Computational modeling and analysis of prominent T-cell epitopes for assisting in designing vaccine of ZIKA virus

Article history: Received on: 07/04/2017 Accepted on: 09/05/2017 Available online: 30/08/2017 The Zika virus disease or Zika fever, regularly shows no or just mellow side effects, like an exceptionally gentle type of dengue fever. It spread eastwards from 2007-16 over the Pacific Ocean to the Americas, whereas in 2015 to 2016, Zika virus scourge achieved epidemic levels. In this study, the antigenic epitopes of Zika virus (ZIKV) were predicted and modeled. The highest binding scorers among the predicted ones and their correlating MHC class II alleles were further subjected to binding simulation studies. Immunoinformatics tools were applied to analyze the viral antigenic proteins that may be helpful in designing vaccine for ZIKV. The promiscuous epitopes of MHC class II were predicted from the viral proteins using ProPred, an immunoinformatics tool. The chosen epitopes and MHC alleles were modeled molecularly using PEP-FOLD3 and CPH model 3.2 servers respectively. The viral glycoprotein having epitope/peptide YRIMLSVHG bound to DRB1*01:01 MHC class II allele demonstrated the most noteworthy binding score. The anticipated peptide has high possibility of inducing T cell mediated immune response and it might be helpful in designing vaccines based on epitopes after continued future trials.


INTRODUCTION
Zika virus (ZIKV) is an arbovirus of the genus Flavivirus and the family Flaviviridae (Lanciotti et al., 2008).It is a positive-sense single-stranded RNA which incorporates a few other mosquito-borne infections of clinical significance such as dengue virus (DNV), yellow fever virus (YFV), and West Nile virus (WNV) (Lanciotti et al., 2008).The other individual from its clade, the Spondweni virus, is its nearest relative (Lanciotti et al., 2008;Kuno and Chang, 2007).The genome of Zika virus contains 10,794 nucleotides which encode 3,419 amino acids Kuno and Chang, 2007).The Zika virus is similar to other flaviviruses, that is, it is made out of 2 non-coding regions (3' and 5') by which a polyprotein is encoded and a single open reading frame (ORF) is flanked.This encoded polyprotein gets cleaved into a triad of structural proteins, viz.precursor of membrane (prM), envelope (E) and capsid (C); and seven non-structural proteins (Kuno and Chang, 2007).The day-time active mosquitoes, predominantly of the Aedes (Stegomyia) genus, such as A. aegypti and A. albopictus, transmit the ZIKV (Malone et al., 2016).Aedes mosquitoes are generally circulated all around the globe, and most of the mosquito species inhabit warm subtropical and tropical local natural surroundings (Kraemer et al., 2015).A. albopictus does not yet give off an impression of being a noteworthy vector of ZIKV.But we can't ignore its role in the Gabon flare-up of 2007, it's widespread all through the US and absence of ZIKV when this particular species Aedes sp. was limited (Kraemer et al., 2015).
Inspite of various researches carried for treating ZIKV infection, there is no proper standard treatment available till date for curing this deadly infection.The administration is steady and incorporates rest, liquids, antipyretics, and analgesics.Headache medicines and other non-steroidal anti-inflammatory medications should not be taken until dengue is treated completely as this possesses the danger of hemorrhage (Kuno and Chang, 2007).
So vaccine development against the ZIKV is very essential.The most antigenic part of a virus is its surface or envelope proteins which are often considered as a great contender for inoculation.The adaptive immune response primarily targets the envelope protein which mediates the viral entry which makes them crucial for vaccine development (Cerdeño-Tárraga et al., 2003;Trent and Qureshi, 1971).Over the past few years, immunologists have discovered the epitopes which can be identified by T cells and B cells are different and this has resulted in designing of more potent candidates for vaccine.Intercellular foreign antigenic peptides recognition majorly involves Major histocompatibility complex (MHC) and hence it takes part in developing both cell-mediated and humoral immune responses.Antigen-presenting cells (APCs) contain T cells on their surface which recognize antigenic fragments only when the exposed surface MHC molecules of all vertebrate cells are combined with them (Shekhar et al., 2012;Mohabatkar and Mohammadzadegan, 2007) .
The heterodimeric glycoproteins, MHC molecules, induce the activation of T cell by presenting a varied peptide set on the cell surrounding surface (Viret and Janeway, 1999;Tambunan and Parikesit, 2011).In a population with various alleles, the binding regions of an allele may not trigger the immune response due to high polymorphism of MHC molecules.So it's more necessary to identify viral peptides that are promiscuous and possess binding capacity with multiple MHC alleles (Tambunan and Parikesit, 2011).
Epitope or peptide-based vaccines are more specific, worthwhile, easily producible, safer and more time efficient than the conventional vaccines.The epitope or peptide-based vaccines can deliver high immunogen dosage at lower cost (Von Hoff et al., 2005;Tang et al., 2012).In order to design a potent synthetic peptide vaccine candidate, In Silico modeling techniques and immunoinformatics approaches have been applied including the use of bioinformatics software and tools which are based on various machine learning programs and other statistical approaches.The active candidate for vaccine must possess antigenicity and it must be responsible for pathogenicity (De Groot et al., 2002;Verma et al., 2011)

