Immunoinformatics-Peptide Driven Vaccine and In silico Modeling for Duvenhage Rabies Virus Glycoprotein G

Copyright: © 2017 Albagi SOA, 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. Immunoinformatics-Peptide Driven Vaccine and In silico Modeling for Duvenhage Rabies Virus Glycoprotein G Sahar Obi Abd Albagi1*, Omar Hashim Ahmed2, Manal Abdalla Gumaa3, Khoubieb Ali Abd_elrahman4, Ahmed Hamdi Abu-Haraz1 and Mohammed A. Hassan1 1Africa City of Technology, Sudan 2Department of Pharmacology, Faculty of Pharmacy, International University of Africa, Khartoum, Sudan 3Medical Microbiology, Khartoum University, Sudan 4University of Medical Science and Technology, Sudan

protected against only the Phylogroup I lyssa viruses. Similarity in glycoprotein antigenic sites leads to cross protection between some Lyssavirus species by neutralizing antibodies [9]. Even with the presence of vaccines in effective form, rabies infection stays to kill as a minimum 55,000 of people every year, and many of the animal fatal infections. The highly mortality rate is reported in India with more than 16,000 deaths in just 2010 year 15. So, studying of new vaccines is important to produce an effective and low cost one instead of egg-based or cell-culture derived vaccines [15,16]. Vaccinating animals limited human cases, which transferring through dogs, bats and raccoons, in many countries [17,18]. Initial vaccines virus of rabies were formed in living tissue then substituted by vaccines produced in embryonated eggs and tissue culture. Vaccines of viral vector or DNA have been used successfully with greatly protection volume against the infection of rabies [19,20]. However, no one of these vaccines is lead to product a licensed for personal used until now 4. Furthermore, the immunization with P-G gene of rabies virus did not protect against viruses from genotypes 3 and [21][22][23].
Post-exposure prophylaxis consists of immediate cleansing of the wound after a minor animal injury [24,25], then taking of human rabies immunoglobulin and human rabies vaccine. If the people vaccinated before, fewer vaccine doses and no rabies immunoglobulin are given. If the patient is immunocompetent or immuno-suppressed or unvaccinated, Post-exposure prophylaxis is highly effective if it is begun soon after exposure. When rabies symptoms develop there is no ideal treatment is recommended, some treatments with aggressive therapy have a very high risk of failure, like vaccines, antiviral agents, antibodies to rabies virus, ketamine and the induction of a therapeutic coma are ineffective [14]. The replication cycle is depended on the highly Conservancy of the genome [26,27]. And due to the significance of the envelop matrix-protein M and glycoprotein G in virus infection and virus release [28]. G protein has the major role in the pathogenicity of genotype one lyssa viruses [29,30]. The outer membrane of glycoprotein G is only protein which is related in virus entry besides give a protective immune response [31]. Vaccines are designed for treatment of post exposures to the infection [32]. In animal studies, there is no combination treatment was done in animal models. Some limited access to post exposure prophylaxis may be expected. Put in Duvenhage Virus from a Bat in Kenya, the Treatment failure with Coma-Induction with Ketamine and Antiviral Drugs [33]. So, in silico modeling of epitopes protein remains would help in manufacture of peptide vaccine, which is highly immunogenic and with low allergenic effects [34,35]. Our goal is to project a vaccine for Duvenhage virus using peptide of its envelop glycoprotein G as an immunogenic part to encourage protective immunity.

Sequence of protein recovery
A total of 26 Africa strains of rabies Duvenhage virus strains' glycoprotein was retrieved from the NCBI database (https://www.ncbi. nlm.nih.gov/protein) in November 2016. These 12 strains sequences were recovered from many parts in the world (consist of 11 were isolated from South Africa, Zimbabwe and one from Holland-Netherlands). Glycoprotein strains were Retrieved and area of isolated or collection and their accession numbers are listed in Table 1.

