engages all three arms simultaneously is more likely to reproduce the protective immunity seen after natural infection or live vaccination, without the attendant risks.

The dendritic cell-targeting peptide fused to the epitope ensemble is the design element that most directly distinguishes this construct from earlier multi-epitope vaccines. Dendritic cells are the sentinels of the immune system, stationed in tissues and mucosal surfaces where pathogens first enter. They capture antigen, migrate to draining lymph nodes, and present processed peptides to naive T cells, effectively deciding whether and how an immune response begins. The efficiency of this presentation step strongly influences the magnitude and quality of the resulting response. Targeting peptides, short amino acid sequences that bind receptors preferentially expressed on dendritic cells, have been used in experimental cancer vaccines and antiviral constructs to concentrate delivered antigen in these professional presenters. By fusing such a peptide to the Brucella epitopes, the designers aimed to raise the local concentration of vaccine antigen inside dendritic cells, thereby enhancing both class I and class II presentation pathways. This strategy is particularly relevant for intracellular pathogens, where the class I pathway and cross-presentation, the process by which dendritic cells display ingested antigens on class I molecules, are critical for generating cytotoxic T cell responses.

Before any laboratory work began, the construct was subjected to a battery of predictive evaluations covering stability, hydrophilicity, toxicity, and allergenicity. These filters matter for practical reasons. A vaccine protein that aggregates or degrades rapidly is difficult to manufacture and formulate. Toxicity predictions guard against constructs whose peptides might damage host tissues or interfere with normal physiology. Allergenicity screening is especially important because vaccines are administered to healthy populations, and any propensity to provoke IgE-mediated reactions would be disqualifying. The reported profile of the construct, described as stable, hydrophilic, non-toxic, and non-allergenic, suggests the design passed these standard computational gates, though such predictions always require experimental confirmation.

Molecular docking and molecular dynamics simulations provided the next layer of evidence. Docking algorithms predict how two molecules, in this case the vaccine construct and its immune receptors, fit together, scoring the shape complementarity and interaction energy of the predicted complex. The receptors examined were major histocompatibility complex molecules and Toll-like receptor 4. TLR4 is a pattern recognition receptor on dendritic cells and macrophages that recognizes bacterial lipopolysaccharide and certain other microbial products, triggering the innate signaling cascade that matures dendritic cells and licenses them to activate T cells. Including TLR4 engagement in the design implies an intent to couple antigen delivery with adjuvant-like signaling, so that the vaccine not only supplies epitopes but also provides the danger signals dendritic cells require to abandon their resting state. Molecular dynamics simulations then tested whether the docked complexes remain stable over simulated time, monitoring parameters such as root mean square deviation of atomic positions, radius of gyration, and hydrogen bond persistence. Stable trajectories across these metrics indicate that the predicted interactions are not fleeting artifacts of the docking algorithm but represent physically plausible, durable binding.

Immune simulation using the C-ImmSim platform extended the computational analysis from structural interactions to population dynamics. C-ImmSim models the interacting populations of immune cells and molecules over time, generating simulated curves for antibody titers, T cell expansion, and cytokine concentrations after virtual immunization. Such simulations cannot replace animal studies, but they serve as an inexpensive triage tool, flagging constructs whose predicted immune dynamics appear weak or skewed before laboratory resources are committed. The robust simulated responses reported for this construct supported proceeding to the experimental phase.

The laboratory work followed a well-established pipeline for recombinant protein production. The gene encoding the vaccine construct was synthesized and inserted into the plasmid pET-28a(+), a vector widely used for expressing proteins in Escherichia coli. This plasmid places the insert under control of a promoter induced by isopropyl beta-D-thiogalactoside, allowing precise regulation of expression, and it adds affinity tags that simplify purification by immobilized metal affinity chromatography. Expression in E. coli offers speed and low cost, though it carries the caveat that bacterial cytoplasm lacks the machinery for many eukaryotic post-translational modifications. For a vaccine built from bacterial protein sequences, this limitation is minor, since the target antigens are themselves bacterial and require no mammalian-specific modifications.