METHODS AND MATERIALS Retrieving Zika virus envelope glycoprotein sequence and Protein antigenicity determination
Zika virus envelope (outer membrane) glycoprotein sequences were extracted in FASTA format from UniProtKB (http://www.uniprot.org/)(Gaunt et al., 2001).Theoretical isoelectric point (pI), molecular weight and amino acid compositions were computed by using the Expasy's ProtParam server for all the 40 glycoprotein sequences (Chandra et al., 2010).These sequences were then investigated with VaxiJen (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html) with default parameters to discover its antigenicity, to predict the protective antigen as a vaccine (Doytchinova and Flower, 2007).Amongst all the antigenic proteins, the one having the highest antigenicity score was chosen to carry out further studies.

T-cell epitope-bound class II HLA expectation and identification
To identify the promiscuous T cell epitopes various distinctive prognosis procedures have been utilized.Epitopes are known as "promiscuous" when more than one MHC allele and more than one T cell clone can identify different T cell epitopes.

Modeling favorable epitopes and MHC II alleles and their computational validations
The GP epitopes of ZIKV were modeled at PEPFOLD3 (http://bioserv.rpbs.univparisdiderot.fr/services/PEP-FOLD3/) for each MHC allele.In the Protein Data Bank (PDB) the 3-D structures of MHC alleles were inaccessible, so in order to secure their 3-D coordinate's files CPH model 3.2 server which uses their respective structural templates, was used.Afterwards, numerous online servers were used for verification and structural analysis of MHC alleles' structure.The servers being used are Errat, ProQ, ProSA, RAMPAGE, etc. Errat (http://services.mbi.ucla.edu/ERRAT/) is an algorithm for verification of protein structure which is used specifically for analyzing non-bonded interaction statistics between various atom types (Colovos and Yeates, 1993).ProQ (http://www.sbc.su.se/~bjornw/ProQ/ProQ.cgi) is a protein quality prediction tool that predicts the nature of a protein model on the basis of MaxSub and LG score, the two quality measures (Wallner and Elofsson, 2003).ProSA (https://prosa.services.came.sbg.ac.at/prosa.php) is a web-based tool which analyzes the structure of a protein (Wiederstein and Sippl, 2007).It determines the model quality both local and overall (Z-score) and for a particular protein structure it calculates an overall quality score.RAMPAGE (http://mordred.bioc.cam.ac.uk/~rapper/rampage.php) is a web tool for assessing protein model on the basis of Ramachandran plot (Lovell et al., 2002).

Docking prediction of epitopes and alleles
After modeling the structure, chosen peptides and alleles were docked using rigid docking method with the assistance of PatchDock followed by refinement through FireDock to decide the values of binding score.

Epitope-HLA allele molecular dynamic simulation
The molecular dynamic simulation was performed by utilizing Desmond (Klepeis et al., 2009;Raj et al., 2015).Initial co-ordinates were obtained from the best peptide-allele complex for dynamics simulation process.The system building process was carried out by adding Simple Point Charge water model.Then the system was neutralized with the addition of counter ions followed by the system minimization step using Broyden-Fletcher-Gold farb-Shanno LBFGS algorithms.The whole system was further subjected for MD simulation studies for 20 ns at 300K & 1.0325 bar pressure and the RMSD was recorded to check the stability of the above said complex (Raj et al., 2016;Kamal et al., 2016).

Recovery and prediction of antigenic Protein
Sequence analysis of envelope glycoprotein segregated from Zika virus was done on the basis of the structure.Fasta format of an aggregate of 40 polyproteins was recovered from UniProtKB.The physiochemical properties for all these 40 poyproteins hasve been summarized in Table 1.The protein having the highest immunogenicity was predicted using VaxiJen.The envelope protein, UniProtKB ID Q91KX7, was predicted with highest antigenicity with an aggregate score of 0.6269 at a threshold value of 0.4.This prediction corresponds to a previous observation where immunogenicity of the envelope glycoprotein was found (Roehrig et al., 1983).

Prediction and analysis of MHC Class II binding peptides
Based on the scores generated by the Propred (Table 2), peptides showing maximum binding with the same alleles of MHC II were taken.3).
Further we determined the contribution of the hydrogen bonds to the global binding energy utilizing FireDock, it was found to be -2.40,-2.13 and -0.78 respectively.

Fig. 1 :
Fig. 1: The graphical representation of ERRAT result of DRB1*01:01 allele.On the error axis, two lines are drawn to show the certainty of which it is conceivable to reject regions that surpass that error values.

Table 1 :
Physiochemical Properties of proteins.

Table 3 :
Docking and post docking results of epitopes with alleles.