Employing computational tools to design a multi-epitope vaccine targeting human immunodeficiency virus-1 (HIV-1)

Abstract

Background: Despite being in the 21^st century, the world has still not been able to vanquish the global AIDS epidemic, and the only foreseeable solution seems to be a safe and effective vaccine. Unfortunately, vaccine trials so far have returned unfruitful results, possibly due to their inability to induce effective cellular, humoral and innate immune responses. The current study aims to tackle these limitations and propose the desired vaccine utilizing immunoinformatic approaches that have returned promising results in designing vaccines against various rapidly mutating organisms. For this, all polyprotein and protein sequences of HIV-1 were retrieved from the LANL (Los Alamos National Laboratory) database. The consensus sequence was generated after alignment and used to predict epitopes. Conserved, antigenic, non-allergenic, T-cell inducing, B-cell inducing, IFN-ɣ inducing, non-human homologous epitopes were selected and combined to propose two vaccine constructs i.e., HIV-1a (without adjuvant) and HIV-1b (with adjuvant).

Results: HIV-1a and HIV-1b were subjected to antigenicity, allergenicity, structural quality analysis, immune simulations, and MD (molecular dynamics) simulations. Both proposed multi-epitope vaccines were found to be antigenic, non-allergenic, stable, and induce cellular, humoral, and innate immune responses. TLR-3 docking and in-silico cloning of both constructs were also performed.

Conclusion: Our results indicate HIV-1b to be more promising than HIV-1a; experimental validations can confirm the efficacy and safety of both constructs and in-vivo efficacy in animal models.

Keywords: Acquired immunodeficiency syndrome; Bioinformatics; Computational biology; Human immunodeficiency virus; Immunity; Immuno-informatics; Toll like receptor-3; Vaccinology.

Fig. 1

Immune simulations results of HIV-1a as obtained from C-IMMSIM. A Production of various types of Igs and immunocomplexes formation induced by antigen. B B cell population. C T-helper cell population per state. D Total TH cell population. E Cytotoxic T cell population per state. F TC cell population. G Elevated levels of interleukins and cytokines. Inset plot shows high production of IL-2 along with danger signal

Fig. 2

Immune simulations results of HIV-1b as obtained from C-IMMSIM. A Production of various types of Igs and immunocomplexes formation induced by antigen. B B cell population. C T-helper cell population per state. D Total TH cell population. E Cytotoxic T cell population per state. F TC cell population. G Elevated levels of interleukins and cytokines. Inset plot shows high production of IL-2 along with danger signal

Fig. 3

Predicted tertiary structure, Ramachandran plot and model quality analysis of HIV-1a. A Refined structure. B Ramachandran plot of the refined structure. C Overall model quality showed that protein structure lies in the allowed region. D Local model quality through a window size of 40 amino acids also showed that most part of the protein falls under the knowledge-based energy

Fig. 4

Predicted tertiary structure, Ramachandran plot and model quality analysis of HIV-1b. A Refined structure. B Ramachandran plot of the refined structure. C Overall model quality showed that protein structure lies in the allowed region. D Local model quality through a window size of 40 amino acids also showed that most part of the protein falls under the knowledge-based energy

Fig. 5

RMSD, Rg, RMSF, and partial density of HIV-1a. A The RMSD graph has the peak around 17,000 and 29,000 ps of almost 0.9 nm and 0.85 nm respectively, which is still much less and still in the required zone of 1.0 nm. The overall graph shows the stability, with an average RMSD value of almost 0.65 nm. B Radius of gyration shows a little peak at the start but has overall stable Rg value and shows that the protein is stably folded. C RMSF doesn’t show much deviation as per atom, which determines that the structure is stable. D Protein density graph is smooth and stable as per the graph for HIV-1a

Fig. 6

RMSD, Rg, RMSF, and partial density of HIV-1b. A The RMSD graph is almost stable around 0.8 nm and has no recognizable peak. Although the value is a bit higher as compared to HIV1a but is still less than 1.0 nm. The overall graph shows the stability, with an average RMSD value of almost 0.8 nm. B Radius of gyration shows a little peak at the start but has overall stable Rg value and shows that the protein is stably folded. C RMSF shows some peaks, suggesting that the protein structure has flexibility in some areas but still is in quite acceptable range, which determines that the structure is almost stable. D Protein density graph is smooth and stable as per the graph for HIV-1b

Fig. 7

Secondary structure of (A) HIV-1a, and (B) HIV-1b

Fig. 8

HIV-1a-TLR-3 complex structure and interaction details. A HIV-1a-TLR-3 complex refined structure. B Summary of HIV-1a-TLR-3 interactions. C Interacting residues of HIV-1a and TLR-3. HIV-1a forms 8 salt bridges, 21 hydrogen bonds, and 179 non-bonded contacts with TLR-3. D Eigen value of HIV-1a-TLR-3 complex

Fig. 9

HIV-1b-TLR-3 complex structure and interaction details. A HIV-1b-TLR-3 complex refined structure. B Summary of HIV-1b-TLR-3 interactions. C Interacting residues of HIV-1b and TLR-3. HIV-1b forms 5 salt bridges, 11 hydrogen bonds, and 100 non-bonded contacts with TLR-3. D Eigen value of HIV-1b-TLR-3 complex

Fig. 10

Cloned (A) HIV-1a, and (B) HIV-1b in pET28(a) + plasmid. CAI value of both multi-epitope vaccines suggest high expression levels in E. Coli system

Fig. 11

Flowchart of the methodology followed in this study