Determinant of conserve regions
The aligned of retrieved sequences was done to obtain conserved regions using multiple-sequence alignment (MSA). Sequences aligned were supported by Clustal-W as Applier in the BioEdit-software, version 7.0.9.1 (Hall, 1999) to find the conserved regions in all 12 retrieved in Rabies spike glycoprotein G sequences Figure 1. At the same time; then   analyzed these candidates epitopes by various prediction tools from Immune Epitope Database 'IEDB' examination resource (http://www. iedb.org/) [36,37].

Prediction of B-cell epitopes
The B-cell epitope is the part of an immunogenic, which determine with B-lymphocytes. B-cell epitope was categorized by hydrophilic reachable and in a turn of beta region. Thus, the classical susceptibility scale methods and hidden programmed for Markov model software's from IEDB as resource of analysis were utilized for the sequential tests [37,38] (Figures 2-3).

Linear Prediction of B-cell epitopes
BepiPred-test from immune epitope database (http://tools.iedb. org/bcell/result/) [39] it has been done as linear prediction of B-cell epitopes when selected from conserved region with a default threshold value of 0.148.

Surface accessibility prediction
via prediction of Emini surface accessibility tool of IEDB (http:// tools.iedb.org/bcell/result/) [40]. Epitopes of the surface accessible were predicted from the region they were conserved and the default threshold holding value was 1.0.

Antigenicity prediction of epitopes sites
The antigenicity method of kolaskar and tongaonker was proposed to determine the sites of antigenic epitopes with a default threshold value of 1.042, (http://tools.iedb.org/bcell/result/) [41].

Binding predictions of MHC class I
The peptide binding Analysis to Major Histocompatibility complex class I molecules was evaluated by the IEDB MHC-I estimated tool at (http://tools.iedb.org/mhci/), MHC-I peptide molecules presentation to T lymphocytes submit to a number of steps. The linking of fragment peptides to MHC molecules step was proposed its reached by Artificial neural network (ANN) method [42,43]. Previous to prediction, all lengths of epitope was set as 9 mers, all internationally conserved epitopes that binding to alleles at score less than or equal to 500 halfmaximal inhibitory concentration (IC50) was obtained for additional analysis [43].

Binding predictions MHC class II
Binding analysis of peptide to Major Histocompatibility complex class II molecules was evaluated by the IEDB MHC-II estimated tool at (http://tools.iedb.org/mhcii/) [44,45]. In this step, references of human allele were used. In MHC class II groove has the capability to binding  to peptides with various lengths. This binding variability makes difficult prediction as fewer accurate [46]. IEDB take a five methods of prediction for MHC II as tool of prediction, we used the artificial neural networks method 'NN-align' that allows for simultaneously identifying binding core epitopes of the MHC class II and affinity to binding. Method of the consensus uses the medium ranking of the three methodologies as the final score of prediction [47]. All epitopes were preserved as conserves that binding to various alleles at equal score or less than 1000 half-maximal inhibitory concentration (IC50) it was selected for additional analysis (Figures 4-6).

Population coverage calculation
All probable binders of MHC I and MHC II from rabies virus of Africa duvenhage strain Glycoprotein was evaluated the coverage of population compared to the whole world and Africa population with a selection of MHC-I and MHC-II related alleles by calculation   of the IEDB population coverage tool at http://tools.iedb.org/tools/ population/iedb_input [48].

Homology modeling
The modeling has been done by Raptor X structure prediction server. Which is submitted the reference sequence of duvenhage strain on 2016-12-21. The files result is received at 2016-12-22, which containing the predicted 3D model and image are attached noreply@ raptorx.uchicago.edu.
Then we used visualized tool of UCSF Chimera (version 1.8) to get a 3D-structures of protein, this tool was presently available within the Chimera package and accessible from the chimera web site (http:// www.cgl.ucsf.edu/cimera) [49].

Docking
MHCI alleles Epitopes was predicted to get bind with fewer average of percentile-rank were designated as the ligands, PEP-FOLD as online tool of peptide modeling it has been used to given epitopes structure [50]. The MHCI allele's 3D structure was achieved from PatchDock algorithm stimulated by object recognition and segmentation techniques of image used in Computer Visualization. It tries to match two parts by selection one part and search for the matching one as complementary [51]. The FireDock server statements used for problem of refinement in protein-protein docking resolutions; The method at the same time targets the flexibility problem and scoring of solutions was produced by docking algorithms as fast rigid-body [52][53][54].