The in vitro validation experiments examined three complementary readouts. Indirect enzyme-linked immunosorbent assay measured antibody binding to the vaccine protein, confirming that the B cell epitopes embedded in the construct are accessible on the folded protein surface and that sera recognize the recombinant product. Cytometric bead array assay quantified a panel of cytokines in culture supernatants, providing a multiplexed picture of the inflammatory and helper signals elicited when immune cells encounter the vaccine. Flow cytometry characterized cell populations directly, distinguishing which cell types responded and measuring activation markers on their surfaces. The statistical significance reported across these assays, at P less than 0.01, indicates that the differences between vaccine-treated and control samples were unlikely to arise from random variation, though the magnitude and biological relevance of the effects depend on the specific values detailed in the full results.

Collectively, the findings position this construct as a candidate that activates dendritic cells and improves antigen presentation efficiency, the mechanistic claims that flow cytometry and cytokine measurements were designed to test. The convergence of computational predictions and multiple independent in vitro assays strengthens confidence in the construct’s immunogenicity, even though the critical questions of protection and safety in a living host remain ahead. Animal challenge studies, in which vaccinated animals are exposed to virulent Brucella and monitored for bacterial clearance and disease signs, will be the decisive test of prophylactic potential. Formulation and delivery route, whether the protein is admixed with adjuvants, encapsulated in particles, or administered mucosally, will also shape the response in ways that in vitro culture cannot fully anticipate.

Beyond brucellosis specifically, the study offers a transferable blueprint. The workflow, antigen selection by reverse vaccinology, epitope mapping across class I, class II, and B cell recognition, fusion with a dendritic cell-targeting peptide, computational validation by docking, dynamics, and immune simulation, followed by recombinant expression and in vitro immunogenicity testing, applies in principle to any intracellular bacterial pathogen for which a genome sequence is available. Pathogens such as Listeria monocytogenes, Mycobacterium tuberculosis, and Salmonella enterica share with Brucella the defining challenge of intracellular survival, and multi-epitope designs that engage cellular immunity while targeting professional antigen-presenting cells address that challenge directly. The open-access publication of the work, with supplementary data files provided, facilitates replication and extension by other groups, an important consideration for a field whose progress depends on iterative refinement of shared computational tools and experimental protocols.

It is worth noting the epidemiological setting in which this research was conducted. Xinjiang, in northwestern China, lies within a region where brucellosis remains endemic among livestock and where human cases continue to occur through occupational and foodborne exposure. Research groups embedded in high-incidence regions bring practical urgency to vaccine development, since the populations they serve face ongoing transmission that livestock vaccination programs have not fully eliminated. Human brucellosis, when untreated or treated late, can progress to chronic forms involving bones, joints, and the nervous system, imposing long-term disability and substantial healthcare costs. A safe, effective human vaccine would complement existing control measures, protecting the groups at highest risk, including herders, veterinarians, abattoir workers, and laboratory personnel, for whom repeated exposure is an unavoidable feature of their work.

The study also exemplifies the growing integration of immunoinformatics into mainstream vaccine research. The tools employed, epitope prediction algorithms, structural modeling servers, docking engines, dynamics simulators, and immune population models, have matured to the point where a complete computational design cycle can be executed in months rather than years. Their predictions are not infallible, and the discrepancy between simulated and real immune responses remains a recognized limitation of the field. Nevertheless, when computational triage is paired with rigorous in vitro validation, as in this study, the probability that laboratory and animal resources are invested in a genuinely promising candidate increases substantially. The convergence reported here, in which structural stability, receptor binding, simulated immune dynamics, and measured dendritic cell activation all pointed in the same direction, illustrates the value of that paired approach and suggests the construct merits continued development through the preclinical stages that separate a validated in vitro candidate from a vaccine ready for human trials.

Subject of Research: …

Article Title: Immunoinformatics design and in vitro evaluation of dendritic cell-targeting brucellosis multi-epitope vaccine based on Omp28 and GroEL

Article References: Zhang, X., Chen, N., Shi, J., Luo, J., Chai, Z., Li, C., Zhang, S., Pang, N., Zhu, Y., & Zhang, F. (2026). Immunoinformatics design and in vitro evaluation of dendritic cell-targeting brucellosis multi-epitope vaccine based on Omp28 and GroEL. Scientific Reports. https://doi.org/10.1038/s41598-026-74212-0

Image Credits: AI Generated

DOI: 10.1038/s41598-026-74212-0

Keywords: Immunoinformatics, design, vitro, evaluation, dendritic, cell-targeting, brucellosis, multi-epitope, vaccine, based, Omp28, GroEL