Elizabethkingia anophelis is described as a multidrug-resistant opportunistic pathogen associated with severe neonatal meningitis, sepsis, and hospital outbreaks. The pathogen carries high mortality and presents limited therapeutic options, motivating a prophylactic approach. The authors applied a reverse vaccinology and immunoinformatics workflow to develop a hybrid multi-epitope vaccine (MEV) and parallel mRNA construct as a preventive strategy against E. anophelis infections.
The study predicted antigenic epitopes for B cells and T cells (both MHC-I and MHC-II) from four outer membrane and secretion-associated proteins of E. anophelis. Candidate epitopes underwent stepwise in silico screening for antigenicity, allergenicity, toxicity, and cytokine induction potential. Selections were further guided by binding affinity metrics and global population coverage considerations to prioritize epitopes likely to elicit broad and safe immune responses.
The final MEV construct incorporated eight MHC-I epitopes, eight MHC-II epitopes, and eight B-cell epitopes. Epitopes were joined using appropriate linker sequences, the vaccine was adjuvanted with Human Beta Defensin-3 to enhance immune stimulation, and a 6 × His affinity tag was included for downstream purification or detection purposes. This hybrid design combined peptide-based epitope assembly with an adjuvant and purification tag in a single recombinant construct.
Population coverage analysis reported a global coverage of 99.38% for the selected T-cell epitopes, indicating that the epitope set potentially addresses diverse HLA allele distributions. In silico immune simulations predicted a robust Th1-biased response, prominent antibody production with IgG class switching, and development of long-term memory B and T cell populations, outcomes consistent with effective vaccine-induced adaptive immunity.
A three-dimensional model of the MEV was generated using AlphaFold2 and subsequently refined. Quality metrics reported include 98.0% Ramachandran favored residues, an ERRAT score of 98.621, and a ProSA Z-score of −4.68. These metrics were presented as evidence of a high-quality, well-refined structural model suitable for downstream receptor interaction studies and stability assessments.
Molecular docking experiments showed strong binding between the MEV construct and TLR-2, a pattern recognition receptor relevant to innate immune activation. To test the durability of the predicted interaction, the vaccine–receptor complex underwent 200 ns molecular dynamics simulations, which the authors reported as confirming the stability of the complex over the simulation interval.
Immune simulations performed in silico forecasted a Th1-skewed immune profile, consistent with cellular immunity supportive of intracellular pathogen control. Simulations also predicted high titers with IgG class switching and the formation of long-term memory B and T cells, suggesting the construct could promote both humoral and cellular adaptive responses if validated experimentally.
For recombinant protein production, the nucleotide sequence encoding the MEV was codon-optimized for Escherichia coli K12; the reported codon adaptation index (CAI) was 0.982, indicating strong compatibility with the expression host. The optimized sequence was in silico cloned into the pET-28a(+) expression vector to support efficient recombinant expression workflows. Parallel mRNA construct analysis included secondary structure assessment, which the authors reported as indicating high mRNA stability for potential mRNA-based delivery.
The computational pipeline produced a multi-epitope vaccine candidate and an mRNA construct with favorable in silico immunogenicity, structural quality, receptor binding, stability in molecular dynamics, high predicted population coverage, and expression-compatibility metrics. The authors emphasize that these results are computational and conclude that experimental validation is required to confirm immunogenicity, safety, and protective efficacy in vitro and in vivo. Specific experimental methods, timelines, or results beyond the in silico analyses were not reported in the source.