Prediction of B-cell epitope
The Duvenhage rabies glycoprotein (GP) reference sequence was examined by Bepipred linear epitope prediction, Kolaskar and Tongaonkar antigenicity and Emini surface accessibility methods in IEDB, to confine the epitope linearity, which is surface rise and the immunogenicity respectively supported test (Figures 7-9).

Cytotoxic T-lymphocyte epitopes Prediction and interaction with MHC-1
The duvenhage rabies virus reference sequence of glycoprotein G was examined using IEDB MHC-1 tool of binding prediction to predict T cell epitopes to given interact proposed with types of different MHC I alleles built on ANN-align by (IC50) ≤ 500, the list of epitopes were a conserved and their MHC-1 alleles corresponding are shown in (Tables  2-4).

T helper cell epitopes Prediction and interaction with MHC II alleles
The Duvenhage rabies virus reference sequence of glycoprotein G was examined using IEDB MHC II tool of binding prediction based on NN-align by half-maximal inhibitory concentration (IC50) ≤ 1000; In supplementary- Table I were shown the list of all epitopes and their correspondent binding MHC II alleles while the list most promising epitopes that had Required to binding affinity with MHC II alleles along with their positions in the glycoprotein G Duvenhage rabies virus were presented in (Table 5).

Population Coverage Analysis
The test of Population coverage was done to observe the all epitopes world coverage binds to MHC1 alleles and MHC11 alleles also to joined

Population coverage for MHC1
All epitopes in population coverage results binding to MHC1 alleles of Duvenhage rabies virus was shown in supplementary- Table (II) and the results of population coverage for the most candidates promising three peptides was shown on (Table 6).

Population coverage for MHC11
All peptides in population coverage results binding to MHC11 alleles with total HLA hits and their coverage were shown in supplementary- Table (II) and the results of candidates three epitopes promising were presented on (Table 6).

Population coverage for both MHC1 and MHC11 alleles
The performed of this test to candidate most one promising epitope (YFLIGVSAV) and it was establishing to bind 12 alleles with the world coverage of 99.36% (Figures 10-12).

Docking
Many servers used as the complementary tool to give final resolute by achieved two input options for ligand and receptor of (2 PDB) files with each docking solution, The output is a list of ranked in all the input     resolutions. All epitopes were docked Epitomized a Molecular Structural Shape, which have good affinity to binding with lowered binding global energy, the method we are used at the same time targets to problem of flexibility and recording of solutions created by fast rigid body docking algorithms. This gave a set of up to 15 possible docking candidates of three promising peptides bind with MHC1 Homo sapiens. In Table  7 selected three best result with lowest binding of global-energy score necessary for all epitopes besides the attractive and Repulsive softened van der Waals interactions VdW. The developed complex structure is created for up to three candidate's low-energy for individually structure. Epitope FVGYVTTTF is showing the best one binding energy Figure 10.

Discussion
The importance of selecting Duvenhage strain 'genus Lyssavirus' in the present study was based on many contributed including high mortality rate and the used of traditional method to produce rabies vaccine without deeply testing studies on the immune system response can be selected a conserve peptide to give a universal covering of population and who have made a strong binding between MHC homo spins with these peptides. Some study that supports previous studies indicating all African Lyssavirus species that can be lethal when inoculated via the i. m. route and that the pathogenicity of a Lyssavirus species should not be concluded based on a single isolate [53].
In this study we used, the immunoinformatics tools to predicate the candidate conserved peptides to give a good binding and faster recognized by immune system, especially in B and T lymphocytes in adaptive immune response to produce specific antibodies against Duvenhage virus. All protein's sequences of glycoprotein G were selected from NCBI website were aligned to get a conserved region that representative Duvenhage Lyssavirus isolates is shown in Figure 1. Then used reference sequence of target strain ''Duvenhage rabies virus glycoprotein.
G'' to be subjected in Bepipred linear epitope prediction test, Kolaskar and Tongaonkar antigenicity test and Emini surface accessibility test using IEDB, to define binding affinity to B cell, presence in the surface and to exam the immunogenicity.
In Bepipred test, were presented of 23 conserved epitopes that have the affinity of binding to B cell while there are 17 epitopes surface predicted that according to Emini test for surface accessibility and 18 epitopes presence antigenic when detected by Kolaskar-Tongaonkar antigenicity test but only three epitopes established as overlap all performed B Cell tests (AHYK, YTIPDKL and SLHNPYPDSH). Furthermore, the reference Duvenhage rabies sequence of glycoprotein G was tested by IEDB-MHC 1 of bind predicted tool to obtain the predict result T cell epitope was proposed to interact with various types of MHCI alleles were subjected on Artificial neural network (ANN) through the half-maximal inhibitory concentration (IC50) ≤ 500; 23 conserved peptides were predicted interact with various MHC-1 proposed alleles. The peptide FVGYVTTTF from 93 into 101 had greater affinity to interact with 9 alleles ( The analyzed for reference glycoprotein G strain was performed via IEDB MHC II prediction of binding tool established with halfmaximal inhibitory concentration (IC50) on NN-align ≤ 1000; around 39 conserved epitopes remained predicted found to interact with alleles of MHC-II, The peptides LLESSVVPL, LLQQHIELL and YFLIGVSAV required the affinity to bind the maximum number to these alleles of MHC11.
The Results in World population coverage for total binding epitopes to MHC1 alleles were 99.51% and the peptides have most promising were FVGYVTTTF, FVDERGLYK and FLIGVSAVA with population coverage in the world 56.71% and total HLA hits 9, 5 and 4 respectively.
World population coverage result for totally epitopes that take binding affinity to MHC11 alleles was 99.99% while population coverage in world of the best promising three epitopes LLESSVVPL, LLQQHIELL and YFLIGVSAV was 99.68% with HLA hits 16, 18 and 18 respectively. The peptide YFLIGVSAV shows excellent population coverage outcomes for MHC1 and MHC11 alleles together of (99.36%) with total HLA hits 18 different alleles.

Recommendations and Conclusions
In the current study, we achieve that to our information this was the first one study to suggest peptide vaccine through previous steps for Duvenhage rabies virus up to now. Furthermore, the immuno-informatics methods representing a smart in silico tools, it's gives a highly supported outputting epitopesdriven vaccines for communicable pathogens could be extended to efficacy trials stage in humans, after demonstration a several effective result in animal models must be doing as the next step in our vision. In the summary, we are in the procedure to developing a vaccine of epitope-based through the present study to get a trial promising epitopes vaccine; in this article designates our in silico progress through many pathway to reach our goal. We trust that our rabies vaccine of immunoinformatics-driven peptides development approach could have been many good expected over other approaches: (1) rapidity to response, (2) safety and low adverse allergic significant, (3) broadimmunogenicity (delivery of conserve multiple epitopes, derived a specific protein, faster in recognition when predicated on B cell, MHC1, MHC11 tools) and (4) world coverage with good docking to MHC homo spins properties.
So among the different test for peptides to both B cell and T cell this study planned an interesting T-cell epitope (YFLIGVSAV) that has a strong bind affinity to MHC1 and MHC11 alleles together, and it can be binding with 18 distinctive alleles with population coverage of the world (99.36%), indeed this candidate epitope-based vaccines are adept to inducing more potent reactions than vaccines of the whole  protein. We also trusted that the methods defined here, which will lead to development of a multi-epitopes Duvenhage vaccine, could being a step in the right direction over the development of a kind of safer, good effective bio defense with this promising candidate epitope vaccines. We recommend to doing further studies to propose a peptide driven vaccine for the other strains of Rabies Lyssavirus, here will be a possible to find public conserved candidates epitopes for multiple strains. In addition, we recommend applying these peptides to evaluate the in vitro and in vivo methods.